In this article
Why is software so much cheaper to build now?
Experienced engineers now use AI-assisted tools for much of the routine work: drafting, testing and connecting systems. People still design the product and decide what goes into it, but the gap between an idea and a working product has shrunk to weeks. Almost everything else in this article follows from that.
Bryan Simkins at Transparent Partners describes small teams bringing a product concept to life in weeks.
For an SME this is good news. A custom tool used to need a large budget and a lot of patience. Now it is a realistic option, and you have more choices than you did a few years ago.
It also means the build is no longer a reason to pick one software development agency over another. Every AI software development company can build quickly. Speed is the entry ticket, not the difference.
What is workslop, and why does it matter for software?
Workslop is AI output that looks finished and has nothing in it. Researchers at BetterUp Labs and Stanford's Social Media Lab coined the term in Harvard Business Review in September 2025, to describe polished but empty memos, reports and emails at work. Their survey of 1,150 US desk workers found that 41% had received some in the previous month, and each case took nearly two hours to sort out.
I think the same thing is now happening to software. Products get built quickly, look like every other product, and stop at the point where the demo works. The login is fine. The dashboard is fine. Nobody decided what the thing was for.
Simkins tells the story of a founder who built a credible-looking platform in six weeks, then asked how long it would take to win a competitor's customers. His answer was probably never. A working product and a working business are different things.
What does taste mean in software?
Taste is judgement you can repeat. In a software project it shows up in three decisions, and all three happen before anyone writes much code.
The first is the right problem. Most failed builds solve a problem the business did not really have, or treat the visible symptom instead of the cause. Choosing the problem is the most valuable decision in the project.
The second is a clear idea of what good looks like. Someone should be able to describe what the finished thing does for the person using it on a busy Tuesday afternoon. If nobody can, the team will build whatever is quickest, and it will look like everyone else's.
The third is what to leave out. The VC Corner argues that taste is mostly editing, and I agree. Features are cheap to build now, which makes it tempting to build too many. Every extra screen is something your staff have to learn and someone has to maintain.
Why does the order of the work matter so much?
The order is the strategy. Two teams with the same feature list will get different results if one of them builds the part that pays for itself first.
When building was expensive, the budget forced you to prioritise. It no longer does. You can build ten things in the time one used to take, so the discipline has to come from people.
I would rather ship one tool that a team uses every day, and learn from it, than launch a platform that does eight things adequately. The first gets used. The second gets a training session and a login nobody remembers.
Where does taste come from?
It comes from shipping products to a lot of people and watching what happened next. You learn what good looks like by getting things wrong in front of real users and fixing them.
That is why we built Naitiv around senior people rather than headcount. Simon Rutter was Executive Vice President at PlayStation. He was part of the growth from PlayStation 3 to PlayStation 5 and of all the PlayStation digital services. Joel Baynes was Chief of Staff at UNiDAYS, which millions of students in the UK and worldwide use, and COO at Profusion and at Ecologi. He also worked at Rakuten. Alex Price was CTO at Ecologi, where he built the backend and software, and CTO at Just (justabout.com).
My own background is fintech products, PlayStation's BAFTA-winning Buzz! franchise, which tens of millions of players use, Infinite Inside, and the Peaky Blinders and Doctor Who games at Maze Theory.
None of that makes us right about your business. It means we have seen what happens when a product meets people at scale. You can read more about the team behind Naitiv.
How does Naitiv use taste on a real project?
We start with a business consultant, then build a prototype, then publish the price. The order is deliberate.
The consultant comes first because the first job is finding the problem worth solving. That conversation decides what good looks like for your business before any code exists. If you want the wider view first, our guide to working out where AI fits in your business covers it.
Then we build a prototype quickly. AI-assisted tools make this affordable, so we use them. You see something working early, and the prototype is where our judgement gets tested against your staff and your customers rather than a slide.
Then you see the cost. We publish what builds cost, so you can judge us before you speak to us. As one example of the result, our work on the BUUK pitch took each pitch from three days to five minutes. We also built LeadScout.
There are cases where you should not hire us. If an off-the-shelf product already does most of what you need, buy it. Build only when the software would be the thing that makes you different, and our piece on when bespoke software is worth it goes through that decision. And if you already know exactly what to build and have a spec, a cheaper team may suit you better. You would be paying us for judgement you already have.
If you want to test this, book a first conversation. Bring one process that costs your team time every week. We will talk through what the problem is, whether to build or buy, and what a prototype would involve.
