Preqin~4 min read

An ESG score that meant nothing to the fund reading it.

Three versions of the module shipped. The two experiments after them both failed.

  • Role: Product strategy, design, and research
  • Timeline: 4 weeks
  • Team: 1 PM, 1 engineer, me (plus investors and customer service)
  • Impact: 7 contracts
  • Platform: Web

A number with no context

Preqin had ESG data for private-market funds, but presented it as a static snapshot. Two users could see the same figure and neither could say what it meant for them.

The product needed to put a number in context: this type of fund, this geography, these industries. The modules moved from snapshot to momentum and specific gap analysis, taking three versions to get there.

Then I tried twice to finish the job those modules opened. Both attempts failed.

What I owned

I owned

  • Product strategy
  • User research with both user types
  • Three module versions and the reasoning behind each change
  • Both follow-on experiments and the read on their failure
  • Components contributed back to the design system

Decided with others

  • Feature sequencing, scored with the PM rather than set by me alone
  • What the underlying ESG dataset could support, with engineering

The problem

Two user types, and neither could make sense of the score.

“I care about what this number means for this type of fund, in this geography, investing in these industries.”

A disclosure score without context is unreadable. The same figure means one thing for a European infrastructure fund and another for a North American buyout fund. The product asked every user to make that comparison themselves.

The work wasn’t about surfacing more data. It was about moving beyond a static snapshot to show momentum over time and specific gaps against a relevant peer set.

From snapshot to momentum

What helps someone see where to look first?

Three versions of the module went out: tables of exact values, then a multi-dimensional risk view, then heatmaps. Version one used tables, which worked for exact numbers and operational detail but not for deciding where to look first. Version two represented risk across its real dimensions but hid prioritization: everything was visible and nothing was ranked.

We used heatmaps to spot risk quickly. A heatmap shows where to look but not how much, and these users still need exact figures for diligence. The tables remained underneath so exact values stayed available. Version three shipped, and the commercial results are attached to it.

The modules opened a job they couldn’t finish

The product surfaced the data but did not finish the job.

Both follow-ons failed

Could the product help prepare for the meeting?

I could put a guided preparation drawer beside the data, build a wizard that walked through priority gaps and exported a brief, or leave the job to the user. Gap analysis identified what to ask about, then stopped; every investor rebuilt the same brief by hand outside the product.

We built both. The drawer put questions beside their gaps so users could assemble a brief without leaving the ESG tab. The wizard surfaced priority gaps, made questions editable, and exported a PDF. Both moved the product from reporting into advising, making it different from the product customers bought, and neither mitigation worked. Both failed for the same reason. They challenged what the product was for. Limited partners wanted their own thinking confirmed, and handing them a finished brief took away the part they were paying for.

That was the project’s most useful finding, and it doesn’t show up in the results. A data product can credibly identify a gap and lose that credibility when it tells users what to ask about it. Identifying a gap is evidence. Recommending questions is advice. These customers were buying something to check their reasoning against, not advice.

What the modules returned

0

Contracts closed

0+

Pre-sales calls

0%

Engagement

Contracts and pre-sales calls are commercial outcomes the sales team can verify against its records, so they are the two figures here to weight most heavily. Separately, 80% of survey respondents marked the modules very helpful. That is self-reported and should be read that way. The record does not say what “60% engagement” measures or how many accounts it covers.

The finding was right. Both answers were wrong.

Investors had a real gap, and both products built to fill it failed. The drawer and wizard answered the same finding. Investors identified a gap in the product, then went elsewhere to turn it into questions.

The problem was not execution. Both crossed a line customers had drawn, without saying so, between a product that shows where to look and one that tells them what to think. I know two points on the wrong side of that line, but not exactly where it sits.

Nobody revisited this while I was there, so gap analysis still identifies work it leaves the user to finish. Somewhere between showing evidence and giving advice there is a line customers won’t let a data product cross, and I never found where it sits.

Sam Cusano