TRAMLINES

TRAMLINES are pleased to announce our partnership with Teal+ to co-build the AI-native layer for AdTech as a category. As humans, we have been trying to reach the right audience for millennia. The first-ever written advertisement was found in the ruins of Thebes in Egypt, on a papyrus dated around 3000 BC.  Five thousand years later, the problem is still the same.    AdTech, or Advertising Technology, is one of the most complex ecosystems of the digital economy today, expected to be valued a trillion-dollar industry in 2026. Most of that value sits within the platforms who own the data. And yet, it is inaccessible to most publishers, who, despite serving huge audiences, simply do not have the data layer needed to prove what that inventory is worth. Their impressions go to auction under-described, under-bid, and undervalued.  Understanding market gaps like this one comes with years of experience in the field, and it is exactly what Ade Clayton (Teal CEO), Matthew Whaley (Founder & CSO), and Oliver Burns (CTO) represent. Matt spent over two decades in the industry, previously founding Triple13, a programmatic and Prebid-server specialist that was acquired by Freestar, a monetisation platform for digital publishers. Freestar’s Chief of Staff at the time was Ade Clayton, who has now become the CEO of Teal, supported by Olly, who has spent his career building and shipping complex software.   Between them, they have been in the delivery seat, the client seat, and the executive seat.  It is exactly the kind of domain expertise we look for at TRAMLINES. Building the data and AI layer underneath has the capacity to become the default way in which independent publishers will run advertising.   Through Tramlines, Teal is now supported by Glen Robinson, former Microsoft National Technology Officer and one of the first four AWS employees in Europe, as their AI Readiness Lead, Albert Azis-Clauson 🦚, CEO of Tramlines, on commercial execution, and Andrea Madaschi as CFO.  The advisory bench includes Craig Donaldson, founding CEO of Metro Bank, who took it through a £1.6bn IPO, and Andrew Winters, previously a Managing Partner at Deloitte Technology & Digital. To Ade, Matt, Olly, and the team at Teal - we are very excited to be building with you. 

In a world where anyone can build software, the only real defensibility is structured domain expertise.  Today, we are proud to back Hybrid AI Solutions. Steven Ellis spent 25 years in ERP, including nearly a decade leading ERP at a £1bn publicly listed medtech company. Tom Marshall grew a Microsoft Business Central practice from three consultants to twenty-three in four years.  Between them, they have been in the delivery seat, the client seat, and the executive seat across hundreds of implementations.  That is exactly the kind of domain expertise we look for. What they have built is a Business Central implementation service, completed in 3 to 4 months, with AI handling the documentation, data mapping, configuration, and testing administration that traditionally consume most of the budget and most of the project timeline. Senior consultants remain throughout, owning the governance, the change management, and every decision requiring human judgement.    The result:   BC implementations that cost half as much, take a third of the time, and cause far less disruption to the businesses going through them. Hybrid AI Solutions is supported by Glen Robinson, known for being the Microsoft National Technology Officer and one of the first four AWS employees in Europe, as the team’s AI Readiness Lead, Albert Azis-Clauson 🦚, CEO of Tramlines, as one of the Directors supporting the commercial execution, and Andrea Madaschi as the CFO. The advisory team includes Craig Donaldson, founding CEO of Metro Bank, who led it through a £1.6bn IPO and Andrew Winters, previously a Managing Partner at Deloitte Technology & Digital. Over 100,000 UK businesses are still running on legacy ERP. The reason they have not moved is cost, risk, and time. Hybrid AI Solutions removes all three. Congratulations to Steven, Tom, and the whole team.  We are very glad to be building this with you. #BusinessCentral #MicrosoftDynamics365 #HybridAI #DomainExpertise 

The TRAMLINES team is at Women's Health Founders Day on Thursday 9 July. A full-day event for founders, investors, and ecosystem partners across B2B and consumer health tech. Two sessions from us. Erica Young, our Director of Networks (formerly Atomico, Foresight, and Anthemis), is giving a morning presentation on network building, followed by a lunchtime practical workshop on how to map and assess your network. Eva Dobrzanska, our Investor Communications and Relations Lead, is delivering one of her signature playbooks: how to do investor outreach and fundraising strategy that actually lands meetings. Free to attend. Practical tools you can put to work the same week. The TRAMLINES methodology, applied. Event link in the comments.

In 1882, John D. Rockefeller set up the first modern family office, after realising he had become wealthier than any single bank could prudently advise on. The structure he built was deliberately conservative - investing in bonds, treasuries, land, and later on, some blue-chip equities. The point was preservation, rather than growing wealth. For nearly a century, that was the dominant model for family office investments, but what I've seen across my conversations with some of them is how it is changing. UBS published their 2025 Global Family Office Report: US family office allocations to alternatives (PE, VC, direct deals in startups) sit at 54%, and in Europe at 44%. Twenty years ago those figures were closer to 15%. The categories that used to be "speculative" are now the core of the portfolio, and whilst some VCs focus on what this means for deals, the bigger implication is on the LP side. For traditional institutional LP base what we seen across the few past years is that Endowments and pensions are squeezed by the DPI drought. Secondaries and adjustments are common, and they are necessary just to maintain existing commitments. Meanwhile, as family offices have stepped into the world of backing VC funds, they do not behave like institutions. They are faster, more thesis-driven, and that is working to their advantage - they can enter the funds that are filling up fast. #familyoffice #LP #vc #venturecapital

Only 17%. That's how many $750M+ VC funds have ever returned more than 2.5x to their LPs. Whereas when you look at the smaller counter-parts (sub-$350M funds), the percentage grows to nearly 25%. So why does the money keep flowing to the largest funds? In the words of a tale as old as time, “Nobody gets fired for buying IBM”…
 And in the venture space, what happens is that institutional allocators aren't always optimising for better returns, but they're likely optimising for their own career safety, too. Losing capital in a famous fund = rounding error, plus you still get to say you’re backing a Tier 1 fund Losing it in an emerging manager = performance review trigger. In psychology, this is called agency dilemma. It’s a conflict of interest when one party (the agent) is delegated to act on behalf of another (the principal), but their goals and motivations are not perfectly aligned. Understanding this dynamic matters whether you're raising a fund or backing one - ultimately, personal incentives can shape behaviour far more than best intentions would… #venturecapital #vc #lp #fundraising [ DATA SOURCE: Santé Ventures' November 2023 whitepaper: Why Venture Capital Does Not Scale? Based on PitchBook data covering ~1,384 U.S. VC funds (vintages 1979–2018) ]

Mega-funds are structurally dependent on mega-rounds. For a $5B fund, writing a $50M check is relatively small. They also need billion-dollar positions to return the fund. This means, …mega-funds end up crowded into the same three names, at the same inflated valuations, alongside every other large fund in the room. Only 30 VC funds control 75% of all LP dollars, with a16z at the very top. “Nobody gets fired for buying IBM", after all, but… Santé Ventures analysed 1,384 U.S. funds between 1990-2018. $750m+ funds return an average 9.7% IRR Funds <$350m average a 17.4% IRR Larger funds need to deploy capital at scale which, in today's market, means writing enormous checks into the only rounds large enough to absorb them. OpenAI raised $1B in 2019 + $2B in 2021 + $10B in 2023 + $6.6B in 2024 + $40B in 2025 + $122B in 2026 Plus a few more secondary rounds & extensions in between, totalling a whopping $180B+ raised in total. To put things in perspective, the Apollo Programme (putting humans on the moon, arguably the most complex engineering endeavour in human history) cost the equivalent of roughly $200B in today’s money, spread across 13 years. OpenAI’s March funding is nearing to be its equivalent. 🌓 The power law was always venture’s defining feature - a small number of outlier outcomes drive almost all the returns. And, as for the mega-funds, they don’t escape it, but rather, deepen the dynamic. #venturecapital #vc #lp #fundraising

Only 17% of >$750M funds delivered >2.5x return to investors, and yet, the top 30 funds secured 75% of all LP capital deployed in 2024. “Top”… by what metric? The average IRR for a large VC fund is 9.7%, compared to 17.4% for smaller funds. Genuinely seems like most LPs are backing the worst-performing tier in the market, and doing so with increasing conviction. And there are consequences. When the same 30 firms control 75% of the capital and all chase the same thesis, you lose the conditions that produce contrarian thinking. At that point, investing stops being about identifying the best deal and starts becoming a Keynesian beauty contest: you’re not picking the winner you believe in, you’re trying to guess what everyone else will pick, and then picking that. #LP #vcmetrics #venturecapital