The number I do not publish

kolab. has a figure we don’t put in decks. It’s our active user count.

We are early. Our download and active-usage numbers reflect that, and a number that low sitting beside a strong satisfaction figure invites a reasonable person to conclude the satisfaction figure is doing work it can’t support. So, we leave it out; and we say that we leave it out, rather than hoping nobody asks.

I have been asked whether that is a form of hiding. I think the honest answer is that it would be, if we replaced it with something inflated. Our practice is to publish what we can defend and to name what we can’t, and the difference between those two positions is visible to any funder who looks.

Here is what we do publish. When we report that 82.1% of teachers rated our spelling module ideas as extremely useful, the sentence carries the source, the place, the date and the sample: our Spelling Test Feedback Form, from a project we conducted in KwaZulu-Natal, August 2025, 39 responses. When we report 100% teacher and learner satisfaction, we name the Teaching Experience Report of October 2025 and its scope of six teachers across five schools. Six teachers is a small study, and I would rather you knew that from me.

Every number in that paragraph is weaker on its own than a headline figure with no scope attached. Together, I would argue they are worth more, because a funder can check each one.

The sector makes this difficult, and I have some sympathy for why. A CSI manager comparing four proposals sees confident round numbers in three of them and a set of scoped samples in the fourth, and the honest document reads as the least impressive one in the pile. I have watched that happen. The temptation to round up, to combine two figures into a third, or to describe an internal moderation as an independent evaluation is a commercial temptation, and it is real.

We are endorsed through internal moderation by JET Education Services. That is a meaningful endorsement from a serious institution, and it is not independent certification. The two phrases are one word apart and they describe different things, so I use the accurate one, even where the other would land better in a meeting.

The reason isn’t principle alone. A funder who discovers an overstatement discovers it after signing, and everything else in the file becomes suspect at once. A small sample named with its scope survives that inspection; a confident number doesn’t.

So, the figure we don’t publish stays unpublished, and I’ve now told you it exists. I would rather a funder hear that from me in an essay than work it out from a gap in a slide.