The unsaid bit last year after my hospital admission was the diagnosis: locally advanced colon cancer. By “locally advanced” they meant it was huge, had been there ages (maybe 10 years by my count) but by some stroke of fortune hadn’t at that point spread anywhere else.

Aggressive chemo followed, which itself led to several A&E admissions, notably for sepsis, and even gave me diabetes. Surgery followed in April to remove the bastard, and then some more chemo. Today I am back, rid of the diabetes, and following further restorative surgery which in my head closes the chapter. The oncologist disagrees, says I’m stage 4, and is very pre-occupied with answering the question “well, where is it then?”. But in my head this chapter is closed, and I’m determined it won’t be reopening.

An experience like this teaches you a lot. I’ve learned loads about myself - fair to say I’ve changed too - and I’ve learned a hell of a lot about other people.

First and foremost, I remain irreconcilably apoplectic at the bloody Government. Being told, as I effectively was by a junior doctor, that I had only a few weeks (which wasn’t true then and isn’t true now BTW) is one thing. Coming home to a young family with that news, and finding my first priority was making sure they weren’t going to be made homeless because I’d had the insolence to try and create something, is quite another. Their vindictive, spiteful policies destroy lives and I won’t forget that. I won’t forget Aviva, who cancelled my life insurance after they cocked up the direct debit (and allocated my ongoing payments to my wife), or Airwallex, who blocked all our cashflow while I was in intensive care, or the competitor whose third offer to acquire Simwood was an undeniably predatory 50% of their second, with the only negative change our side being my health. I won’t forget any of it.

What I will dwell on is the wonderful selfless people I’ve met along the way. I don’t just mean medical professionals, and I’ve seen from the very best to the very worst there, but humanity generally. My (now) wife and family have of course been totally supportive but in a very business as usual kind of way. On getting back from the latest surgery my youngest asked three questions in succession with no difference in the importance and clear linkage: “Is the surgery done now?”, “Is the cancer gone?”, “Can we go to a water park?”. Simwood colleagues, led by Charles, have been truly amazing; I’ve not really had any time ‘off’ but that isn’t because I couldn’t. Business has carried on with very little, if anything, needing me. They’ve also somehow perfectly balanced letting me know they care with giving me as much space as I needed, and actually, no awareness or judgement of how much space that actually was. Then there’s the acquaintances who have become friends, rallying round to help my family with things I was unable to do, and of course the friends who stepped up without question. Last but by no means least, those of you amongst customers, professionals and industry friends who have just been there, checking in. I’ve long thought you can weigh your life by how many people show up at your funeral (cheery I know!) and an experience like this gives an insight into who turns up, including those I never expected to see. Thank you all so, so much.

And that brings me to myself, and therein, hopefully some insight for anyone else. I’ve been petrified of anything medical since I was a child (blame the dentist), which is a large part of why I was so late getting diagnosed. It was stupid! While I don’t advocate blind trust of Doctors (I am still default distrust), the fear of the process turned out, for me, to be far worse than the process. From getting faint at the prospect of a needle to some of the beastly things I’ve had done is quite the journey. I’ve also increasingly had people - medical professionals mostly - asking me how I do things because they think I’m some kind of freak. Last week, a sister took me down to the operating theatre and I then passed her in the corridor an hour or so after coming round, where she quizzed whether I’d actually had surgery. After lots of immediate expletives, I think, about getting back in bed, she came to see me several times over the week to talk about state of mind. I’d never had surgery before April so refusing pain relief and being up and about so quickly didn’t strike me as odd. Apparently it is. I don’t say this to sound like some sort of hard-man because I’m not, and I’m as perplexed as everyone (especially my wife) on the pain side of things, but I genuinely didn’t have any. I had exactly two shots of morphine, both of them because the staff made me test the button worked. I credit the surgeon and the anaesthetist with that more than myself, but it did make me realise the power of mind, and that’s what I talked about with the sister and the other staff who came to quiz me separately.

I’ve always been incredibly goal oriented but this was far and away the biggest. It would be easy to get sucked into the negativity, and frankly the Doctors push you that way, but in my head dying was not an option and wallowing in self-pity just felt like a slope to that. The journey was simple: chemo to knock the bastard back, cut it out, get on with life. It got extended a little with the need for a second lot of chemo and surgery but each time I went for surgery it was the final hurdle and the chapter was closing afterwards. The goal was always clear and nothing, and I mean nothing, was going to stand in the way of that. I knew I was tenacious and plenty of people will tell you I’m strong-willed, but after this, I think I’d even be afraid to get in my own way!

So where does that leave us? Well Simwood has gone from strength to strength this year. I’ve not been away but I’ve not been in the weeds, and that seems to be a good thing all round. My health is probably on a high with most people I meet telling me how well I look, and I feel 15+ years younger than I did even before diagnosis. The Doctors will keep looking. That’s fine: they’ll keep looking, I’ll keep turning up, and I’m not going to live in it. I know what and who I love and what I do it all for, and that feels a massive privilege. Some people never know all that until it is too late; I feel like I’ve got privileged insight.

Thank you again. I’m back baby!

What has been announced is not caller authentication, it is not STIR/SHAKEN, and it has nothing whatsoever to do with knowing who is ringing you. It is transport encryption: the signalling that sets a call up travelling over TLS instead of in the clear, and the audio travelling as SRTP instead of as plain RTP, which anybody with a tap and a copy of Wireshark can turn back into a recording in roughly the time it takes to make a cup of tea. It is the right thing to do, it has always been the right thing to do, and the only question is why it is being announced in the fourth quarter of 2026.

We turned TLS and SRTP on for outbound calls in January 2013, and by that December the capability was there on inbound too, on every number we deliver over SIP, which I wrote up at the time in TLS/SRTP and why you need them!. It has been free on every call since. Not a security bundle, not a premium tier, not a conversation with your account manager. Free, because charging somebody extra not to be intercepted isn't a product, it's a toll.

That was thirteen years ago. In the same few weeks, Jessops, HMV and Blockbuster all went into administration. Blockbuster, remember them? Thirteen years in which this has been so ordinary here that nobody internally thinks of it as a feature, it has never appeared on a price list, and the last time I wrote about it at any length I was arguing with people who thought it was a waste of money.

One platform is not a portfolio

The newsletter's standfirst promises "secure upgrades on our global voice portfolio". The item underneath it covers one platform. Global Communications Enablement is Gamma's international voice platform for service providers, launched in April 2025, and it is the only thing named in the announcement.

Gamma did not build it. Its own account of the launch says Coolwave, the Irish wholesale business it bought, "brought their own network capabilities, session border controllers (SBCs) and data centres", and that GCE is "just a portfolio name to bring all of those to light". What was actually bought is set out in Gamma's 2024 half-year report: net assets acquired of £7.5m, of which tangible fixed assets were £0.1m and technology intangibles £6.0m. Gamma's own accounting policy says what that technology line consists of: "software licences purchased from third parties" and "rights over network interface identifications". A hundred thousand pounds of kit and six million pounds of permission to use other people's.

One platform is not a portfolio. What the rest of it does, I do not know, and neither does anybody who read that newsletter, which is a peculiar thing to be true of a security announcement. Which products, and since when? Gamma has not said.

In 2017 I wrote a post called Encryption is "Pointless"!, named after what somebody in this industry had put to me in writing. "We know most other wholesalers don't offer it," I wrote then. "We also know 'me too' doesn't offer it so their shiny suits will proclaim that the market doesn't want it." "Me too" was not a category. It was a company, and people did ask me which one. The market didn't want it because almost nobody was selling it, and almost nobody was selling it because the crypto licence on a big vendor's magic box costs an order of magnitude more than the basic one, which is a point I made again last October and which nothing since has done anything to disturb. Why can't anybody just say that out loud? The licence was expensive, we didn't buy it, our customers' audio went across the public internet in the clear for as long as it took us to give a shit.

Unencrypted media is never abstract. It is readable by anybody in the path, and the audio is somebody ringing their GP, somebody ringing a solicitor, somebody ringing a helpline they would rather nobody knew they had rung. That is what is in the packets, and it is why I have been banging this drum for thirteen years while being told the market doesn't want it.

The same joke, twice

If this were the only thing that had arrived late I would put it down to a bad quarter.

In September 2024 I wrote Resilience is not a magic button, and the ask in it was not complicated: "we lean so heavily on DNS and encourage all customers to use fully qualified domain names (FQDNs) rather than direct IP addresses". Use a name and you have multiple A records, short-lived records you can update straight off your monitoring, SRV and NAPTR. "SRV is the daddy here," I wrote, "because it also enables all of the aforementioned yet gives a pre-configured failover and load-balancing schedule." A name is how the internet tells you where something is. It is the difference between a supplier being able to move your traffic around a problem and a supplier having to ring you up about it.

Gamma's own technical service description for SIP Trunks said, for years and in terms, that DNS capability, including SRV and A record lookup, was not supported, and that the customer's equipment was identified instead by a static public IP address assigned by the customer's ISP. That document sits in public, as published on the Crown Commercial Service's G-Cloud marketplace, where anybody can go and read it. Gamma's own developer site makes the same point rather more briefly: the SIP Trunks API specification published there gives you exactly one variable for telling Gamma where your equipment is, and it is an IP address.

Names did eventually arrive, on the carrier side, in the early part of last year, and the document explaining how to use one ran to thirteen pages. Thirteen pages, for a DNS lookup. Whether the same is now true of a SIP trunk, I could not tell you, and that is the second time in one post that the answer to "which products, and since when" is not written down anywhere a customer can read it.

The domain name system was specified in 1983. It is older than the company and older than most of the people who will read this. So the complaint is not that Gamma does not support names, it is the forty-two years it took. Same magic box, same licence economics, same decade of asking, same small announcement when it finally lands. Twice is a pattern.

What a buyout house is actually buying

On 1 September Gamma's board recommended an offer from Epiris, taking the company private, expected to complete in the first half of next year. I wrote about it at the time: a fund does not buy a business in order to own it, it buys one in order to sell it again, and the price it paid has to come back out of the business through price, through cost, or through both.

So look at what has actually shipped and ask where else the return could come from. Transport encryption on one platform in the fourth quarter of 2026. Hostnames on the carrier side four decades after the specification. An international voice business whose acquired value was, on Gamma's own numbers, four fifths licences and rights. Whatever Epiris is paying a premium for, on this evidence it is not the pace of development, because nobody underwrites a return on a roadmap that is delivering 2013 on a good day. Which leaves the price list and the cost base. The cost base is people, and the price list is you.

Then there is the other half of that page

Gamma has been ranked number three in the world for Software and Telecommunications in Newsweek's World's Most Trustworthy Companies 2026, produced with Statista. It is a perception study. It measures belief, not truth.

Newsweek's own ranking page carries the line "Are you on the list? Click here to learn more about the licensing options", and that link leads to Statista's award portal, where the logo, the certificate and the trophy are licensed products. Being trusted is free. The trophy is not.

A trust survey measures what people outside a building think is going on inside it, and it cannot measure what is actually going on inside it, because not one of the people surveyed had any way on earth of knowing whether the signalling was encrypted. So you can be the third most trustworthy telecommunications company in the world, on a methodology that is entirely honest about measuring belief, and announce on the same page that the audio was crossing the public internet in the clear until last month.

Same email. Sent to partners.

Four questions

They aren't for Gamma. They've simply had the misfortune of being the most recent example in my inbox, but they are not alone at all. Ring whoever carries your traffic and ask: does my traffic run over TLS with SRTP, right now, on every call in and every call out? Since when? Does it cost extra? And can I point my equipment at a name instead of an address?

If the answer to the third one is yes, you already have the answer to the first two. If the answer to the fourth one is no, ask what else is still waiting for a licence somebody didn't buy.

The same morning the Secretary of State for Digital, Culture, Media and Sport, Lisa Nandy, issued a public interest intervention notice under the Enterprise Act. Continuity of telecoms supply is not one of the public interest considerations that Act specifies, so the government has said it will lay an Order before Parliament as a matter of urgency to add new grounds: continuity of the supply of telecommunications services, and the protection of public services, critical national infrastructure and vulnerable customers. Incidentally, Pete is spitting feathers this morning, because he’s been telling the powers that be this exact thing for more than 10 years. 

The CMA has been asked to report by 19 October. BT says that pending the review, it and TalkTalk will operate separately and continue to compete.

That move happened inside an administration, which is not a process anybody chooses or deserves to find themselves in. We have friends and colleagues at TalkTalk and at PXC, people we have worked alongside and argued with for years and been glad to have on the other end of the phone. Our thoughts are with all of them this week, and with the families who got the news over breakfast.

What has actually gone

TalkTalk was a force for good in this market. For two decades it was the reason a consumer could ring their provider, say they had seen a cheaper price somewhere else, and have that sentence mean something. It made broadband a thing people shopped for rather than a thing people were allocated, and you did not have to admire everything about how it went about that to recognise what it did.

At wholesale it mattered more, and almost nobody outside this industry ever noticed. TalkTalk unbundled over three thousand exchanges when unbundling was hard, and then opened that network up to the channel, which is how a generation of small providers got wholesale access to products the incumbent would only sell you if you were already big enough not to need the help. What that business had been folded into by the end was PXC, selling to something approaching a thousand wholesale partners and able to reach around 98% of UK premises, its own network and everybody else's alike. Every reseller, MSP and platform in this country has at some point been quoted better terms because somebody in a meeting could credibly say "or we could go to TalkTalk".

It did not lose

What finished TalkTalk was its balance sheet. BT puts its revenue at about £1.2bn, and it was losing money doing it. Ares, the private credit house, is reported to have put more than £380m into the group since 2024. The reported bidders for PXC included an overseas group, a management buyout backed by a buyout house, and at one point or another a fund or two circling. Almost every candidate to rescue the independent alternative was the same kind of capital that creates the problem.

Scale stopped delivering margin in this industry some years ago, which is why we run a materially higher EBITDA margin than a peer a hundred times our size, and why acquisitions nobody then bothers to integrate destroy the efficiency they were bought to create. The buy and build model that produced most of the consolidation of the last decade bought businesses on borrowed money and ran them on interest cover, and I argued then that survivability deserves a premium regardless of size, because a customer whose supplier has stopped paying its debts cannot migrate once the service has already gone. Last month, when Gamma agreed to go to a buyout house, I wrote that a fund does not buy a business in order to own it, it buys one in order to sell it again, and that the price it paid has to come back out of the business through price, through cost, or through both. I would much rather those posts had aged badly.

A business with that revenue, a national network and two and a half million customers has gone to the one buyer in the country big enough to absorb it, because nothing else in the market was solid enough to catch it. The losers here are not TalkTalk's shareholders. They are everybody who was quietly relying on TalkTalk existing.

The road here

None of this arrived out of nowhere, and the published record is there without anybody needing the hindsight that everybody has this morning.

In July 2021 Openreach notified the Equinox offer: FTTP pricing conditional, in Ofcom's own words, on the volume and range of services purchased, and running for ten years. We responded to the consultation that August. That September Ofcom decided to take no action. CityFibre appealed, the Competition Appeal Tribunal dismissed the appeal the following summer while urging Ofcom to keep a close eye on what the offer did to the market, Openreach came back with Equinox 2, and in 2023 Ofcom again decided to take no action.

Equinox made Openreach cheaper for TalkTalk, and TalkTalk was one of the largest buyers of Openreach FTTP in the country, so on the day it landed it was good news for them. That is the half anybody can see. The other half is what a ten year discount conditional on volume does to the alternative. TalkTalk had built things before. It unbundled three thousand exchanges. It built FibreNation and sold it to CityFibre in 2020. A company that had been a builder was now offered a decade of cheap access on condition it kept buying, and buying got cheaper exactly as building got harder to justify to a board. It took the discount. It never built again. By this year it was reportedly paying Openreach £60m to £80m a month and had nothing of its own to move onto.

That is what I objected to at the time, and I put it then as BT being able to introduce a loyalty scheme to kill off infrastructure competition with the regulator's blessing. Pete made the same point from the other end that December, as CityFibre's appeal went in, observing that Equinox had already caused at least one potential new entrant, us, to decide against entering the market at all. The objection was never that the discount was too generous. It was that a ten year discount conditional on volume delivers exclusivity without an exclusivity agreement ever being signed, and it does that to the buyer just as surely as to the builder. The instrument that made TalkTalk's inputs cheaper is the same instrument that left it with nothing of its own to move onto, and that second half took five years to show up.

Last Monday Ofcom blocked an Openreach commercial offer for the first time, finding it not fair and reasonable because a reasonably efficient rival might be unable to match the charges while recovering its costs. I welcome it, genuinely. It is not a vindication of anything I said in 2021, because it points the other way: that was a discount blocked to protect the people building networks, and blocking a discount raises what a buyer pays. Used in 2021 it would have made TalkTalk's input costs higher, not lower. Ofcom allowed the other offers in the same statement. What I would observe is that it took five years of this argument to produce a first intervention, and that it arrived seven days before it was too late.

No one of those decisions put TalkTalk into administration, and there is no evidence that any of them did. TalkTalk's last filed accounts, to February 2025, show revenue down from £1.51bn to £1.41bn, a statutory pre-tax loss of £465m and 420,000 broadband customers gone. That is a retail failure with a leverage problem on top, and Equinox is not in it. What Equinox is in is the answer to the other question, which is why there was no alternative left standing for any of it to land on.

The £100m is money Openreach was owed and will not now receive, and BT puts it in its own release as part of the cost of the morning (although Pete will be keeping a watchful eye on how much BT tries to manipulate its regulated products to recover this loss from everyone else). Openreach was made a legally separate company in 2017, with its own board and its own governance, so that the group's commercial interests and the regulated access everybody else has to buy could not be weighed against each other by the same people. The debt was owed to the access business. The rescue was carried out by the group. Both are in the same transaction, announced in the same release, before breakfast.

BT's part in it

BT probably had very little choice about this morning. There is no supplier of last resort regime in telecoms of the sort energy has, but there is an industry process by that name, and there is now a government willing to say in a published notice that the failure of a provider this size is a risk to life and to public services. Nobody was going to let two and a half million lines go dark, and there is exactly one organisation in this country with the scale, the network and the balance sheet to absorb them over a weekend. Once TalkTalk was going to fail, BT buying it was close to the only outcome available. I do not think BT wanted this and I do not think it is pleased with what it cost.

What is also true is that BT contributed to the condition that made the rescue necessary, and that BT benefits from the result. PXC was reportedly paying Openreach £60m to £80m a month, which made BT by a wide margin its largest supplier, on terms a regulator sets rather than a market. The pricing structures the regulator waved through for a decade are what made building the alternative irrational. And this morning the company that kept the incumbent honest at wholesale, and was a bigger force still at retail, became part of the incumbent. That is what happens when one organisation is at the same time the regulated supplier to its competitors, the largest creditor of its competitors, and the only available buyer when one of them fails. Structures produce outcomes whatever anybody intends, and this is the structure that separation was built to prevent.

The accounting for it will be worth watching. The opportunity in owning TalkTalk accrues to the group: 1.5 million retail customers, a million wholesale ones, and a national wholesale platform the group did not have on Friday. The costs do not obviously sit in the same place. The £100m of arrears is Openreach's money and Openreach is the entity that will not now receive it, and an Openreach cost is not an ordinary commercial matter. Where Ofcom controls what Openreach may charge, under the Telecoms Access Review 2026-31, it does it by reference to the costs attributed to the access business in BT's regulatory financial statements, which are prepared to a methodology Ofcom directs and against an assumed level of efficiency Ofcom decides Openreach ought to reach. A cost that lands in the access business is therefore a cost that everybody buying access can eventually expect to meet in a price. Which side of that line each part of this morning ends up on is a question for BT's next set of regulatory accounts, and it is a better question to ask now than after the charges are next set.

The competition argument

A Virgin Media spokesperson said on the day that the deal "has all the characteristics of a stitch up masked as a rescue deal in the public interest", and that days after the regulator proposed potentially blocking a nexfibre and Netomnia deal, rules might now be watered down so the incumbent could roll its tanks over competition. That is their characterisation and it is not mine. I do not think this was a stitch up. I think it was the only thing left, which is the worse problem of the two, because a stitch up can be unpicked and a dead end cannot.

Ares told officials, in a letter reported the day before the deal, that waving it through would damage the UK's standing with international investors. BT's answer is that millions of citizens and businesses were at risk and somebody had to step in, and the Secretary of State evidently accepted that at least far enough to reach for the public interest regime and to start writing new grounds into the statute to make it fit.

That one is for the CMA on 19 October and for ministers afterwards, and anybody telling you today what remedies or undertakings will come out of it is guessing. The review is being asked about the future of competition, and the answer to the question about the present is already in. The company that kept the incumbent honest at wholesale is now owned by the incumbent. Whatever gets decided on the 19th, that happened this morning.

Who was actually in the room

The two houses whose names have been all over the coverage for a fortnight sit on opposite sides of the table, and the reporting rarely stops to say which is which. I had them the wrong way round until this week. Ares is the incumbent money: a US private credit house and TalkTalk's controlling creditor, and lending on that scale for that long stops being lending somewhere along the way. Epiris is not that. Epiris is a UK buyout house and it was the rival bidder, backing a management buyout of PXC. Ares was indeed manoeuvring, and it was doing it over the consumer arm, where it is reported to have been the front runner to take retail itself before BT stepped in. The fight over the wholesale business was somebody trying to get in rather than somebody trying to stay.

What the rival bidder reportedly needed in order to win was BT. The ask was not a £300m write-off, whatever is in circulation. As reported it was a three month holiday on future Openreach charges worth somewhere between £200m and £250m, plus roughly £100m of arrears already outstanding, with Openreach offered a share of the proceeds of any future sale above an undisclosed threshold. BT declined, which was its own money and its own decision. Days later TalkTalk went into administration and BT bought both halves of it, and in doing so lost about £100m of those same arrears anyway, took on about £60m of trading losses for the rest of the year, and paid the transaction and administration costs on top. Whatever its reasons for refusing a deferral, the money did not stay in BT's pocket. A rescue by a rival needed the incumbent's agreement. The rescue that actually happened did not, and it cost the incumbent considerably more.

Epiris is also, and this part is an announced and recommended offer rather than reporting, the house taking Gamma private, expected to complete next year. So the same house is in both of this autumn's stories, and it lost one and won the other. If both land as they currently stand, the two largest independent wholesale platforms in this market finish up in opposite kinds of hands: PXC inside the incumbent, answering to a group strategy, and Gamma inside a buyout house, answering to a hold period and an exit that has to be planned from the first day of ownership. Neither of those is the independent whose name is on the contract you signed.

Retail choice was already narrower than it looked

The same thing has happened one floor down, at retail. BT Group's consumer division sells under three brands: BT, EE and Plusnet. Plusnet sits at the value end of the shelf and EE at the other, and both have belonged to BT for years. A household comparing those three on a price comparison site is choosing between three shop windows belonging to one company, and whether the household doing the comparing knows that is quite another matter. TalkTalk was one of the retail names that genuinely was not BT. This morning it joined them.

Enshitification is what tends to happen next in a market shaped like that. Once everybody is selling the same thing to the same people, squeezing the customer becomes the only route left to profit, which is where seven year contracts, hardware lock-ins and games with number portability come from. It is not anybody's moral failing. It is what the arithmetic rewards once there is capital to service and nothing much left to compete on.

What this leaves you with

If you run a CPaaS or CCaaS platform, an MSP, a reseller, or one of the AI voice businesses that has appeared in the last two years, the thing to do this week is not to panic. It is to draw your supply chain on one side of a page and then put a circle round every supplier that now belongs to the same group. For a good many of you that circle got bigger this morning without you touching anything.

Access, transit, numbers, voice termination, inbound: if those arrive from different logos that roll up to one parent, you do not have a supply chain, you have a concentration with extra invoices. That is a commercial exposure long before it is anything else. It decides what you can negotiate, what you can escalate, and what happens to you on the day that one group changes a price, retires a product, or decides to compete with you directly in the market you sell into.

Do that exercise to us and BT will be on the page, and making that circle as small as it can be has been the work of years here. The standing policy is to buy from BT only where there is genuinely nothing else to buy, and what that policy turns into in practice is direct interconnection with everybody who is anybody, all the major mobile operators and all the major fixed ones, so that wherever we can manage it the commercial relationship sits with the operator the customer actually chose. TalkTalk was a significant part of that. This morning it folds back into BT, and the list of places we have an alternative is shorter than it was on Friday.

That policy only works so far, and what limits it is the way number portability works in this country. Numbers port one at a time, and a ported number is still reached by sending the call to the operator that holds the range and having that operator hand it on, for the life of the number. So the network a call to a ported number reaches you from is not the network the caller is sitting on. It is the range holder. Where a number was originally BT's and has been ported away, BT is the donor network operator, and 100 per cent of calls to that number go to BT first before they ever reach the provider the customer actually chose. Ofcom's own market data puts BT's share of fixed originated call minutes at a little over 40 per cent, which is the number interconnection can do something about, because it measures where calls start. The donor mechanism is about where calls have to go. No amount of direct interconnection fixes that, because it is not an operator's to fix. We asked for it to be fixed in the Narrowband Review in 2020, and we said the move to IP was a prime opportunity to fix porting. It was not taken. That failure is Ofcom's, and we have said so many times.

The thing a buyer can actually act on is not whether the incumbent is somewhere in the chain. It is how many of the other names in the chain are also the incumbent, and who owns the company you ring when it breaks.

Competition rarely disappears on the day of an announcement. It disappears from a price list, a line at a time, over the two years afterwards, and nobody sends a newsletter about that.

I set out two tests for choosing who you buy from in July. Capability: have they already built the thing you need, or do they start when you ask? Alignment: does their growth depend on winning the customers you are chasing? Both got sharper this morning. A supplier that is part of the group you compete with fails the second test by construction, however good the people are, and there are a great many good people at BT.

Simwood is privately owned and founder controlled. The group at the top of us is our own, there is no fund behind it, people who work here hold shares and options in it, and our growth comes from our customers growing rather than from us becoming them. That is a structure rather than a slogan, and structures are the only promises worth anything in this business. Whether it matters to you is your call.

Nobody knows yet what BT will do with PXC. It may run it at arm's length for years. It may not. I do not know, and neither does anybody who is currently telling you they do. The argument I made about dinosaurs not surviving an ice age was about people still selling the lines and minutes of the nineties over SIP, and the point of AI as an extinction level event was that the asteroid is not aimed at anybody's job, it is aimed at the businesses too slow and too inefficient to move, which lose not to the technology but to competitors who use it. None of these outcomes is ever decided by what a business says it intends to do. It is decided by what the arithmetic it is standing in requires of it, which is why the interesting question about any new owner is not the promise made in the first twelve months but what the numbers demand of it in the third year.

So when you get a quiet hour this week, draw the page and count the logos. Then count the groups. There are fewer of them than there were yesterday.

The one that went this morning is why a generation of businesses got the chance to develop at all, and why a great many people are working from home today on broadband they would probably never have had if BT had been left to its own devices.

Thank you, TalkTalk. We will miss you.

As usual, these changes are colour-coded in our full rate files available through the portal as below. Where your account has custom rates, these are now reflected in the rate files. Remember, if you'd like more granular prefixes (inc. NPA-NXX) or a more compact rate file, we've got you covered.

Unlike other operators, we are not passing on UK origin surcharges, relying instead on our advanced technology to protect you and us from this industry's latest sleazy money-grab. We will maintain this position as long as possible so our customers do not need to worry.

Rate files Full rate sheets with codes are now only available through the portal or from our API for active customers. Once logged in to the portal you can find them in the "commercial" section.

The SMS A-Z rate sheet is now available on our portal or via API.

Please contact our sales team if you are not a customer and would like to see our rates.

Ancillary charges (inc. SMS & fax) Our charges for ancillary services can be found here.

100% SLA Our Service Level Agreement (SLA) for our UK carrier services customers is a guarantee of 100% Simwood service availability. Read more here.

MSA Our MSA is here.

Acceptable use policy Our AUP is here.

Hardware supply schedule Our Hardware supply schedule is here.

Hosted pricing Looking for Hosted pricing? Click here.

And what a seven years it's been!

When I joined Simwood in 2019 (well, Sipcentric before the acquisition merger), I was moving into my first role in tech. Before that, I’d worked in management and customer service roles at various companies for almost 10 years. I had plenty of experience with people, customers and the general chaos that comes with day-to-day operations, but technology up to that point had mostly just been a hobby.

I joined at Support Level 1, and I didn't really know what this career jump would hold for me. Would I enjoy it? What would the next step be? And would that next step be at Simwood?

It turns out there were plenty of opportunities to grow and progress within Simwood, and I enjoyed the culture, the people, and there were some pretty good benefits too. So I guess I got my answer.

Fast forward to 2022, and I moved into the Operations Manager role. That was a big change for me and gave me a completely different view of the business. Instead of just looking at the issue in front of me, I was looking more at the bigger picture and how things worked across the teams, where processes could be improved and, importantly, how all of that affected the customer. I learned some invaluable skills and gained some great insight during this time from working with the rest of the senior management team.

At the beginning of last year, I stepped down from the Operations Manager role. It might sound like a strange move to some people, but for me it made sense. I wanted to focus on improving one specific area of the business.

What area is that, I hear you ask?... Number portability (I know I’m a glutton for punishment).

There’s a lot that happens behind the scenes when a customer wants to move a number, and when everything works properly, you probably don't even think about it. When it doesn't work properly, however, it causes no end of frustration for you, the customer.

Porting has long been a key focus for our development team, and my role has been to build on that momentum, working closely with Operations and our brilliant devs to keep raising the bar for you, the customer. A big part of that has been the porting automation, which I hope you've already seen the benefit of.

While stepping down gave me the opportunity to take a slightly different approach to my career, it also made room for something else that was about to become rather important...

In October 2025, I became a mum after welcoming my daughter and headed off on maternity leave. This may explain why you haven't seen me in your inboxes lately. Did you miss me?

Maternity leave was a really nice break, as much of a break as looking after a tiny human can be, anyway! It gave me the chance to focus on being a mum and enjoy that time with my daughter, but towards the end I was always looking forward to getting back into work and some sort of routine.

And now that time has come.

I’m looking forward to getting stuck back in, catching up with everyone and seeing what’s changed while I’ve been away. Spoiler alert: not a lot when it comes to industry-wide porting changes! I was hoping that the porting industry might have made some great leaps forward while I was away. Sadly, I can confirm that it has not, so there is still plenty of work to be done!

So, to sum up for those who already know me: Hi again!

And for those I haven't met yet: I'm looking forward to it.

The power loom was a British invention. Britain had around 37,000 cotton handloom weavers in 1780 and 240,000 by 1820, because mechanised spinning had made yarn cheap and somebody had to weave it. Then the power loom matured: 2,400 of them in 1812, 100,000 by 1832, 400,000 by 1860, and the share of British cotton cloth woven by power went from 5% to 95% between about 1815 and 1845. By 1860 there were 10,000 handloom weavers left. The trade grew more than sixfold and then all but vanished, in the same century, in the same county. Weaving itself multiplied. The handloom weaver did not survive it.

I have since gone looking for whether that was a one-off, because it is so relevant to where we are today with AI. It is not a one-off.

James Bessen, an economist at Boston University, took the 271 detailed occupations the US Census used in 1950 and followed every one of them through to 2010. Plenty of them disappeared. Boardinghouse keeper went because nobody wanted boardinghouses. Telegraph operator went because the telegraph went. But in exactly one case could the disappearance of an occupation be put down mainly to automation.

One! In sixty years! It was the elevator operator.

Sixty years spanning the mainframe, the minicomputer, the PC, the spreadsheet, the industrial robot, the barcode, the ATM/cash-machine, the internet and the smartphone. Every one of them arrived with a confident, specific, published list of the jobs it was about to erase. The count after six decades is just one.

Bessen's own conclusion is the part worth noting: that period saw enormous automation, and almost all of it was partial. Machines took tasks. They did not take occupations.

The ATM/cash-machine

The bank teller was the prediction everyone was most sure about.

A machine that dispenses cash and takes deposits does the visible part of a teller's job, faster, at three in the morning, for no wage. Over 400,000 of them went into the US. The number of tellers employed went up, from roughly 500,000 around 1980 to roughly 550,000 by 2010.

The mechanism is the thing, not the outcome. Bessen measured that the number of tellers needed to run the average urban branch fell from twenty to thirteen between 1988 and 2004. A branch that needs seven fewer people is a branch you can afford to open in places that could never previously carry one, so the banks opened them: urban branches up 43%. More branches needed more tellers in total, and those tellers spent their day on relationships and selling rather than counting notes, which was the part the machine could not touch.

Same arithmetic as the power loom, a century later, in a service industry, which is the harder case. Automation makes a unit of output cheap. Cheap output gets bought in far greater quantity. Greater quantity needs more people, doing the part that was never the bit being automated. The mechanism has a name and it long predates any of this: William Stanley Jevons pointed out in 1865 that every improvement in the efficiency of the steam engine raised Britain's total coal consumption rather than lowering it, because cheaper coal found more uses than the efficiency saved. Jevons paradox is the loom and it is the ATM, and it has been running underneath every automation panic since.

The end of the teller story is less comfortable. The effect ran out. US teller employment peaked around 2007 and is down to roughly 347,000, with another 13% decline forecast to 2034. What finally did it was online and mobile banking, not the ATM: a change in what customers wanted to do at all, rather than a machine doing the same job cheaper. Thirty years of automation grew the job. A change in demand shrank it. Those are different causes and it matters enormously which one you think you are facing.

None of which is American, incidentally. The first cash machine in the world went into the wall of Barclays in Enfield on 27 June 1967, built by De La Rue to John Shepherd-Barron's idea, dispensing £1 notes against a cheque impregnated with carbon 14. By the time the Bank of England reviewed twenty years of the major British banks in 1991, ATMs roughly equalled branches and about 60% of retail deposit withdrawals went through them, and the Bank's own account of what that did to staffing was this: rapid growth in staff numbers over most of the 1970s and 1980s, with investment in technology "helping replace staff in back-office processing tasks, releasing them for front-office sales tasks". That is Bessen's finding, in the Bank of England's words, a quarter of a century before he published it. The British version differs in one respect. Our branch numbers were already falling while American ones rose, so we got the reallocation without the branch expansion that drove the headcount growth over there. The ending is the same as well: bank and post office clerks are down 83,000 since 2001, and nobody blames the cash machine for that either.

The jobs are not the same jobs

So the honest version of what I argued in July is not "your job is safe".

David Autor and his colleagues at MIT went through eight decades of US occupation data and found that roughly 60% of all employment in 2018 sat in job titles that did not exist in 1940. Among professional occupations it was 74%. Not jobs that had evolved - job titles that had to be invented, because the work had not previously existed to be described.

The British series runs longer and says the same thing. Deloitte's economists went through the census of England and Wales for every decade since 1871 and found agricultural labourers down from 6.6% of the workforce to 0.2%, and 200,000 people washing clothes for a living in 1901 against 35,000 in 2011, in a population that had gone from 32.5 million to 56.1 million. What replaced them was not nothing. Accountants went from 9,832 in 1871 to 215,678. There is now one hairdresser or barber for every 287 people in England and Wales, against one for every 1,793 in 1871. The caring professions went from 1.1% of employment to 12.2%, and the muscle-power trades from 23.7% to 8.3%. Weavers and knitters, since we started there, were down to 4,961 by 2014 from 24,009 in 1992, over a period when total UK employment rose 23%. And after a century and a half of all of that, the ONS puts the three highest employment rates in its entire 1861 to 2018 record in 1872, 1943 and 2018, all at 76% of the working-age population.

The work is safe and expanding. The job description is not.

And before anyone in telecoms takes any comfort from Bessen's number, I should point out that our industry ran the single largest counter-example in the data. Between 1920 and 1940 AT&T automated mechanical switching across more than half of the US telephone network and destroyed the telephone operator, which was then one of the largest occupations open to young women in America. In the cities that cut over, operator employment among 16 to 25 year old women fell by between a half and four fifths.

Feigenbaum and Gross went and looked at what happened to those people. The cohorts that followed were fine: the jobs came back as clerical and service work, including categories of work that had not existed before. The women already doing the job were not fine. A decade later they were disproportionately in lower-paid occupations or out of work altogether.

That is precisely the handloom weaver, and it is the reason I keep saying the technology is not what ruins anybody. Weaving boomed. Telephony boomed beyond anything the operators could have imagined. What got destroyed in both cases was the people and the businesses that kept doing it the old way while the world reorganised around them.

Where the new work actually comes from

Which leaves the question nobody in this debate ever asks. If automation keeps making things cheaper, and employment keeps going up, where is the extra work coming from? It cannot all be the same people buying more of the same thing.

Clayton Christensen named it in The Innovator's Solution in 2003: the growth does not come from competing for customers who are already buying. It comes from competing against non-consumption. The people who were not in the market at all, because the thing cost too much, took too long, or needed an expert they could not get hold of. Make it cheap and simple enough and an industry appears where there was not one, and that is where the unpredictable new job titles come from. You cannot forecast an occupation that serves a market which does not yet exist.

Which is the bit Jevons does not cover. Jevons is the same good, cheaper, bought in far greater quantity; non-consumption is demand that was never in the market to be counted.

It is also exactly why incumbents read all of this backwards. Christensen's real finding was never about technology. It was that thoroughly competent people, inside a business whose cost base demands a certain margin, will reflexively screen out the low-margin opportunity that creates the new market, and will explain very persuasively why they were right to. "AI is for our customers to build" is that sentence, dressed for work. My colleague Peter has written about the four AI strategies currently on display and they are all recognisable from this one insight.

The non-consumption in our industry is the conversation

Telecoms is sitting on one of the largest pools of non-consumption I have ever seen, and most of the industry is looking straight past it.

Email is indexed. Chat is logged. The CRM is structured and reportable. And the conversation, which is where customers state what they actually want, where compliance is met or missed, where fraud is attempted and where emergencies are declared, has historically produced a recording nobody ever listened to. Not for want of wanting to. Reviewing an hour of calls costs an hour of a person, so every business on earth reviews a token sample and bins the rest.

That is not a market being served badly. It is a market that has never existed. Almost every conversation your business has ever had went unexamined, and everyone quietly accepted that as the natural order of things.

Make it cheap and you do not get a smaller contact centre. That is the ATM mistake, it is the Klarna mistake, and it is sitting in a lot of board packs right now. What you get is a contact centre doing things it could not previously attempt at all: signals that fire while the call is still happening, guidance arriving in a handler's ear mid-conversation, the conversation becoming a record that compounds rather than a file nobody opens. The handler with an Operator watching their call is not a cheaper handler. They are a better one than they could ever be unaided, and the work that creates is new work, not recovered work.

We built it on that principle and we built it that way ourselves, right down to the internal plumbing nobody sees. Amplify the people. Do not replace them.

What the elevator operator was actually for

Come back to the one job that did go, because the reason it went is the warning, and it is not the warning people take from it.

The elevator operator did not lose to a cheaper elevator operator. They lost because the entire job was the interface. Their function was to be the human who pressed the button on the passenger's behalf, and the moment the passenger could press it themselves there was nothing underneath. Every other occupation on Bessen's list had judgement wrapped around a task. Automate the task and the judgement is still there, now applied across far more surface area than before, which is what happened to the weaver, to the teller, and to the generation that followed the switchboard.

So the question is not whether AI can do part of what you do. It can, and sixty years of census data says that has never once been enough on its own. The question is whether there is anything underneath. If your whole contribution is standing between a customer and a capability, passing the request along and adding a margin for the service of pressing the button, you are not a business facing disruption. You are an elevator operator, and I have written about your customers before.

Which is the point, and the reason July's post said what it said. Sixty years, one occupation eliminated. There is no equivalent register for companies, and if there were it would run to volumes. Automation has been extraordinarily kind to workers and utterly ruthless with firms, and everybody in this industry asking "how many people can this replace" is asking the question that has been reliably wrong since 1950, while ignoring the one that has been reliably right.

The asteroid was never aimed at your staff. It has your slower competitors' names on it, and if you stand still long enough, yours.

So: what is under your button?

To that end, we have one humongous project ongoing in relation to Carrier Services billing. Our call billing is mature and untouched but when it comes to other things, like billing number rentals, trunk charges etc. it is overly simplistic. Not the kind of simplistic that gets marketed as ground-breaking, but simplistic enough to have been a major constraint over the years. It just isn’t right that we can solve technical problems only to be constrained by billing agility and while the early temptation is always to deal with exceptions one-by-one, over time that becomes an unmanageable knot. So, for perhaps the last decade, we’ve just accepted the constraint. No more - newer services have so many moving parts and we need to be able to bill like a software company, not be wedded to the very deep but very narrow paradigm of telephony billing. That means bundles, allowances, multiple meters, bespoke subscriptions and all manner of good stuff. That is an ongoing project whose first outing is AI and WhatsApp (before taking over billing existing services), meaning we have a blank canvas. Yay!

Billing this stuff is complicated and when we first started on this journey a few years ago, we wanted to simplify it. It was so insanely complicated it was nigh on impossible to tell what a service was going to actually cost. That’s fair enough for a hobby project, not for building a business on. Along the way we’ve seen multiple providers dramatically change pricing and either row back or publicly apologise. We’re not even into enshitification yet, they seem to be just trying to optimise what they can grab. That isn’t our style but we’ve come to appreciate that simplification dramatically raises risk and either leads to all customers funding the edge cases, or us taking massive arbitrage risk. Best value comes from us exposing some complexity, offset by that which is eliminated by Simwood’s unique position (the only actual carrier globally offering these services on-net), and pricing in a way that enables our customers to price in a genuinely competitive way against those seeking to underpin a bubble!

So, our pricing has 5 moving parts, well, 6 if you consider the big bit that isn’t there - telephony. You don’t need to deal with a reseller of a carrier and pay 2c a minute for the privilege, because all our services are on-net on calls you already pass, on numbers you already have. You can use BYoC if you’re unlucky enough to have numbers locked into a dinocarrier or the mustard morons, but porting to Simwood is within your gift there.

As to the remaining, we’re offering 4 tiers of service and these are available to any production Carrier Services account and are unaffected by service level (Virtual Interconnect, Managed Interconnect, etc.) as the economic benefits of those are exposed through telephony and we want to encourage innovation and growth here. Each tier trades a monthly subscription against consumption economics, but even at the headline lowest commit level, you should be able to compete with any of the bubble-operators. At the top-end, we have completely bespoke pricing for those achieving scale or looking to bring on a 1,000-seat call centre or two. The per minute rate covers all agent and intelligence functions and there is a generous allowance baked into the subscription, the subscription being the cheapest way to buy those minutes. This covers the core functions of our agents: listening, speaking.

The notable exception here is ‘thinking’, i.e. the LLM costs and this is where the real complexity lies. Different models have dramatically different economics, and different customers or end-user profiles have dramatically different consumption patterns - we’ve proved this in the beta. Trying to blend these leads to us overcharging you or taking on significant risk, with fairness being an edge case. So instead, just as you can choose your model, we’re going to be passing on LLM costs at cost. That includes the substantial benefit from LLM-caching, where it is achievable. BYoLLM is on our list but you get the economic benefits of that from day one. The remaining benefit is really operational, e.g. charges at the same rate on your Anthropic bill rather than ours.

There is another critical function which is a huge Simwood USP and that’s ‘remembering’ - storing your knowledgebases for agents or indexing past conversations for state. Other providers are severely limited here, both in terms of functionality and price. ElevenLabs for example will (crudely in our opinion) apply a limit across the entire account - pretty useless - from 2MB up to 1GB on published plans. To access 1GB you’d be spending $990 a month and $0.08 per minute. Twilio is more generous - they’ll charge you $0.018 per GB per hour for as much as you like - so $1,296 for 100GB, before you retrieve any of it at $0.005 per time! We’re storing and indexing on-net, and are not grotesquely greedy, so we’re allowing as much storage as you need with charging storage at a sensible rate, i.e. £0.0018/GB/hour. Furthermore, unlike others, we’ll only charge based on the original document, not the multiple of that after vectorisation!

The final moving part is concurrency and this one is hard. Nobody in the market is overt about their limits and true constraints; it is all fuzzy language and obfuscation. Concurrency risks being incredibly expensive, particularly as we progress to full on-net processing, so we need to find a balance. Accordingly, every tier comes with a small number of included concurrent channels and additional channels can be purchased on top. These are account level limits which you will be able to divide and contend by trunks as you do with other limits on Simwood accounts. They will all be best-effort and have a hard limit, in the interests of fairness. Until POA territory these will be the same rate regardless of subscription level.

So, that’s a subscription which includes minutes, overage at a rate which reflects the subscription, included levels of concurrency increasing with subscription, and the ability to scale that as you need. Oh, and LLM usage at cost - you pick the model and entirely control your own cost profile.

Additional concurrent channels £5 a month each, same price on every plan. LLM tokens passed through at provider cost on every plan.

While LLM costs will be passed through at cost it is helpful to record what those costs are (at the time of writing) for anyone wanting to benchmark; the tables are at the end of this post. These are in USD so will vary for billing in GBP. For reference, if you’ve been using our services in beta, you will have used Gemini 2.5 Flash, GPT 4.1 or 5 Mini for Conversational AI and Haiku and Sonnet for Conversation Intelligence. Frontier models are available but absolutely not required. E&OE of course, and these will change, not least due to promotions, new models etc. One worth knowing today: Google’s current Gemini 3.x Flash pricing is promotional and doubles on 1 January 2027, and because we pass cost through, that increase reaches you.

Last but by no means least, we need to talk about data residency. Our ultimate aim is for all of this processing to occur on-net and we’re quietly moving towards that. We don’t intend to narrate every element as in all likelihood the transition will require hybrid working, but where this stuff is processed matters to customers and we know some are using it as a selling point. In every case possible, we are electing for UK processing first, EU if the UK is not available. Google don’t allow nomination but their services are anycast so logically we will naturally be using UK and EU instances first. When it comes to T&Cs however, we think it is more important that services work, so we will permit a fall-back to other instances (e.g. the USA) in the event of the preferred instance not being available.

We hope you’ll share our excitement at this and, naturally, look forward to feedback in our Community Slack or via account managers. When comparing, please do look at others like Twilio, ElevenLabs, Vapi, Retell etc. In every case they either can’t provide, or charge separately for, the connectivity, and the basic costs are substantially higher than ours - you guys should be able to compete with any of them on this pricing before we get into bespoke pricing for legitimate scale.

Finally, as regards timing, the billing is the remaining constraint here. As soon as that is done, these products will come out of beta, and we estimate that’ll be this quarter as billing is a Q3 objective and Charles doesn’t miss! Pete on the other hand…

**

LLM pass-through rates

Tokens are passed through at provider cost, so these are costs, not prices. Per 1,000,000 tokens, from each provider’s own pricing page, retrieved 11 September 2026. FX GBP 0.740295 to USD 1.00, ECB reference rates, 11 September 2026, one rate in every row.

Fast, the live turn loop

Standard, a capable agent turn and post-call analysis

Frontier

A note on caching. Google charge no premium to write a cache, so their write figure is simply the standard input rate, and the storage charge they publish applies to caches you choose to store rather than the automatic caching in use here, so there is nothing extra to pass on. Where a model is marked not available, Google publish no cached rate for it at all, so there is no caching benefit to pass through at any price. The minimum matters as much as the rate: below it a request is simply processed uncached, with no error and no warning.

Some of you have known me for a long time. There is a post from 2018 where I answered much the same questions much like these as the founder of a company called Sipcentric. It's still online, which is either brave or careless! Reading it back was a useful exercise. Most of what I said then, I would still say now. The scale is what changed.

The long way round

My route into telecoms was not direct. Before any of this I had launched an online trading company and an IT consultancy, built a small property portfolio, and opened a couple of bars. I have always been more interested in businesses and people than in any single technology, but IT and telecoms was where my expertise sat, and internet-based communications was where the interesting problems were.

Sipcentric was formed in late 2010, and we rebranded to Nimvelo in 2015. Over nine years we built a hosted platform, and the team to run it, without any significant outside investment. That constraint shaped everything. When there is no funding round to hide behind, you architect carefully, because you are the one who will be supporting it at 2am.

I also said in 2018 that we were working on something new, that I could not say much about it, and that its purpose was to help teams have more productive conversations. Hold that thought.

October 2019

Simwood acquired Sipcentric in October 2019. Simon was characteristically direct about why: he reckoned we had built the best-architected hosted platform in the market and that almost nobody had noticed.

Shortly afterwards I stepped up to Group CTO, a role Simwood had not had before. Going from running your own company to running technology inside a larger one is a genuine change, and it takes some adjusting to. What made it work was that this was not the usual sort of acquisition. Several of the Sipcentric team took senior positions in the combined business rather than being absorbed and quietly forgotten. David Maitland, who wrote in this series back in March, was Sipcentric's second employee at nineteen and is still here running infrastructure and systems. That says more about how the two companies fitted together than any press release did.

What I gained was scale, and colleagues who had been operating networks at serious scale since 1996. What Simwood gained, I hope, was a platform team used to shipping.

What the job actually is

I am responsible for the ongoing reliability and interoperability of every network across the group, and for the products we build on top of them. In practice the work splits three ways: keeping what exists running properly, building what comes next, and making sure the people doing both have what they need to get on with it.

The middle one gets the blog posts. The first one is where the credibility comes from.

A lot of that foundation is decidedly unglamorous. Database clusters that need to fail over cleanly rather than theoretically. Identity and single sign-on, migrated without locking anyone out. An intermittent lock-up that only ever appears in production, under real traffic, at the least convenient hour. None of it makes a good headline. All of it is the reason customers can put us in the call path and not think about us again.

Some of it has been running longer than the products it now supports. Kamailio has been in Simwood's stack since 2005, and it is still there, doing the same job in a completely different architecture. I told that story on stage at Kamailio World in May: twenty years of change with one constant in the middle of it. Contributing back to the open-source projects we depend on matters to me. We would not have a business without them.

Back to that thought

In 2018 I was talking about helping small teams have better conversations, from inside a PBX. Today the AI engineering team and I are building Conversation Intelligence, which does something similar from an entirely different place: the carrier layer.

Media is tapped passively at the network layer and mirrored into an AI pipeline. Nothing touches the call, nothing adds latency, and nothing routes through a third party, because we are the carrier and it is our network. Operators watch the transcript as it unfolds and emit typed signals — a fraud score, a detected intent, a compliance flag. Those signals fire webhooks into your platform mid-call, or whisper guidance to the handler while they are still talking. Every conversation becomes a vCon, an open IETF-standard record, held in your storage rather than ours. We genuinely do not want your data.

The pitch is simple: act while the call is still happening, instead of finding out next quarter. Fraud caught before the transfer is authorised. A missed disclosure prompted before the call ends. If you would rather read endpoints than prose, it is all in the developer docs.

Different scale, same idea I had eight years ago. It only works now because we own the whole stack, which is precisely what the 2019 deal put together.

How I try to work

My answer in 2018 was that I liked being one of the team and would never expect anyone to do something I would not do myself. That has not changed, and it is easier to live by here than it was, because Simwood already worked that way.

What I would add now is a preference for candour. I want to know the difference between what we have shipped and what we intend to ship, in plain terms, early. Optimism about a roadmap is fine. Confusing it with the current state of production is not, and it is the fastest way I know to lose a customer's trust. The same goes upwards: if I am wrong, I would rather be told in the room than discover it in an incident review.

The flip side is ownership. Nobody here waits to be told to fix something they can see is broken, and that is recognised rather than merely expected. You will find the same theme running through most of the posts in this series.

Away from the network

Home is a few rural acres, which supplies a reliable stream of problems that cannot be solved with more compute. There is a garage block mid-conversion and a planning process attached to it, so I have developed a new appreciation for systems where the specification is deliberately ambiguous. The rest of my time belongs to my family, who are unmoved by SIP.

Still here

Eight years on from that first post, and seven from the acquisition, I am still doing a version of the same job: building communications infrastructure that other people can rely on without having to understand it. The difference is the size of the network underneath it, and the calibre of the people I get to argue with about how it should work.

That is why I am still here. It is a rare combination, and it has not gone stale yet.

I want to start by saying well done, and I mean it. Gamma has been built into a business turning over £645.8m last year, growing at 11%, with 89% of its revenue recurring and more than 1,500 UK channel partners selling it. A £1bn cash exit is a result, and the people who did the work earned it. There are serious, capable people there, and nothing that follows is about them. It is about the model, and it is written for the partners - Gamma's first among them, but far from only them - who woke up on 1 September to find the ground under their supplier had moved, and who are now, quite reasonably, thinking about their own futures.

We have written about this before

Last July I wrote about buy-and-build: acquisitions funded with cheap debt at eight to twelve times EBITDA, and interest cover collapsing. Survivability, I argued, deserves a premium.

In April I wrote about enshitification, the point at which a business stops creating value for its customers and starts extracting it from them. Extraction is not a moral failure. It is the equilibrium of a commoditised market with capital to service, and public markets had already marked carriers down after overpriced acquisitions.

And in March I wrote that scale does not deliver margin in this industry, that seven to nine in ten deals fail, and that integration destroys the efficiency it was meant to create. Two years before that I had already called the pattern "debt funded roll-ups".

Three threads. On 1 September they tied themselves together in one announcement.

The premium in context

The headline was a 53% premium to the undisturbed closing price of 732p on 7 April. But a premium is measured from wherever the price happens to be standing, and the question is how it got there.

Gamma's shares peaked at around 2,350p in September 2021. The price the premium is measured from was down roughly two-thirds from that peak. The offer, at 1,120p, is under half of what the market once thought the business was worth.

So the market had made its decision about this business years before Epiris did. The board's stated reasons for recommending the offer include cash certainty and the "removal of execution risks from standalone strategy", and other proposals were on the table. A board with a good, growing, cash-generative business chose a certain pound today over its own plan for tomorrow. It tells you what the board believed the standalone plan was worth, and what the market had been telling them for several years.

Where the money comes back from

A fund does not buy a business to own it. It buys a business to sell it, in a few years, for more than it paid, usually having borrowed part of the price to do so. The interim facilities for this deal come from Ares. How much was not disclosed, but the shape is the shape. A price has been paid, and it has to come back to the fund, with a return on top, from somewhere.

There are only three somewheres. Price: charge the existing customers more. Cost: spend less on serving them. Growth: win more of them, or sell them more.

If you are a channel partner, you are on the receiving end of two of the three. Price is your margin. Cost is the service your customers experience and the support you rely on. Growth is the one everybody hopes for, and it is the hardest, which is why the other two get pulled first.

Which brings me to the AI. The announcement promises "increased investment in product innovation and AI adoption", and Gamma's own results in March leaned on AI as the growth lever: an AI concierge that answers calls, early voice-agent revenue in Germany growing fast off a small base, and management, to their credit, warning against extrapolating from it. Good. It is the right thing to be building and I would say so whoever was building it. But under a fund that investment sits on the same P&L as the debt service, and that P&L has to produce a return inside the fund's timetable. AI spent on growth is a bet that pays back, if it pays back, in years. AI spent on taking cost out of the business pays back with something close to certainty, on the cost line, which is the line the fund most needs. Which kind does the arithmetic choose?

As I argued in July, that is the distinction that decides who survives this. The companies AI finishes are the ones that use it for replacement and cost-out. The ones it makes are the ones that use it for amplification, so the same people do a great deal more, and amplification compounds: by the time the second car pulls away from the line, the first is not merely ahead, it is over the horizon. Compounding cannot be bought late, and it cannot be bought at all from a standing start, because it has already started elsewhere. The alternative I described in July is to spend telephone numbers buying customers instead of building, which is the model this whole post is about, one level up.

The announcement says there will be "no material headcount reductions" in the first twelve months, and no material changes for channel partners. I believe that is sincerely meant. I also note that twelve-month assurances have twelve-month horizons. Completion is expected in the first half of 2027. Count forward from there.

None of this needs bad faith. Nobody here is a villain. It is what the structure does: debt has to be serviced, a return has to be earned, an exit has to be prepared, and the people inside the structure, however decent, execute what it requires. That is what I meant in April. Extraction is an equilibrium, not a choice.

The third test

In July I gave you two tests for a supplier. Capability: have they already built what you need? Alignment: does their growth depend on winning the customers you are chasing? I want to add a third, because 1 September made it unavoidable.

Who owns your supplier, what do they need from it, and by when?

A business is run for its owners. That is not cynicism, it is more or less what company law says. When a supplier is owned by a fund, it is run for the fund's return, on the fund's timetable, and every partner of that supplier is a line in the model that produces it.

A side-point. Gamma's people, like those at most listed companies, have had share schemes, and I hope a good many of them see something from this. But the directors' own holdings committed to the deal come to about 0.13% of the company, and the equity is overwhelmingly institutional. The result belongs to the people who built it. The cheque, overwhelmingly, goes elsewhere.

Simwood is owned by the people who work in it. There is no external fund. We are largely debt free, and we published our interest cover last year rather than ask you to take my word for it. Apply the third test to us and the answer is short: the people who need a return from this business are the people answering your tickets, and what they need is for you to still be here in ten years. That is a different set of incentives from a fund with a fund life.

Three things to do this week

First, read your change-of-control and termination clauses now, not when the deal completes. Know what notice you can give, what notice you can be given, and what happens to your numbers and your customers' data if you leave.

Second, send the You buy from who questions to your supplier in writing: whose network your traffic ends up on, who else is in the path, and what they are building next year that you are scoping to build yourself.

Third, add the new one to the same email. Who owns you, what do they need from the business, and by when do they need it? A supplier who cannot answer that plainly has answered it.

The ground moved

If you built your business on Gamma, or on any supplier that has since been bought, you did nothing wrong. You chose a sound business on the information you had. The ground moved under you. It moved because the public market decided what the business was worth, a board decided cash was better than the plan, and a fund decided it could make the numbers work. None of those decisions involved you. Every one of them lands on you.

So the question is not whether you were foolish then. You were not. The question is whether the next contract you sign is with a business whose owners need you to win, or with a business whose owners need something from you. One of those is a partnership. The other is a line in somebody else's model.

I know which one we are. Who owns your supplier?

Ask anyone who knows me, and they'll confirm: I box, I play tennis twice a week, I've recently picked up padel, and I have a deep love for maths, numbers, and anything technical. I'm also a firstborn daughter, which basically means I'm professionally trained to take care of everyone around me, whether they ask for it or not. It's a lifestyle.

So when the opportunity came to join Simwood, a UK company, in a field I had zero experience in, telecommunications, I did what I always do with something new and slightly intimidating: I said yes. I joined as an Account Manager, and honestly, the part of the role I enjoy most is the communication side of it, getting to speak to different people, understanding what they need, and building those relationships. So if you're reading this and we haven't had a chance to speak yet, send me an email; or book a call with me. I'd love to catch up!

Six months in, here's where I am.

The team is welcoming and helpful, and has a good sense of humour, which makes a difference when you're navigating a new field and asking a lot of questions. There's a good balance between getting proper support when you need it and being pushed to figure things out yourself, which is honestly the best way to actually learn.

The telecommunications world turned out to be exactly the kind of subject I enjoy, full of logic, structure, and enough complexity to keep things interesting. A bit like learning tennis. Confusing at first, satisfying once things start clicking.

This is also my first time working for a company outside of Georgia, with a UK-based team, remotely with them from a great office in Tbilisi. Being remote doesn't save my colleagues from me, though; Roam (Virtual office platform) makes sure of that. Distance is simply not a barrier when you have questions, and I always have questions. One thing worth mentioning is that, even though Simwood is based in the UK, we also have a small team here in Tbilisi, Georgia, which made the transition much smoother. Getting to work alongside some great Georgian colleagues while also being connected to the Bristol team really does give you the best of both worlds.

That said, I'm very much hoping to make it to the Bristol office before the year is out. I'd love to experience the UK work culture in person and finally put faces to all the names I've been messaging.

One of the things that genuinely surprised me during these three months was discovering what Simwood has actually built in this space and how much it can change the way a business communicates. Coming from outside the industry, I didn't expect to find something that felt this practical and immediately useful. Simwood has developed AI agent integration directly into tools people already use every day, such as WhatsApp, Microsoft Teams, and similar platforms. These agents can take calls, transfer them, generate summaries, handle first responses, and basically take care of all the front-line tasks that quietly eat up hours of someone's working day. I say this as someone who came in with no background in this at all. Once I saw how straightforward Simwood has made the integration process and what it actually delivers for a business in terms of time and efficiency, it was hard not to appreciate what's been built here. If you haven't had a chance to explore our new offerings to share them with your users and make their life easier, it's genuinely worth a conversation.

If this is new territory for you as a client, make sure to check out our other blogs where we go into more detail on all of this, or just reach out to us directly, and we'll make sure to walk you through everything you need to know.

These three months have been a solid introduction to a field I knew nothing about, and I'm looking forward to seeing where it goes from here.

Next time I write one of these, hopefully I'll have a Bristol visit to report back on too.

Until then, Mariam

PS: If you haven’t seen our newest podcast episode, you really should here.

That has in turn done a few things for our customer base. It has vindicated our long-held position on scrotes, although the burden on our sales team as a filter has ballooned with so many scrotes looking for a home for their traffic. Moreover, our non-US customers have had to themselves become compliant with new requirements, agencies, and filings. Some have taken this in their stride, others have made a massive ordeal out of it and resent us for changes the FCC has imposed, but we generally support. However you skin it, our costs in the US market have risen exponentially.

There are still operators pretending to offer service in the US and bemoaning us for being compliant. One "peer" advising our customers on US compliance when they have zero presence and zero licence there took the biscuit. If you resent compliance to protect consumers, please feel free to use them and see how long service lasts.

This is a long way of saying that our pricing is woefully out of date, and effective September 1st (for billing October 1st), our rentals on US numbers will increase as below. Naturally, Managed Interconnect customers with a bespoke committed rate will continue to benefit from that rate, at least until contract renegotiation.

Further, while e911 was never relevant to us as a wholesale operator, our customer profile has changed, and it now is. Effective immediately, you can upload address data through our portal and API. It is in API v3 rather than the newer architecture for easy compatibility with UK 999 integrations and will be deprecated/migrated alongside in due course.

Commercials for e911/911 vary in the US to what our UK customers will be used to. We are not applying any kind of database update fee - they're free of charge - and calls to 911 are also free of charge. However, numbers (a.k.a. DIDs/DDIs) which are e911 enabled (i.e. have had data uploaded against them) are charged a monthly uplift to the rental to cover the elements which are not charged. This is not our model; we've followed local convention.

Our monthly per-number pricing for the US is now:

All prices are per telephone number, per month, ex VAT at the time of this blog being published. Current rates are always on our rates page.

Our existing statutory investigation fee is unaffected by this change and will apply wherever a call is routed to 911 and no e911 data exists. In this circumstance, the call cannot be routed to the appropriate Public Safety Answering Point (PSAP), the functional equivalent of UK EHA, because there is no address. This not only risks lives but incurs additional cost and process all round. We will take an extremely dim view of any customers who seek to minimise rental costs by chancing the investigation fee and risking lives in the process.

We will route calls to 911 where the call originates from a US number. Calls for numbers allocated through Simwood should be sent to Simwood, and the same with other providers - same preference as in the UK. However, as in the UK, we will connect any call and deal with the compliance afterwards.

911 should not be tested, but to make this possible, you can configure 933 to route identically and make test calls to 933. The call will answer and read back the address information configured against the number. Furthermore, API updates do not require the overnight batch processing Calypso requires despite the user base being roughly 5x the size. Updates via our API should be reflected relatively instantly. This is useful for testing, but you should not build production logic requiring it!

Hopefully, this is a welcome update that'll help all our customers with US estates improve their offer and compliance.