Why “We Should Use AI” Is Not a Strategy: Setting Goals That Actually Deliver Value

Sequel to “Is Your Business AI Ready? 6 Essential Foundations Every Australian Business Must Check”

“We should use AI” is not a goal. It is a direction without a destination, and it is currently the most expensive sentence in Australian business.

In our piece on AI readiness, we named clear business goals as one of six foundations a business needs before AI investment pays off. This is the deep dive on that single foundation, because the research shows it might be the one costing businesses the most.

The scale of the problem

McKinsey’s Global AI Survey, published November 2025, found that 88 percent of organisations now use AI in at least one business function. Only 39 percent report any measurable impact on earnings. That gap, 88 percent adoption against 39 percent measurable value, is not a technology problem. Every one of those organisations has access to the same tools everyone else does.

McKinsey’s own numbers reinforce the same pattern from the other direction. Businesses that tie their AI initiatives to specific, measurable business goals are 3.5 times more likely to succeed than those that do not. Gartner’s research points at the same failure mode from a different angle: through 2026, Gartner predicts organisations will abandon 60 percent of AI projects that were never supported by AI-ready data and a defined outcome in the first place.

Why “use AI” fails as a goal

A goal needs three things to be useful: it has to be specific, it has to be measurable, and it has to make the next decision obvious. “We should use AI” fails all three. It does not say what improves, by how much, or how you would know if it worked.

Compare that to goals we help businesses define in practice.

Reduce the time spent on weekly compliance reporting from four hours per team to under thirty minutes. Specific. Measurable. The moment you set that goal, the tool choice becomes obvious: you need something that pulls data automatically from the systems already producing it, not a general-purpose AI subscription.

Ensure every client enquiry receives an automated follow-up within two hours, without anyone manually triggering it. Again specific, again measurable, and it immediately tells you the goal lives in the connection between your website, your CRM and your communication tools, not in a chatbot bolted on top.

Eliminate the manual data transfer between your CRM and your finance system, so a won deal automatically generates an invoice without anyone touching it. This goal does not even require what most people picture when they hear “AI.” It requires systems integration. That is precisely the point. A genuinely clear goal often reveals that the real blocker was never intelligence, it was connection.

What a real example looks like

One of the clearest illustrations of goal-first thinking we have worked on was not originally framed as an AI project at all. A client of ours, inCommunity, needed a way to catch tenants at risk of losing their housing before it became a crisis. The goal was specific from day one: get property managers a fast, low-friction way to refer an at-risk tenant into support services, and measure whether it actually got used.

That platform, PM Assist, which we built as inCommunity’s technology partner, delivered 60 referrals in 28 days and went on to win Innovator of the Year in the Property Management Platform category at the 2026 REB Innovation Awards. Nobody started that project by asking “should we use AI or automation.” They started by defining exactly what solved looked like, then worked backward to the right technology. That sequence, and not the tools involved, is why it worked.

How to set a goal that will actually hold up

Three questions separate a real goal from a wish.

What exactly improves? Name the specific process, report, or workflow. Not “operations” or “efficiency” in general. One named thing.

How will you measure it? A number, a time, a percentage. If you cannot picture the after-state as a measurable fact, the goal is not ready yet.

What decision does this goal make for you? A working goal should make the next choice obvious, whether that is which systems need to talk to each other, which process needs redesigning first, or which platform is actually relevant. If the goal still leaves you needing a vendor to tell you what to buy, it is not specific enough yet.

Businesses that can answer all three before spending anything on AI are, per Gartner’s own numbers, 3.5 times more likely to see it pay off. That is a meaningfully better return for doing the thinking before the buying, not after.

Where this fits with the rest of AI readiness

A clear goal does not operate in isolation. It is what makes the other five foundations, connected processes, accessible data, people readiness, a flexible technology stack, and governance, actually worth building. Without a goal, those five are infrastructure with nowhere to point. With one, they become the specific things standing between you and a measurable result.

If you are not yet sure whether your business has the underlying foundation to make a real AI goal achievable, our free health check gives you a plain-English picture in under five minutes.

Take the free health check: linkingintegrating.com/health-check

If the results point to something worth a proper conversation, the next step is a direct 20-minute call about what is fixable and what the right sequence looks like for your business.

Book a 15-minute call with Mark Cooray: bookings.cloud.microsoft/book/SpeaktoLinkingIntegrating@linkingintegrating.com

Frequently asked questions

What makes an AI goal “specific enough”?

It names one process, includes a measurable before-and-after, and makes your next decision obvious. “Improve efficiency with AI” is not specific enough. “Cut the time to compile our monthly board report from four hours to thirty minutes” is.

Do we need a data strategy before we can set an AI goal?

Not before setting the goal, but you do need it before acting on it. Define what solved looks like first. That definition will usually tell you which systems and data need to be connected to get there, which is exactly the sequence that works.

Why do so many AI projects fail even with executive support?

Deloitte’s research found leadership involvement without a defined objective is still a failure pattern. Sponsorship gets the budget approved. It does not replace the work of deciding, in specific and measurable terms, what the project is actually meant to achieve.

Is this different for a not-for-profit or membership organisation?

No, the same principle applies with different language. A goal like “reduce the time to match survey responses back to member records from a week to a day” is exactly as specific and measurable as any commercial example, and just as likely to succeed under the same research.

How do we know if our goal is realistic given our current systems?

That is what a systems and data audit tells you. A goal can be perfectly well written and still be blocked by disconnected systems or inaccessible data. The health check is built to surface that gap before you commit budget to a goal your current foundation cannot support yet.

Mark Cooray is the Founding Director of Linking Integrating, a Melbourne-based digital transformation and systems integration consultancy helping growing businesses across Victoria and Australia define measurable outcomes and build the systems foundation to reach them, before recommending any tool, AI included.

  SOURCING NOTE — verify before publishing: McKinsey Global AI Survey (Nov 2025) 88%/39% figures, Deloitte 62% no-clear-objective figure, and Gartner 3.5x / 60%-abandonment figures were sourced via web search summaries of secondary coverage, not read directly from the primary reports. Confirm exact wording and figures against McKinsey’s and Gartner’s own published pages before this goes live.