The tool that fixes your problem, the tool that talks to it and the subscription you forgot you were paying for. There's an app for almost everything. For most founders, that's the problem.

The reflex is understandable: A stage of the business feels chaotic, so you buy a tool to manage it: HubSpot to bring your sales pipeline under control. Monday.com to track projects. 10 AI tools to help sort your finances. Soon, you have a stack, and you're paying R5 000 a month in subscriptions, but your data is scattered.

Theunis Duminy, founder of Vertiqal, an SA AI implementation consultancy, spends his days helping businesses connect AI to the actual systems where their data lives. He's seen the software sprawl problem at every company size, and he's built a clear framework for deciding what to buy and what to build.

Some business problems are already fully solved. Accounting has rules, tax codes, and regulations. There's no competitive advantage in reinventing double-entry bookkeeping. But most of the software founders reflexively buy fall into a different category: Tools that impose someone else's logic onto your business.

"When you buy software like Salesforce or HubSpot, they force you to work in a specific way. HubSpot has an opinionated way of designing pipelines, but that's not necessarily yours. Founders end up retrofitting tools, using features the wrong way or omitting features they're paying for."

The alternative is building a custom operating system for the parts of your business that are uniquely yours. It used to require a development team. That's no longer true.

Before you open a browser and look for tools, Theunis has a test: Can you write your process on a piece of paper? "If you can't write your process on a piece of paper, software likely won't save you. It's a process issue."

Founders frequently buy software to manage a process they haven't diagnosed. A company knows its sales process isn't working, so they buy a CRM. But they never asked: how do we actually want to do sales? The result is a well-designed tool running a broken process, which is worse than no tool at all. Map the workflow first.

Some domains are genuinely solved. Accounting is the clearest example: the rules of double-entry bookkeeping are the same for everyone, SARS doesn't care about your preferences, and no version of it is uniquely yours. Buy QuickBooks or Xero and move on. Basic project management (Trello, Monday, Asana) falls into the same bucket. 

The signal for buying: the problem is solved, the logic is universal and there's nothing about it that's specifically yours.

Everything else, like how you find clients, run deals, deliver work or manage relationships, is a candidate for building. Not primarily to save money (though it does), but because building lets you imprint your own logic into the system.

"Personal software lets you imprint your specific workflow, your logic, your way of thinking about clients and deals and service delivery into the system. For the first time, you can build software that works the way you work."

Theunis uses Airtable as the foundation: One base linking companies, contacts, deals, projects, tasks and invoices. Build it once and it runs your entire business. The alternative is paying for five tools that fragment your data and create a new problem to solve.

There's a specific warning sign worth watching for: automations whose only job is to move data between two tools. If you've built a Zapier workflow just to push records from HubSpot into Monday.com, you're not creating leverage. You're patching a gap that only exists because the two tools don't share data natively, and you'll maintain that patch forever.

"Get your data structure right, and everything else, AI, automations, reporting, becomes trivially easy." If most of your automations are connecting tools rather than triggering real actions in the business, the underlying architecture needs fixing before you add more tools on top.

Theunis treats every new tool purchase as a decision that deserves real scrutiny. Before signing up:

SA founders are running lean. The seat-and-licence model that global SaaS companies are built on scales linearly with your team. R500 here, R900 there, suddenly R5 000 a month just to operate. Building a centralised custom system often costs less than one of those subscriptions.

Most enterprise tools were designed for Silicon Valley-scale companies with full operations teams. The assumptions baked in don't always match how SA founders run their businesses. Building your own means you're not retrofitting your process into someone else's product.

"One person with the right system can do what previously took 10 people. Solo operator or two co-founders can run what used to take a full department." Theunis Duminy 

When founders get this right, the effect compounds. A centralised system with clean data makes every other investment in the business more valuable. AI tools, automations, reporting, client dashboards: all trivially easy when the underlying data architecture is sound.

When it's not, every new tool you add makes the problem worse. It takes a weekend to set up the right foundation. It saves months of accumulated technical debt.

This post covers the buy vs. build decision from Theunis Duminy's Founder Collab masterclass, Automations Every Founder Needs in Their Stack: A Practical Founder OS. The full session goes much deeper. Here's what's inside:

The Founder Collab has 40+ masterclasses from SA's best operators across sales, UX, fundraising, paid media, automations, and more. Join The Founder Collab to access the full session.

Get more SA tech and business news and subscribe to The Open Letter.