1.6: Now Is The Time To Automate
Author: Matt Belcher, Afor Director
Author Introduction
"In more than two decades working in software test engineering, I've learned to be sceptical of anything billed as a turning point. Most 'revolutions' turn out to be incremental. The current wave of AI is the exception. It is genuinely disrupting how businesses operate, how markets move, how applications are built, and what individual roles look like day to day.
Amid all that noise, one practical conclusion stands out to me: now is the time to automate. The tools are more capable, the frameworks are more resilient, and the ROI for automation has become more compelling. This blog follows on from our AI-Enhanced Test Automation series and updates the picture for a landscape that is shifting month to month, as organisations move past experimentation and start building with agentic frameworks.
There is a second shift I want leaders to notice. The old logic of sending work offshore to chase labour savings is being overtaken. When automation is this affordable, the advantage moves back to teams that sit close to the business and can turn that context into working automation quickly. The smart play is no longer to rent effort by the hour - it is to build an automation asset onshore that keeps paying you back across the full lifecycle of your applications."
— Matt Belcher, Director, Afor
Outline
Why AI has reset the automation timing question
The rise of agentic frameworks in the enterprise
How falling automation costs change the maths
Why offshore labour arbitrage is fading
The case for bringing work back onshore
Building automation as a durable enterprise asset
Where AI-enhanced test automation fits
How to get started with Afor
Key Takeaways
AI disruption makes now the moment to automate.
Agentic frameworks are moving from pilots to production.
The cost of automation has fallen sharply.
Offshore labour arbitrage no longer guarantees savings.
Onshore, AI-enabled delivery is increasingly competitive.
Treat automation as an asset, not a project.
Lifecycle savings compound across enterprise applications.
Start with a focused, low-risk diagnostic.
Introduction
The explosion of AI has created significant disruption across businesses, markets, applications and individual roles. As organisations come to grips with both the opportunities and the risks, one conclusion is becoming hard to ignore: now is the time to automate.
The options available today are broader, cheaper and more capable than at any point in the past decade. This article follows on from our AI-Enhanced Test Automation series and updates the picture for a landscape that is changing month to month, as enterprises move beyond experimentation and begin building with agentic frameworks.
For most leaders the question is no longer whether to automate, but how quickly - and where the work should sit.
Why now? AI has reset the timing question
For years, the honest answer to "should we automate this now?" was often "not yet". Tooling was expensive, frameworks were brittle, and skilled people were scarce. Those constraints shaped a generation of technology decisions, including the business case for automation itself.
That calculation has shifted. AI has lowered the effort required to build, maintain and scale automation, while simultaneously raising the cost of standing still. Competitors are compressing release cycles, customers expect faster change, and manual or semi-manual processes are becoming a visible drag on the business.
When the barrier to entry falls and the cost of inaction rises at the same time, the timing question answers itself. The organisations that move early do not just save money - they build the internal muscle and the reusable assets that make every subsequent project faster.
From pilots to production: the rise of agentic frameworks
The most important shift of the past year is the move from AI as a helper, to AI as an operator. Agentic frameworks - systems that plan, act, observe and correct their own work rather than following a fixed script - are moving out of the lab and into production.
Gartner has projected that around 40 percent of enterprise applications will embed task-specific AI agents by the end of 2026, up from less than five percent a year earlier, an eightfold shift in a single year.
The direction of travel is clear, but so is the discipline it demands. Deloitte's State of AI in the Enterprise research found agentic AI usage rising sharply while governance lags, with only about one in five organisations reporting a mature model for overseeing autonomous agents.
The lesson is not to wait, but to adopt deliberately: start with a well-defined workflow, keep humans accountable for outcomes, and build governance in from day one, rather than bolting it on later.
The economics are shifting: automation has become more accessible
The economics that once made automation a hard sell have quietly turned. In software testing, the largest cost has never been the licence - it is the engineering time spent maintaining brittle scripts as applications change. Opportunity now exists for experienced teams to employ AI-enhanced tools and techniques to manage testing at scale more effectively. AI enhanced techniques result in rapid review making engineering resources more productive, and open-source frameworks remove high licensing fees from the equation altogether.
The result is a far more favourable return profile, which we explore in our comparison of traditional and AI-enhanced test automation. When the cost of building and maintaining automation falls, work that was previously "too expensive to automate" moves firmly into scope.
This is exactly why the timing conversation has changed, and why a clear-eyed view of automation ROI now looks so different to the one many leaders carry from three or four years ago.
The offshore calculation no longer adds up the way it used to
For three decades, the logic of offshoring was simple: take a defined, repeatable task, move it to a lower-cost location, and capture the labour arbitrage. It was a reliable play. AI is dismantling the foundation it rested on. As the Harvard Business Review has argued, AI is rewriting the economics of outsourcing by automating many of the routine, rules-based tasks that were sent offshore in the first place.
The savings were also never as large as the headline suggested. Analysis from CrossCountry Consulting notes that arrangements which looked like a 60 to 70 percent saving on paper often delivered only 30 to 40 percent in practice once coordination overhead, quality control, security compliance and the friction of distance were accounted for. As wages in traditional offshore markets have risen and AI has absorbed the routine work, that already-narrow gap has closed further.
The strategic implication for New Zealand organisations is significant. When automation is more affordable and AI helps with the heavy lifting, the advantage shifts back to teams that sit close to the business - people who understand the domain, the customers and the systems, and who can turn that context into working automation quickly.
Bringing work back onshore is no longer a premium choice made in spite of the economics. Increasingly, it is becoming more economically justifiable particularly if the Automation Asset is owned and managed locally.
Build an asset, not a project: the lifecycle view
There is a deeper shift underneath all of this. The old model treated automation as a project: a fixed cost, a delivery date, and a hand-off. The new model treats it as an asset - something you own, that compounds in value across the full lifecycle of your enterprise applications.
An automation asset built on reusable, relational test logic and open-source frameworks does not depreciate with the next sprint. It scales across platforms and pipelines, absorbs change through self-healing, and lowers the total cost of ownership every year it operates.
Crucially, when that asset is built with your teams rather than for them, the capability stays in-house. You gain internal knowledge and independence, not an ongoing consulting dependency. Over a five-year application lifecycle, the difference between renting effort offshore and owning an appreciating automation asset onshore is measured in both dollars and resilience.
“The old logic of sending work offshore to chase labour savings is being overtaken. When automation is this affordable, the advantage moves back to teams that sit close to the business and can turn that context into working automation quickly.”
Where AI-enhanced test automation fits
This is precisely the gap Afor's AI-Enhanced Test Automation offering is built to close. Too often, automation fails to deliver: heavy tool spend, brittle scripts and a shortage of skilled people mean that up to 70 percent of test automation projects fall short - a pattern we call the Automation Paradox.
Afor takes a different path. Using AI-enhanced, open-source frameworks with pragmatic self-healing, relational data structures for reusable test logic, and our ROAR delivery methodology (Review, Optimise, Adapt, Report), we help you build an automation asset that scales rather than breaks. Delivery is engineering-led and onshore, embedding capability into your Agile teams so the value stays with you.
And because it starts with a focused, fixed-scope five-day Automation Strategy Engagement, you get a clear, low-risk roadmap before committing to build - the fastest way to turn today's favourable timing into a durable advantage.
Next Steps
The disruption is real, the tooling has matured, and the economics have turned. Waiting no longer protects you from cost - it simply hands the advantage to organisations that act. The pragmatic move is to start small, prove the value on a defined workflow, and build from there.
How do I get started with AI-Enhanced Test Automation?
Contact us today to discuss how Afor AI-Enhanced Test Automation can deliver real business outcomes: https://www.afor.co.nz/contact-us
FAQs - Further reading on how to accelerate your AI Enhanced Test Automation Journey
Blog 1: Why Enterprise Test Automation Often Fails: Breaking the Automation Paradox
Blog 2: Building a Compelling Business Case for Advanced Test Automation: Beyond the ROI Numbers
Blog 3: Traditional vs. AI-Enhanced Test Automation – A Pragmatic Comparison
Blog 4: Implementing AI-Enhanced Test Automation: A Strategic Roadmap for Success
Blog 5: Measuring Test Automation Success: Key Metrics That Demonstrate Business Value
Blog 6: Now Is The Time To Automate
