How cloud v2 + AI are flipping the build vs buy equation

• Cloud v2 eliminates infrastructure headaches—no architects, no ops team, no upfront setup.

• AI-powered teams shrink costs and timelines, making custom builds the smarter play.

• For the long tail of small projects the “buy” process slower and more expensive than just building and iterating.

A shift from buying to building

“ We’ve decided to replace the off-the-shelf software we currently buy from external vendors and develop our own custom solutions instead. By moving away from one-size-fits-all products, we’ll be able to focus on deeper customization, greater personalization, and a real explosion of creativity. “

That quote from Cathie Wood, CEO and founder of the $20 billion USD ARK Investment fund, perfectly captures the shift we’re now experiencing. For nearly fifty years, enterprise technology followed a "buy, don't build" philosophy. But that world is changing fast.

We’re entering a new age where it often makes more sense to build small, targeted technology solutions between and around our existing systems, rather than embarking on lengthy procurement and package modification exercises to onboard yet another third-party platform.

Why this makes sense now

This shift hasn’t come out of nowhere. Two major changes have made building viable again: cloud v2 and AI-powered teams.

Cloud v2: no-infrastructure hosting

Even before the recent AI boom, a quieter revolution had already begun. It started with the arrival of a new breed of hosting services—what we might call "cloud v2"—offered first by Google and then by the other major cloud providers.

In 2010, about half the budget for an integration or digital project was spent just getting infrastructure in place. Even though we were "in the cloud", teams of specialists were still needed to architect, provision, configure, secure, and test the environments. Add to that another 20% wasted managing downstream technical dependencies. That left only 30% of the budget to actually build applications and automations.

Operational costs told the same story. Supporting infrastructure often meant increasing internal headcount, while vendor hosting fees could add $50,000 a month or more—even for a modest solution.

Cloud v2 changes the game. These modern, automated services are pre-architected, pre-designed, pre-secured, and available on demand. We pay only for what we use. There's no need to build or manage infrastructure, no need to add infrastructure specialists to the team. We just deploy code and start using the solution.

AI-powered project teams

As I wrote recently in "From planning to rollout: AI speeds up every step of a project", we can now build the application side of a solution much more quickly and cheaply than we could just two years ago.

Project teams that effectively harness AI—across analysis, design, UX, coding, and testing—can reduce the team size from ten people to three or four, often with faster delivery.

From talking to building

Either of these developments—cloud v2 or AI-powered teams—would be transformative in isolation. Together, they're profoundly reshaping how we deliver technology.

In many cases, it's now faster and cheaper to build and test a working prototype than it is to write the business case and requirements document. And if we throw out a few versions along the way, so be it—we're still ahead.

We also skip the 6-to-12-month vendor selection and onboarding process, which means we realise business benefits earlier.

Perhaps most exciting is the long tail of small projects that previously didn’t stack up. Their costs outweighed their benefits. That equation has flipped. With today’s tools and techniques, these once-unviable opportunities are now viable—as long as we get better at making quick, confident decisions.

When you're ready, here's how I can help:

Andrew Walker 
10 years AI strategy & implementation for executives in staff-heavy organisations, often with mobile workforces.

Did someone forward this email to you? Want your own subscription? Head over here and sign yourself right up!

Back issues available here.