Build vs Buy: This could kill you.

I had an interesting conversation with a founder last week. He'd spent 8 months building an internal tool that he could've bought for $130/month.

Made me think about all the build vs buy decisions I've made at Madsapiens. Some brilliant, some spectacularly stupid.

The stupid ones almost killed us.

Time is rarely the obvious calculation you think it is.

Yeah, that $130/month tool sounds expensive annually. But I once watched a great engineer spend 3 months building a "simple" analytics dashboard (the founder eventually shifted into Mixpanel). The opportunity cost of that developer's time was likely more expensive.

The interesting part, though? Sometimes building really is the right call.

Tools can be for internal or external use

1. When it is external, ask yourself the question "Is this something that is specifically customised for your clientele's experience? Something that your competitors can't do by using off-the-shelf solutions?" Then yes, build it. Customise it. Sell it. If nothing else. becomes a peacocking strategy

2. If internal - The primary aim of using internal tools is to automate and streamline workflows. From Slack to Monday to Superhuman, these are all tools that increase productivity. If you are using a tool internally that helps with productivity, but you need customisations that will add about >30% to your team's efficiency, then apply a decent chunk of investment and time into building an internal clone of that tool that works for your company.

For example, we have cloned the following tools for our own use internally:

1. Airtable

2. Otter - We use OpenAI's capability to build our own notes taking system that is customised for our processes

3. A few Applicant Tracking Systems - Built our own extensive ATS that caters to our needs

On the other hand, we built our own programmatic pacing dashboard instead of using existing solutions, from the very start of our company.

Expensive decision. Took forever. But it became our core IP - combining automated spending patterns with custom client reporting. Over a few iterations, we now have a product that our competitors can't easily replicate.

But let's talk about outsourcing/buying tools. People overcomplicate this.

I'd say use others' tools when:

• It's a solved problem (seriously, don't build another CRM)

• You need it fast (we use Salesforce because sales can't wait)

• It's not your secret sauce (our accounting software? All external)

Real example: We tried building our own client communication portal. Waste of time. Switched to a mix of Slack and Monday. Clients love it. Team loves it. Nobody misses our buggy portal (we named it MadChat, for goodness' sake, what was I thinking?)

Most people throw around advice like "build your core, buy everything else." Real life is messier. We bought basic DSP access when starting out (which is our core), then built our custom optimization algorithms later when we actually understood what moved the needle in programmatic advertising.

The clearest sign you're making a bad build decision? When you start with "How hard can it be to build XYZ?".

The answer is always: Harder than you think. Longer than you estimate. More expensive than you planned.

But here's what nobody talks about: The real cost isn't the building. It's the maintenance. The updates. The security patches. The "oh shit, this breaks every time we update our main stack."

Want to know if you should build something? Ask yourself: "Am I ready to maintain this forever?"

If that question makes you uncomfortable, there's your answer.