London Tech Week - What Should Series A Investors Ask?

London Tech Week - What Should Series A Investors Ask?

Having gone through the speaker list and the nine themes, the thing that jumps out, maybe unsurprisingly, is AI, AI, AI, and more AI! ElevenLabs, Perplexity, Lovable, OpenAI, Anthropic, Notion, Wayve, Isomorphic Labs on stage, plus themes explicitly built around Applied AI, Responsible AGI, agentic AI, AI infrastructure and sovereign compute. The deep tech and quantum tracks are there, but the centre of gravity is obvious. That shapes what a Series A investor should actually be probing — because most of what you'll meet on that floor will be AI-wrapped, and the easy questions won't separate the real businesses from the demos.

Here's what I'd be asking, and why it matters given that specific line-up.

Defensibility against the model layer. With the foundation labs literally on the same stage as the application companies, the sharpest question for any Series A app is: what stops OpenAI or Anthropic from shipping your feature as a default next quarter? Lovable is the live case study — huge reported ARR off vibe coding, but how much of that moat is product versus distribution timing? I'd want to hear a credible answer on data advantage, workflow lock-in, or proprietary distribution, not "we move fast."

Margins, not just revenue. A lot of these companies will quote ARR that looks like SaaS but actually carries the cost of goods that scales at the same pace as revenue, because of inference. I'd push hard on gross margin and its trajectory: where does it sit today, what's the path to 70%+, and - a crucial one for me - how exposed is the unit economics to a model-pricing change you don't control? Revenue that grows while margin quietly inverts is the trap of this era.

Agentic AI: in production or in the keynote? The Enterprise Growth theme leans on agentic AI and the word will be everywhere. The uncomfortable question is how many of those agents are running unsupervised in a paying customer's environment versus sitting in a sandboxed pilot. I'd ask for named production deployments, the failure rate, who is liable when the agent gets it wrong and what is done to prevent it.

Retention under the novelty curve. Early AI-product usage is flattered by curiosity. I'd want net revenue retention and cohort behaviour at 6 and 12 months — does usage compound after the wow wears off, or does it decay? This is where a lot of impressive 2026 top-line numbers will quietly fail.

Sovereignty as tailwind and trap. George Osborne is there heading "OpenAI for Countries," the UK AI Minister and Liz Kendall are on the bill, and sovereign compute is a theme. Government and public-sector demand is real money, but it's also pilot purgatory with brutal sales cycles. For anyone leaning into that wedge, I'd separate signed, recurring revenue from MOUs and "strategic engagements," and ask how long the cash runway assumes that cycle takes.

Capital intensity and milestone-to-milestone honesty. For the deep tech, infra and quantum companies (Planet, Isomorphic, the quantum track), Series A means something different — the question isn't product-market fit yet, it's how much capital and time to the next de-risking milestone, and whether the syndicate behind them can actually fund that path. I'd be wary of deep tech raising a Series A on a SaaS-shaped narrative.

The survivorship bias baked into "Unicorns' Lessons in Scaling." That whole track is winners explaining why they won, mostly with hindsight. The more useful private question is about the 2021–2022 vintage now down-rounding or quietly dead: what changed, and is the company in front of me repeating those mistakes with an AI coat of paint?

Burn and the talent market. AI talent is being bid up viciously right now. I'd want to see the burn multiple and a clear view on whether the team can be retained without a comp structure that eats the next round.

If I had to compress it to one filter for the week: most companies will pass the "is this exciting?" test. Very few will pass "what do you own that the people one stage over can't take from you, and does the money work when the hype tax comes off?" That second question is the whole job at Series A in this market.

To help you further, here is a quick question sheet. Feel free to use it (download the image and print).

Image below looks cropped, but if you right click -> save as, you'll get it all

If you want to discuss this in more detail and understand what it means for your business right now, click here to contact us.