This ended up a long but worthwhile discussion. The Simwood potato is relevant across the telecoms industry, from legacy telcos needing to leapfrog technical debt, to Simwood competitors missing features and services, to enterprises needing to add features overnight, and all the startup voice AI companies needing to plug their telephony gaps. It’s a surprisingly full list of potential customers.

Simon kicked off with a quick intro to Simwood’s, 30 year history. Which all leads up to the Potato. From the beginning in 1996 when Simon founded eSMS / SIMWOOD, the first global SMS gateway between the Internet and mobile phones, a proto-SMS-aggregator. You can read more about Simon’s eSMS adventures in The Definitive Truth in A2P SMS.

Simon’s gap year in 2021-2022, was in part to think about what’s next, that resulted in the Potato. Simwood realized their path was not just being the most efficient telecom software company, but a reinvention of what that meant given the trends in openness and decentralization.

Simon referenced a claim by McKinsey that $8T has been invested by the telecoms industry over the past 10 years. I’ll avoid my rant on McKinsey. But that $8T investment has delivered declining revenues and shareholder returns, ignoring those that play the non-GAAP game. By contrast, Simwood has delivered 3.8x revenue, EBITDA 6.5x, and a CAGR of 21%.

I consider Simwood to be the techco archetype carriers aspire to become. Simwood has always built their own network using open source. The challenge telcos have is their culture and the legacy ecosystem around them will always stop them achieving the techco vision. But the Potato changes that, as we’ll discover.

The market reality facing all carriers is minutes and messages are moving to the internet. And Simwood can not avoid that. WhatsApp has become pervasive across Asia (and LATAM) business workflows, even within a hotel you can use WhatsApp to get a pillow for your room. Carriers have not met that utility. Rather creating telco standards, consortia, and expensive upfront equipment without customer involvement. Rather than simple web-centric integrations on existing web pages using WhatsApp.

In addition there is massive over capacity in broadband, with 140EB offered to 11EB demand. I’ve said 4G was good enough. Tim Panton also pointed out most 5G is really 4G.

The BUT is, voice remains the killer app. Our children still use voice, just listen to them play online games: the shouting, screaming, hurling of abuse at each other. Global voice continues to grow at a CAGR of 12-13%. UCaaS (Unified Communications as a Service) is still growing in mature markets like the US and UK, and in many regions its still emerging. Simwood provides UCaaS, but 80% of their revenue is carrier services, the nuts and bolts under UCaaS.

Simon pokes at Johnny on CPaaS (Communications Platform as a Service) and non-GAAP (Generally Accepted Accounting Principles), so he goes off on the one well-known SaaS/CPaaS burying marketing costs in their COGS (Cost Of Goods Sold). Making the gross margin negative, but we’ll leave that for another time.

Voice AI has a CAGR 35%, growing to $48B by 2034, and will be part of the Simwood Potato. AI is impacting operations and development. Simon made an interesting point on voice AI, many startups are doing Voice AI in a Gen Z way, by themselves and in the browser. Which opens up opportunities for the Potato, in crossing the chasm between telephony and the Voice AI community.

There are reasons to be upbeat:

• Carriers need to vastly cut costs; their network, its features, and processes are locked in the past;
• Carriers need to get relevant, WhatsApp is running circles around them. GSMA standards are not going to solve that, no matter what telecom analysts claim;
• UCaaS and CPaaS need supporting services; and
• All the AI startups need telephony, voice is the language of AI.

Back to the food analogies, the legacy telco architectures are like sausages and spaghetti. A new service is a new sausage and more interconnecting spaghetti. This is tied to the box and wires thinking in telecoms. Its 2025, a software centric architecture should be possible by now. And this is where the Simwood Potato come in.

The Simwood potato is a globally distributed core network, present on every continent, with real-time consistency everywhere. Because its software, using open source plus their secret sauce, and relevant internet technologies like DNS (Domain Name System, the internet’s phone book), anycast, etc. Not multiple SBCs (Session Border Controllers) with customers tied to specific SBCs, and hence the problem of configuration between them.

Wrapping that modern core are applications (think aluminium foil) such as: transcoding, recording, AI agents, translation, transcription, encryption, tag analysis, nuisance call filtering (think of the robocalling nuisance). Remember this is working at carrier scale globally. On nuisance calls they are blocking hundreds of thousands of calls every day. Because it’s the right thing to do, rather than earn money off robocalling. Currently it’s about 4% of the traffic, but they could turn up the dial to 40% of calls.

And there’s more… Skewers, representing a protocol or a geography. For example, PSTN, Zoom, Teams, SIP, WebRTC, SMS, across Europe, North America, China, etc. And any layer can be turned on, so a SIP call could be routed to Teams with call recording and transcription on. It’s just software.

The thing that I did not realize is its BYoC (Bring Your own Carrier). Simwood will let you bring your own carrier, but they are a carrier? A competitor could use the Simwood Potato to provide services for their customers, not Simwood’s. But what they’ve proven is they’re winning deals, number porting in the UK is a pain. Simwood has encryption, a competitor did not, so they were able overnight to add encryption for that enterprise customer.

Simwood has always been an innovative leader in the market, many of their disruptive startup customers get bought by competitors. But they have not been squeezed out of the market. They tend to remain a customer, while also being a competitor on some aspects of the business. So the Potato’s BYoC had an existence proof.

Another use case is protocol conversion. Teams is sold as a SIP trunk alternative. But what that means is Microsoft has acquired a customer, where the value is being built at the Microsoft layer. The SIP trunk will be commoditized in time, like you do not pay for SMS and voice on the mobile plan. It’s just a seat, so sell both and let the customer decide, as there’s lots of potential for value add services on the SIP trunk.

The cost cutting and feature enhancing is critical for incumbent carriers. The legacy stack traps carriers, so the potato enables carriers to break free and offer a modern feature set. It is radical, but some carriers are working on such a transition. This is aligned with Simwood’s strategy, they do not want to be a licensed operator beyond the UK and US. So this enables most carriers to use the Potato.

New service agility, using an example of a mobile operator in Ecuador needs to add call recording. This could be done per phone number or trunk through the Potato. The one that I think is going to be big, depending on the local regulator, is the nuisance call filtering. Just imagine if the regulator set targets, and fines for dumping robocalling on customers!

Because the Potato is globally distributed they can help carriers enter new markets. Say a fiber operator needs to add telephony, with the Potato they can support the operator to deliver a compliant national voice service through their ecosystem.

The general discussion moves onto how BYoC as a carrier is unique. But Simwood had the experience to realize it was possible within their business. Dave Horton backed up Simon on supporting the AI guys need with telephony. Interestingly, Dave is seeing voice AI companies approach him where they do not yet have a carrier partner. Fingers crossed they can make some joint business happen there.

Johnny focused on where Simon is taking the business, and options for Simwood to access cash to scale the business. Dave mentioned he needs to position against LiveKit, I think there’s another TADSummit session there. And we wrapped on the importance of voice.

I think Simon’s gap year has enabled a revolutionary unique model, the nuisance call filtering should be on every regulators to do list. That could reverse the decline in telephony.