It started as a template I kept being asked for.
I invest in early-stage companies in Southeast Asia, and for years the same request arrived on a schedule. Around the annual budget, and again whenever a founder was about to raise, someone would reach out: can you help us build the projection model. Do you have a template.
I did. I sent it. And then I watched what happened to it, which was almost always the same thing. The model got built once, for the deck or for the board, and then it sat there. Nobody opened it again until the next budget or the next raise, when the whole exercise started over.
A single-effort projection is a tick-the-box exercise. It answers a question on the day it is built and nothing after.
I have been adamant about this with every startup I have worked with. A budget is not a document, it is a comparison you run every month. What did we say would happen, what actually happened, why is there a gap, and what does the rest of the year look like now that we know. Variance analysis and a rolling forecast. Without those two things, the model is a slide.
Doing that by hand is the reason it never gets done. The comparison needs the actual numbers, and the actual numbers live in the accounting. So a plan that is going to be used has to be attached to the books, or it drifts within two months and quietly stops being trusted.
So the obvious thing was to stop sending a template and build the thing the template was always supposed to become.
The forecast was the easy half. Attaching it to the accounting was where every version broke, because of how accounting is actually done at this stage.
Most early-stage companies here outsource their books. A few have an internal team; most do not. In the outsourced case, there is very little visibility between reporting periods. The numbers arrive weeks after the month ends, in the shape the tax authority asked for, because that is what the engagement was bought to produce. Accurate, late, and of very little use for running anything.
You cannot run a monthly variance against books that close on day twenty-five. You cannot roll a forecast forward on actuals you do not have yet. The plan was never going to be used, not because founders were undisciplined, but because the thing it needed to compare against did not exist in time.
The gap was never in the modelling. It was in the books underneath it.
That is how the accounting module came about. Not as a second product idea, but as the prerequisite for the first one. YourBooks automates the bookkeeping so the ledger is roughly current every day rather than correct once a month, and it feeds the forecast directly. Your actuals post themselves and land against your plan, and the variance appears without anyone building it by hand.
You keep your accountant for the statutory work. They are good at it and the deadlines are external. What changes is that the management numbers stop waiting on the compliance calendar.
Forecasting and accounting are one problem. Selling them separately is why so many founders have a model they do not trust and books they never read.
At the early stage there are two functions that deserve headcount: product and sales. Everything else has to be solved with infrastructure, because every ringgit or dollar spent on overhead is one not spent on the business. That means the founder is also the CFO, and the tools have to work for someone who is not a finance person and does not have a team to operate them.
It also means being built for this region rather than adapted to it. Salaries in local currency with the statutory on-costs that go with them, revenue in a mix of local currency and dollars, investors who report in dollars, and a forecast that does not quietly turn every variance into a currency bet. Those are not edge cases here. They are every company.
YourCFO is the template I kept being asked for, built into something that stays alive after the meeting it was made for.
Kevin, Co-founder, YourCFO
If your last model did nothing after the meeting it was built for, this is the version that would have.