What's Next for Software?

This is the most excited I've been about software in my career.

I've found myself saying that to clients, colleagues, friends and family over the past few months.

For me, that excitement comes from a fundamental shift in the economics of software.

The cost of producing and maintaining software is collapsing, and the implications go well beyond developer productivity.

One of the most significant changes is to the economics of the buy versus build decision.

For the past three decades, buying software has usually been the rational economic decision. Building bespoke software demanded significant upfront investment, specialist engineering teams and an ongoing commitment to maintenance. For most organisations, the economics simply didn't stack up.

Packaged software solved that problem, but it introduced another.

Every application embodies somebody else's assumptions about how a business should operate.

Assumptions about business processes.

Assumptions about data.

Assumptions about workflows.

Assumptions about approval hierarchies.

Assumptions about user roles.

When those assumptions align with an organisation, packaged software delivers enormous value.

When they don't, organisations are forced to adapt. They change business processes, invest heavily in customisations or create workarounds outside the platform. Anyone who has spent time in enterprise technology has seen monstrous Salesforce objects, sprawling Excel spreadsheets and manual processes created simply to compensate for software that doesn't quite fit.

They're all born out of people fighting software.

Much of the discussion on AI has focused on developer productivity, but I think the bigger shift is economic.

AI reduces the cost of building software, but it also reduces the cost and risk of maintaining it. Existing applications become easier to understand, enhancements become faster to deliver and long-term ownership becomes less of a burden.

That changes the calculation.

Capabilities that were previously too risky to justify can now make commercial sense. Organisations have greater freedom to build software that reflects how they actually operate rather than adapting themselves to generic off-the-shelf products.

Packaged software will continue to play an important role, but the economic boundary between buying and building is shifting.

I don't think organisations can evaluate that decision using the same mental model they would have used even a year ago. As the assumptions underpinning the decision change, so does the trade-off between buying and building software.

This shift is a big part of why we've been investing in Digital at Vivanti. There's a growing category of problems that would have been difficult to justify solving with bespoke software even a couple of years ago, but where the economics now look very different.

That's the part I find most interesting. Not simply using AI to build the same software faster, but reconsidering what is worth building in the first place.

What This Means for Vivanti

This shift is the reason Vivanti has invested so heavily in its Digital capability.

Our focus is on three things:

1. Identifying opportunities that have become commercially viable because of AI.
2. Building and validating those opportunities rapidly.
3. Operationalising successful solutions so they can scale securely across the enterprise.

We believe the organisations that gain the greatest advantage from AI won't simply use it to make existing processes faster. They'll use it to build capabilities that previously weren't possible.

If these are conversations already happening within an organisation, or there is curiosity about where AI changes the buy versus build equation, we'd welcome the opportunity to discuss it.