The hidden tax of bending your workflow to fit your software

Every business I've worked with pays a tax they don't see on the invoice. It's the cost of reshaping how the team actually works so the software they bought stays happy.

A bookkeeper who exports a CSV every Friday because the accounting tool won't pull from the CRM. An ops lead who renames files a certain way so a Zap doesn't break. A sales rep who logs the same deal in two places because nobody trusts either system on its own. None of that shows up in a SaaS bill. But it's where your margin goes.

The frame is shifting

For twenty years the answer to almost any business workflow was "there's a SaaS for that." And often that was the right call. Building was expensive, maintenance was worse, and a $40/seat tool that solved 70% of your problem beat a six-month custom project that solved 100%.

That math is moving. Andreessen Horowitz recently made the case for what they call disposable software — small, purpose-fit apps that AI-assisted development makes cheap enough to write, use, and replace when your business changes. Bain's 2025 report on agentic AI and SaaS makes a related point from the other direction: workflow lock-in and rigid product surfaces, the things SaaS vendors used to charge a premium for, are starting to look like liabilities.

You don't have to buy the full "death of SaaS" thesis (and I don't) to take the practical lesson. The cost of fitting software to your business, instead of the other way around, has dropped.

Where the tax actually hides

When we scope a project, we look for the seams. They're almost always in the same places:

- Re-entry: the same data typed into two or three systems. - Workarounds with names: "the Tuesday export," "the override sheet," "the thing Sarah does." - Tools used as 20% of what they cost: a $1,200/month platform where the team uses one module. - Process steps that exist only because the tool requires them, not because the business does.

None of that is a SaaS failure. Generic products are generic on purpose. The failure is treating someone else's product roadmap as a constraint on how your business runs.

What to actually do about it

Not every workflow deserves a custom build. Email, accounting, payroll, CRM basics — buy them. The vendors do that work better than you ever will.

The places worth a closer look are the ones specific to how *your* business makes money. The intake form that feeds three teams. The quoting logic nobody else's tool quite gets right. The reporting your operations lead rebuilds in Excel every month.

Those are the workflows where a small, boring, purpose-fit app pays for itself fast. Not because custom software is magic. Because you stop paying the tax.

If you're not sure where yours is hiding, that's the kind of thing a discovery call is for. No pitch. Just a look at the seams.