I deleted the parts of Clarity I would have pitched

I took three things out of Clarity recently. A whole tab, a section of the insights screen, and the card layout on the history screen. Two of the three are exactly what I would have put on a slide if I were selling Clarity to someone.

That is the part worth writing about, because putting things on slides is my day job.

What selling to healthcare buyers taught me

At Hexaware I work in enterprise new logo sales, and a lot of that time is spent with healthcare buyers. When I put together a briefing for clinical teams, the easy version was a catalogue: everything that could be done, grouped neatly, one slide each.

I built it around the problems those teams already feel instead. Hours lost preparing for visits. Time lost to first drafts. Too many places to go for help. People on the clinical and regulatory side answer for cycle time and for being ready when an inspector turns up. They do not need to be shown what is possible. They need to recognise their own week on the first slide.

A catalogue invites a feature comparison. A named problem invites a conversation about the work. Selling teaches you the difference between a nod and a signature fairly quickly, and a catalogue collects nods.

The catalogue in my own product

Clarity is a ledger I build for a dentist I know. It replaced a spreadsheet he kept by hand for about a year and a half. One user, so everything below is a sample of one.

The Monthly tab. It showed revenue, costs, net and margin for a month. Every one of those numbers was already on the insights screen, in more detail, next to things Monthly did not have. A second place to see the same numbers is not a feature. It is one more tab to choose between. I deleted it, and reordered the remaining tabs by how often they are actually used.

Payout Concentration. A breakdown of who the practice pays out to, lab and consultants. It is exactly the kind of section I would put in a pitch, because it looks like insight. But in his real data, payees are not yet linked to transactions, so the section only ever showed one pile labelled Unassigned. In his hands it said nothing at all.

I took it off the screen and left the field underneath it alone, so the narration and the eval suite that check those numbers are untouched. It comes back when the payees are linked.

The history cards. Every transaction had its own card. They took up space and looked clickable, and they were not. It is a table now. Each row names the patient as a link to that patient's page, and the pager tells you which page you are on and how many there are.

How I got there

For a while, one number ran this project: how long it took to enter a transaction in Clarity compared with his spreadsheet. There was a design review built around closing that gap.

Then I asked him. He liked the entry screen. His spreadsheet was faster because it never asked who the patient was, so it was doing less work. He was not worried about the difference.

What he did bring up was a short list, and the first thing on it was too many tabs.

It is the same mistake as the catalogue. I had been reasoning about what he would value instead of finding out, and I built toward the reasoning. The question should have been asked on day one. In sales I would never walk into a room with a deck I had not tested against what the buyer actually said. With my own product I did exactly that.

What carries across

A section with nothing real behind it is a demo, not a feature. Payout Concentration would look fine on demo data. It only failed on his data, which is the only data that counts.

Duplication is not generosity. Showing the same numbers in two places felt like more product. To him it was one more decision about where to look.

Ask before you build the slide. A buyer and a user will both tell you what they will pay attention to. Neither of them will tell you unprompted.

What I have not worked out

Payout Concentration comes back when the payees are linked. I have not decided what linked means. Every payee, most of them, or only the ones that carry most of the money? A chart that is right about most of the payouts and quietly wrong about the rest might be worse than the Unassigned pile it replaced, because at least that pile was honest about what it did not know.

It is the same question a briefing asks of every slide: how much has to be true before a thing earns its place on the page.