September 15, 2026 · John Halvorsen-Jones

Moving On From Clunky, Expensive SaaS: When to Replace It With Custom Software

Ageing vertical SaaS costs more each year and fits less. Three signals that say replace, not renew, and why AI-assisted development changed the maths for NZ.

An old beige monitor showing a cluttered grey enterprise software screen beside a stack of worn licence binders, with a laptop running a clean modern web app, Auckland skyline at sunset behind

Why are NZ businesses moving on from clunky, expensive SaaS?

Many ageing vertical SaaS products now cost a small fortune and fit poorly. Until recently the rational option was to pay and work around them. Putti's view from Auckland is that AI-assisted development has changed that maths for New Zealand businesses, and replacing a bad-fit platform with software you own often pays back faster than renewing.

Many ageing SaaS products are showing their age, whilst still costing a small fortune. Not only that, many aren't even a very good fit for the businesses they serve.

In the past most businesses had little choice but to pay and find a way. AI-driven development is rapidly changing the equation. The ROI on replacing overly expensive, often outdated, SaaS products with fully custom products, specific to individual business needs, is now stacking up.

Why does the SaaS honeymoon end?

SaaS is brilliant when you're small, your workflow is generic, and you can't justify building anything yourself. That's most businesses, most of the time, which is exactly why SaaS became the norm.

But for a growing number of Auckland, Wellington and other NZ businesses, the maths on that has quietly flipped for certain packages. Organisations are finding themselves paying more each year for software they don't fully use, wrestling integrations that were never designed for their workflow, and watching a roadmap unfold that doesn't suit them. This is especially true of older packages, or highly expensive packages for niche verticals.

Ultimately the limitations of SaaS can end up hampering innovation too.

What are the three signals it's time to replace, not renew?

This isn't a call to rip out every tool you own. Most SaaS is still the right call. But watch for these three signals.

Cost has stopped scaling sensibly. Per-seat pricing that made sense at 10 users starts looking absurd at 60. If your annual SaaS spend is climbing faster than your headcount, you're not paying for software anymore. You're paying rent on your own workflow.

The workflow fights you, not the other way round. If your team spends more time bending their process to fit the tool than the tool spends helping them work, you've inherited someone else's assumptions about how your industry operates. Vertical SaaS is built for the average customer in your category. You may not be the average customer.

The vendor roadmap has stopped moving in your direction. Every SaaS company eventually chases its biggest customers, its easiest features, or its next funding round, not your niche requirements. If the features that matter to you have sat in "planned" for 18 months, they're probably not coming.

None of these alone is fatal. All three together is a business case.

Why did "build it ourselves" use to be a bad answer?

For most of the last decade, custom software meant a 12-month build, a six or seven figure invoice, and a real chance the finished product still missed the mark. That risk profile made "just live with the SaaS" the rational choice, even when the fit was poor.

That calculus has changed. Professional AI-assisted development has greatly reduced build timelines without collapsing quality, assuming it's done by people who know what they're doing, not a vibe coder. A scoped, well-architected internal tool or customer-facing app that once took a year can now be shaped, built, and released to your team in only a few months.

This is the real argument for custom web apps to replace vertical SaaS. It was never that software is fun to build. It's that the build risk that used to justify tolerating a bad-fit platform has largely disappeared.

What does replacing legacy SaaS with custom software actually look like in NZ?

We're not talking about rebuilding Xero or Salesforce. We're talking about the narrower, painful stuff: the container-tracking spreadsheet-turned-SaaS-turned-nightmare, the franchise ops tool nobody quite trusts, the customer portal that was clearly built for someone else's business first.

For NZ mid-market companies, the pattern is consistent. A vertical SaaS tool gets adopted early because it's the only option in the category. The business grows into something more specific than the tool anticipated. The workaround stack grows. Eventually, someone does the maths on licence fees plus lost productivity plus the cost of never quite getting what they need, and realises a purpose-built application, owned outright, pays for itself faster than another renewal cycle.

This is why SaaS replacement in Auckland is becoming less of a fringe conversation and more of a boardroom line item. It isn't about chasing the shiny new build. It's about businesses doing the unglamorous work of asking whether their tools still serve their operation, or whether their operation has started serving the tools.

What's the right order of operations?

Don't start with "should we build custom software." Start with "where exactly is the SaaS failing us? Cost, fit, or roadmap?" That answer tells you what problem you're actually solving, and it's the difference between building something that fixes the real problem and building an expensive twin of the tool you're trying to escape.

If the answer points at genuine misalignment on one or more of those three fronts, and not just mild annoyance, that's when to seriously explore a custom software replacement for legacy SaaS in NZ. Get the scope tight, get the build fast, and get out of a subscription that was never really built for you. A working proof of concept is usually the cheapest way to find out whether the replacement holds up before you commit to the full build.

Renewing out of habit is the expensive option. It just doesn't feel that way until you've stopped paying for it.

Talk to Putti about replacing your SaaS

Putti builds custom software for mid-market NZ businesses, using AI-integrated delivery to move at a pace legacy development timelines can't match. We've delivered 100+ projects since 2010 with zero failures, for clients including Spark, Fletcher Building and Hydraulink. If your SaaS platform has stopped fitting your business, let's talk about what replacing it could actually look like.

Frequently asked questions

  • When should a NZ business replace vertical SaaS with custom software?

    When at least one of three things is true, and ideally all three. Licence cost is growing faster than headcount, the team spends more effort working around the tool than working in it, and the features you need have sat on the vendor roadmap for more than a year. One signal is a warning. All three together is a business case.

  • How long does it take to replace a legacy SaaS product with a custom web app?

    A scoped, well-architected internal tool or customer-facing app that once took a year can now be shaped, built and released in a few months using professional AI-assisted development. The timeline depends on scope, the integrations involved and how much of the old system's data has to move across.

  • Is replacing SaaS with custom software cheaper than renewing?

    Often, once you count licence fees, lost productivity from workarounds and the cost of features you never get. Custom software is a one-off build you own, rather than a per-seat subscription that rises every year. The comparison only works when the SaaS is a genuinely poor fit. For generic workflows, SaaS usually stays the cheaper option.

  • Do we have to replace the whole system at once?

    No. The most common pattern is to keep the core system of record running and build a custom layer for the part that hurts most, a customer portal or an operations tool for example, then migrate further only if the numbers justify it. Putti's systems integration work is usually how the old and new pieces talk to each other.

  • What is the difference between AI-assisted development and vibe coding?

    Vibe coding means generating an app quickly with AI tools and little engineering discipline. It's fine for a prototype. Professional AI-assisted development uses the same speed inside a proper engineering process, with scoping, architecture, testing, security and support. The first gets you a demo. The second gets you software you can run a business on.

  • Who builds custom software to replace legacy SaaS in New Zealand?

    Putti, an Auckland custom software and AI agency founded in 2010, with 100+ projects delivered and zero failed projects. The team scopes the replacement, builds it with AI-assisted delivery and supports it long term from Auckland.

← Back to all posts

Got a project in mind?

Let's talk about what you're trying to build, fix or improve.