Product · 27 Aug 2026 · 3 min read
Why we built another accounting platform
For most of software's history, the hard question was "can I build it." Assembling a team, writing the code, keeping it running: that was the moat. AI has mostly closed that gap.
Why we built another accounting platformFor most of software's history, the hard question was "can I build it." Assembling a team, writing the code, keeping it running: that was the moat. AI has mostly closed that gap. Almost any product that boils down to a website and a database is now within reach of a small, determined team. The question that matters has shifted from capability to willingness: I could build this myself, but how much would I pay someone else to do it properly, and keep doing it, so I don't have to. That shift is why finov.io exists.
The pain that started it
Anyone who has run a set of books on Xero or MYOB knows the drift. The tool that once felt sharp and specific slowly fills up with banking apps, payroll add-ons, payment processors, inventory modules, forecasting dashboards, and an AI assistant bolted on somewhere in between. None of it is optional to look at, all of it is optional to use, and the core job, keeping accurate books, gets harder to find inside the noise. That drift isn't an accident. It's what happens when a product has to keep growing revenue for the people who supplied the capital to build it.
Venture capital and the slow rot
Venture money buys speed. It also buys an obligation: the business now has to grow faster than the market it started in, indefinitely, to satisfy the return that capital was promised. A product that did one thing well eventually has to do five things adequately, because five things sell better to a board than one thing done well. This is the mechanism behind enshittification: it isn't that the people building the product stopped caring, it's that "good product" and "return on investment" quietly stopped being the same goal, and the second one has the vote. The casualty is the product manager who just wanted to make something people liked using. That job used to be about saying no: no to the feature that helps one customer at the expense of clarity for everyone else, no to the tenth integration when the first nine haven't been finished properly. Under growth pressure, saying no gets expensive. Saying yes to everything looks like progress on a slide, right up until the product is a poor third-place entrant in five categories instead of the clear leader in one.
Feature sprawl is suburban sprawl
It's the same failure mode as a city that keeps approving new subdivisions without ever finishing the roads, the transit, or the town centre that would have made any of it cohere. Each addition is defensible on its own. The cumulative effect is a sprawl nobody actually wanted, and a much harder job for whoever has to hold it together. This isn't only a venture-backed problem. Open source projects fall into it just as often, usually for the opposite reason: not a boardroom demanding growth, but the absence of anyone with the authority, or the nerve, to turn a contribution down. A maintainer who can't say no, or a governance model built to avoid ever having to, ends up with the same result as a growth-stage board: a product whose value proposition gets muddier with every release. Design by committee and design by shareholder return arrive at the same sprawl by different roads.
What we're building instead
finov.io is a bet that the more useful role for a product manager, especially now that building has gotten cheap, is as a gatekeeper rather than a feature factory. Our job is to decide what finov.io is not, at least as often as what it is. Finov.io does one thing: multi-entity bookkeeping, done properly. A double-entry ledger, bank reconciliation, invoicing and bills, fixed assets, multi-currency, and the reporting a finance team actually needs to close a period and trust the number at the bottom of it. It's built for organizations with more than one set of books, without asking them to merge everything into a single ledger just to get a consistent view across the group. There's no embedded AI in finov.io, and that's deliberate. Statutory bookkeeping has no room for a plausible-sounding number. What it needs is a system that does the same thing the same way every time, and that we can keep priced sensibly because we're not spending on infrastructure the product doesn't need. We'd rather be the clear leader in one category than a contender in five. If that's the kind of software you've been waiting for someone to build again, we built it.