Why the customer who converts can't use the product, yet

A field note from earlier work. The target company was a FinTech.

The setup

The product was a self-directed online platform for investment advice. The output was genuinely proficient: portfolio construction logic, risk framing, allocation reasoning. On paper, a defensible offering.

Adoption was flat. Not catastrophic. Flat. Sign-ups came in. Activation didn't follow.

What the interviews revealed

Two distinct populations were touching the product.

The people drawn to it. First-time investors looking for direction. They could not parse the output. The advice was readable only to someone who already had the conceptual scaffolding. They bounced before they reached the moment of value.

The people who could parse it. Financially literate, self-directed. They did not need the tool. They had the skills to make the same decisions independently.

The product was sitting in the gap between two customer profiles, serving neither.

Why this is a known pattern, not a one-off

The research literature names this directly.

Lusardi and Mitchell, in their 2014 review for the Journal of Economic Literature, frame financial knowledge as a form of investment in human capital. They document something uncomfortable: the populations most in need of financial decision support are systematically the populations least equipped to act on it.

The product-market problem is downstream of that asymmetry.

The U-shaped adoption curve

The mechanism beneath it sits in consumer-learning research.

Moreau, Lehmann and Markman (2001) demonstrated that adoption of genuinely new products requires knowledge transfer. The customer has to map the new product onto existing mental structures.

The relationship between expertise and adoption is non-linear:

• Novices lack the scaffolding to comprehend the offering.
• Experts often don't need it.
• The middle (the segment with enough literacy to benefit but not enough to self-serve) is where the product actually lives.

The platform we were working on had been designed for a customer who didn't exist yet. The acquisition funnel was attracting the wrong end of that curve.

Where adoption actually broke

Not at acquisition. Not at conversion.

At capability transfer: the point where the customer has to do something with what they bought.

Most companies don't instrument this node, because they don't expect leakage there.

The intervention

We added an information layer in front of the tool: a structured course that taught the conceptual scaffolding, connected directly into the platform's logic.

Webinars onboarded users into the course. The course onboarded them into the tool. Adoption moved.

What we had built, in effect, was the knowledge-transfer bridge that Moreau et al. describe. We manufactured it deliberately because it didn't exist organically in the customer's prior experience.

The translation for portfolio companies

Most year-1 VCPs assume the customer lifecycle leaks at known nodes:

• Pricing friction.
• Onboarding drop-off.
• Low NRR.

These get measured because the categories are familiar.

Capability-audience mismatch is a different leak. It shows up as a healthy top of funnel, healthy intent signals, and a quiet collapse at the point where the customer has to do something with what they bought.

Sales decks describe a buyer who can use the product. The actual buyer often can't, yet.

This is more common in PE-backed Mittelstand portcos than the dashboards suggest. Products built by technical founders for technically literate buyers get sold downmarket post-acquisition, into segments that lack the prior knowledge structures the product implicitly assumes.

The growth thesis breaks not because the product is wrong, but because the buyer needs a bridge nobody scoped.

The diagnostic question

It isn't "are customers converting?"

It is "is the customer who converts the same customer who can extract value, and if not, what sits between them?"

If a portfolio company has strong acquisition metrics and underwhelming retention or expansion, the leak is rarely where the dashboard is looking. It is usually upstream of usage, in a layer nobody built because nobody mapped it.