Bad software has nowhere left to hide
For years a great sales team could beat a great product. AI collapsed the cost of building well — and 'it's too hard to build' stopped being a strategy.
For most of my career, a great sales team could beat a great product. You could ship something mediocre, wrap it in good distribution, and win. I watched it happen for years. That era is closing.
I'll use one of my own examples. At Riqra, we never built a visual CMS editor for our customers — the kind of control that lets a merchant change their own storefront without filing a ticket. Not because we didn't see the need. Building it well was complex and expensive, and we always had something more urgent to prioritize. For a long time, that was a defensible call.
The part that should bother you
Now the uncomfortable version. Plenty of companies had the resources to build that experience and chose not to. The math didn't justify it, so they shipped something worse and let their users live with it. Distribution covered for the product. The customer paid the difference.
What actually changed
The cost of building well collapsed. That CMS editor we kept postponing for years? We built it in two weeks. The thing that was 'too expensive to justify' became a two-week project. And when the better version costs two weeks instead of two quarters, 'it's too hard to build' stops being a strategy.
For years the moat wasn't the software. It was that better software was too expensive to build. That moat is gone.
Bad software has nowhere to hide
This is the part founders should sit with. When a small team can build the experience the incumbent skipped, every underserved customer becomes a market. We're now going to the users big companies could have served well and never did — not with a better pitch, with a better product.
So the question for any founder or CTO: where in your product are you still leaning on 'it's too hard to build well' as cover? Someone is about to find out it isn't — and walk in through that gap.
I share the public pieces of this here. When the workflow problem is already real, I also help a few founder-led teams install the system behind it.
The excuse for shipping bad software just expired.