A founder sat across from me last month, still raw from a partner meeting that had gone sideways. The numbers were good. A clean thesis, real retention, a product people used every day. Forty minutes in, the lead partner leaned back and asked one question.
“If you were starting this today, with the models that shipped this quarter, how many people would you need and how much would it cost?”
The founder answered honestly. Three engineers instead of fourteen. Maybe a fifth of the burn. The room went quiet, and the meeting was effectively over. No term sheet came.
He thought he had been passed on for his metrics. He had not. He had been passed on because he answered, out loud, the question every investor is now asking silently about every company they see. There is a cheaper version of you. The partner had already pictured it before he walked in.
The Lens Behind Every Question
For most of the last decade, investors evaluated a company on a fixed set of vectors. The strength of the thesis. The physics of retention. The depth of the moat. Unit economics, technical debt, growth loops, burn. You could be strong on some and weak on others, and the conversation was about which ones mattered most for your stage.
That set of vectors has not changed. What has changed is that there is now a second question stacked on top of each one. Not just “is your moat real,” but “is your moat real once a competitor can rebuild your core feature in a weekend.” Not just “are your unit economics healthy,” but “are they healthy against a cost base that just dropped by half for anyone starting fresh.”
I call it the AI lens, because that is what it functions as. Investors hold every vector of your business up to the light and ask what it looks like in a world where the cost of building software has collapsed. They do this whether or not the word AI ever comes up in the meeting. It is no longer a sector. It is the lens through which every other thing about you is now read.
The founder who loses does not lose because he is behind on AI. He loses because he has never run the lens on himself, so the first time anyone does it is in the room, in real time, in front of the person deciding whether to fund him.
What the Lens Actually Tests
The lens is not a vibe. It resolves into specific, answerable questions on each vector, and a sharp investor moves through them quickly.
On your thesis: does the core insight still hold if the next model release closes the gap you are building into. A thesis built on a capability that is six months from being free is a window, not a market.
On your moat: what is the half-life of your defensibility. If your advantage is a clever workflow or a thin layer of intelligence on top of someone else’s data, the lens treats it as already eroding. What survives is proprietary data, real distribution, and the cost a customer pays to leave you.
On your unit economics: the lens does not compare your payback to your own history. It compares it to the payback of a company built today, on today’s cost base, with a team a third of your size. If your numbers only look good against the old way of building, they do not look good.
On your burn: the question is no longer how efficient you are in absolute terms. It is how efficient you are relative to the version of your company that does not carry your headcount, your legacy stack, or your habits. Burn that was reasonable two years ago can read as bloat now, and the founder rarely feels the ground move underneath the number.
You do not get asked all of these out loud. You get asked one or two, and the rest are scored in silence.
The vectors investors judge you on did not change. What changed is that each one now carries a second question: how does it hold up against a company built from scratch today, for a fraction of the cost. That second question is the one you are losing on.
The Company You Are Being Compared To
Here is the part founders find hardest to sit with. The competitor doing the damage in your raise often does not exist yet.
The partner is not comparing you to the three funded players you track. He is comparing you to a hypothetical team of three that could start tomorrow, reach your core feature set in a quarter, and operate at a fraction of your cost because they never accumulated what you accumulated. That ghost company has no technical debt, no oversized team, no expensive contracts signed in a different era. On a spreadsheet, it beats you on every efficiency metric that matters, because it carries none of your weight.
You cannot out-argue a ghost. You cannot show better slides than a company that has no slides. The only thing that beats it is the set of things the ghost cannot get for free: the customers who already trust you, the data you have collected that no model was trained on, the workflows your users have built their day around, the distribution you have already paid for. Everything the cheaper version of you would have to spend years and real money to acquire.
That is the entire game now. The lens strips away anything a fast follower can replicate cheaply and asks what is left. Founders who raise well in this market are the ones who can point at what is left without flinching.
Why “We Use AI Too” Fails
The instinct, when a founder finally feels the lens land, is to answer it by claiming the technology. We use the new models. We have an AI roadmap. We shipped a copilot.
This is the weakest possible response, and investors hear it ten times a day. Using the same tools as everyone else is not a defense. It is table stakes that proves the opposite of what the founder intends. If your edge is that you adopted a technology available to every competitor and every ghost on equal terms, you have just confirmed you do not have an edge.
The lens does not reward who uses AI. It rewards who is hard to copy in a world where AI is everywhere. Those are different questions, and founders who conflate them spend the whole meeting answering the wrong one.
The strong answer is never about the technology you use. It is about the position you hold that the technology makes more valuable, not less. A proprietary dataset becomes more valuable when models are hungry for data. A deep workflow entrenchment becomes more valuable when switching is cheap everywhere else and expensive only with you. Distribution becomes more valuable when the cost of building product falls to zero and the cost of reaching customers does not.
Run the Lens on Yourself First
The founder I opened with did the work afterward, and it changed his next raise completely. We sat down and ran the lens on his own company before any investor could. We took each vector and asked the uncomfortable version of the question.
What in our thesis depends on a capability that is about to be commoditized. Where is our moat actually a moat, and where is it a feature a funded team could clone in a month. What do our unit economics look like against a company built this year rather than against our own past. Where is our burn carrying weight that a fresh start would never pick up.
It was not a pleasant afternoon. Three of the answers were bad. But by the time he walked into his next partner meeting, the ghost company was not a surprise waiting to ambush him. He raised it himself, early, and then spent the meeting on the things the ghost could never have: a six-figure dataset gathered over four years, two enterprise logos whose entire operations now ran through his product, a distribution channel he owned outright. He had already conceded the cheap ground and moved the conversation to the ground he actually held.
He got the term sheet. Not because his metrics had improved in six weeks. They had not. He got it because he stopped letting an investor run the lens on him and started running it on himself.
Whatever an investor will ask in the room, ask it first, in private, with the harshest possible framing. The founder who has already survived the lens walks in calm. The one who has not gets to experience it for the first time in front of the person holding the check.
What This Means for You
The cheaper version of you is not a threat you can eliminate. It is a permanent feature of the market now, and it sits in every meeting whether or not anyone names it. Pretending it does not exist does not protect you. It just means you find out it was in the room after the pass arrives.
The founders who are still raising well have made peace with this. They concede the ground that cannot be defended, quickly and without drama, and they build their entire case on the ground that can. They know which of their vectors survive the lens and which do not, because they ran it themselves before anyone else got the chance.
There is a cheaper version of you. The only question that matters is whether you have looked at it clearly, or whether a partner is about to introduce you to it for the first time. Look first. Then walk into the room already knowing what you are.