Near the end of a first-customer-success workshop at Draper U, the speaker drew a ladder on the whiteboard.
At the bottom was a discovery call.
Above that, written commitment.
Then money.
Then product shaping.
At the top, evangelists.
It was a crude ladder, but it stayed with me. Mostly because it cut through one of the more comforting lies early-stage founders tell themselves: that visible interest is the same thing as real progress.
A customer will take the meeting.
They will ask for a demo.
They will say the problem is important.
They may even introduce you to someone else.
None of that is meaningless. It is just dangerously easy to over-read.
What I took from that session was not really the ladder itself. It was the shift underneath it. Stop asking whether the customer is interested. Ask whether they have moved. Ask what they are willing to put at risk. Ask what this step proves, and what it still does not.
Not every yes carries the same weight
The useful thing about the ladder is that it turns a fuzzy emotion into a series of commitments.
A discovery call is a small investment of time.
A written yes is a small investment in traceability.
A paid pilot is a budget decision.
Product shaping is organisational attention.
Public advocacy is reputational capital.
Seen that way, validation is not about enthusiasm. It is about what the customer is willing to carry.
That sounds obvious when written down. It is less obvious when you are tired, a bit underconfident, and badly in need of a sign that the company is not imaginary.
That is usually when founders start counting warmth as momentum.
An LOI is not useless. It is easy to misread.
LOIs are where this usually gets messy.
Most founders know, at least intellectually, that a paid pilot is stronger than a non-binding letter. The problem is not ignorance. The problem is that an LOI looks enough like progress to soothe you. It feels more substantial than “sounds great, let me know when it is live”.
So I would not dismiss them.
An LOI can be valuable. It can show that the problem is serious enough for someone to leave a paper trail. It can show that a champion exists. It can show that interest has moved from conversational to slightly more concrete.
But it still usually does not prove three things:
First, that budget will really be released.
Second, that procurement will go through.
Third, that the customer will stay on the hook when adoption becomes inconvenient.
So an LOI often proves direction, not commitment.
YC’s founder-sales material still distinguishes between free pilots, paid pilots and longer-term deals for exactly this reason: each stage proves something different. A free pilot may prove feasibility. A paid pilot is closer to willingness to buy. A longer contract begins to say something about trust and expansion.
Those are not interchangeable signals.
I became suspicious of verbal interest because I had already hit this wall myself
What made me more sensitive to this was not just that Draper U workshop.
It was also a stretch of hospitality BD I had done myself.
The pattern there was not that people disagreed with the problem. If anything, it was the opposite. The hotel operators willing to sit down and talk generally knew OTA commissions were high, guest data was hard to reclaim, and loyalty was difficult to build on their own. The problem was real. They knew it was real. In practice, though, recognising a problem and stepping forward to do something about it were still two different things.
That was where my own thinking changed. A lot of B2B conversations do not stall because demand is fake. They stall because somebody has to carry the cost of implementation, priority and risk. A prospect nodding along does not mean they will carve out time this quarter. Finding the idea reasonable does not mean they will push it internally. Even signing a document does not mean they will stay at the table once adoption becomes inconvenient.
I wrote about that episode more directly in “Everyone sees the problem. Why won’t hotel operators move?”. The main correction for me was not that the market was clueless. It was that a real problem does not automatically produce adoption.
Low friction is not always a virtue
One of the better recent pieces on this came from Sierra’s design partner strategy. They did not optimise for the easiest possible pilots. They asked design partners to pay, commit time and co-build. According to First Round’s write-up, those partners converted into customers at 100%.
The interesting point is not “always charge early”. That would be too neat, and wrong in plenty of categories.
The more useful lesson is that low-friction engagement can be hollow.
A customer who is happy to chat, happy to watch a demo and happy to tell you the idea is exciting may still be doing something much closer to market research than buying. If nothing is at stake for them, the relationship may be informative without being commercial.
That is the uncomfortable part. Once you admit it, you have to start asking less flattering questions.
Who is spending political capital on this?
Who is trying to get internal buy-in?
Who is willing to explain this purchase to finance or procurement?
Who is exposed if implementation gets messy?
That is where movement starts to separate from mood.
Every step should be tied to a proof
This is the filter I now find more useful than a generic “strong interest” label.
Not every company should demand money on day one.
Not every enterprise sale can move quickly.
Sometimes the immediate proof you are after is technical feasibility, not willingness to pay. Sometimes what you have found is a champion, not yet a buyer.
That is fine. The important thing is not to count all those steps as the same type of progress.
I prefer to keep two columns in my head:
- what this step proves
- what this step does not yet prove
A discovery call proves that someone will spend time. It does not prove the pain is acute.
A written response proves they are willing to leave a trace. It does not prove budget exists.
A free pilot proves they are willing to test. It does not prove organisational adoption.
A paid pilot proves some budget willingness. It does not yet prove expansion.
A public case study is closer to proof that your product has become part of how they work.
That framing is not glamorous. It is just less self-deceptive.
Early on, referenceability matters more than founders like to admit
The AWS story from that workshop stayed with me for another reason. Internally, the question after a big win was not only “did we close it?” It was also “is it referenceable?” Can we talk about this customer? Can we turn it into a case study? Will they get on stage and say this worked?
I like that standard because it pushes past revenue into something harder. Not whether somebody bought once, but whether they are willing to lend you their credibility.
That corrected one of my own instincts. I used to overvalue the silent paying customer. Money mattered most, end of story. Later I realised that, especially early on, the most valuable customers are often the ones who lower the next customer’s cost of belief.
A startup usually does not just need a first contract.
It needs evidence that the first contract was not a fluke.
I trust movement more than warmth now
So if I had to reorder early validation, I would not start with whether the market is huge, or whether lots of people “like the idea”.
I would start here:
- Is there a concrete customer willing to step forward?
- What step are they on?
- What exactly are they willing to risk at this step: time, budget, internal attention, reputation?
- If they are stuck, is the problem weak demand, or is the next step simply too heavy?
Those questions usually make the pipeline look colder.
I am fine with that.
Better a colder pipeline than a warm delusion.
Because interest is not nothing.
It is just not movement.
Series | What Reality Corrected
Chinese series name | 被現實修正之後