Democratising professional software development

Purplebricks. A company founded to replace estate agents on the assumption that the value they provide their customers could be reduced to a workflow. On the face of it, this makes perfect sense to an engineer. Control communications, geofence locations, integrate with a payment gateway and bundle together within a Web App and you'll be saving home sellers thousands of pounds. To venture capitalists, this sounds like a revolution in waiting — something that will put the bricks-and-mortar estate agents out of business within a year or two. In 2023 Purplebricks was sold for one pound. Reality demands much more than an automated workflow.

In 2026, we are in year three of a similar hype cycle, this time prophesying that SaaS is dead and professional software development is history. The same hysterical voices who told us that estate agents would be gone within a few years are now repeating the same mistake of over-simplification. The key is in the word "service". Estimates so far this year show engineering employment has risen 11% already.

What's actually changing

What's really happening between the hysteria and the denialism is real change taking effect. Mature AI tooling has increased the potential capabilities of software engineers, allowing them to spend more precious cognitive energy solving business logic problems, rather than wasting it on syntax and dependencies.

The reason we aren't seeing serious companies "vibe code" their way out of their existing SaaS agreements is the same reason you haven't built your own house, pulled your own teeth, or neutered your cat yourself (hopefully). Expertise. Comprehension. Liability.

AI can be genuinely useful in the field of engineering, so the bigger question isn't when professional software services will disappear — they won't — but rather when will the cost of production come down?

Why we can deliver for a fixed fee

Here at Red Shift, we are leveraging AI tooling to enhance the capabilities of our engineers and lower the cost of production to our clients. This is why we can deploy an enterprise-standard mobile app in two months for a fixed fee.

What used to cost £50,000 can now be achieved much more modestly, opening more opportunities for companies to consider what's best for their software, and then having the ability to pivot towards their end-customer demands without being constrained and almost forced to stay on a path which isn't working simply because the investment was too high. That is real, meaningful change that benefits the market — although perhaps not quite so appealing a narrative to overleveraged VCs on LinkedIn.

So, when you do require professional software services, you should not be expecting to pay the same amount you would have three years ago. You should realistically be looking at anywhere between a 5 and 10x reduction in costs. It's not just about cost though: you should expect to have more face time and interaction with the development process, enabling you to have a clearer steer and genuine input into what is happening with your own software.

This also translates to the entire discovery process. As the cost of prototyping has bottomed out, the need for excruciatingly protracted discovery processes has largely gone. Why endure the torture of committee forum after review meeting in between product whiteboarding sessions, when you can develop a prototype within a week based on a conversation and iterate on that until you get the look and feel down?

An honest and responsible use of AI by software engineers should see your costs significantly reduce and friction almost eliminated. If your current software service provider is not delivering this experience, then you should be asking yourself why, and perhaps explore working with a team who does.