Just because ice cream sales and sunburns rise together doesn’t mean one causes the other.
Yet in IT investment conversations we still see decisions justified on correlation alone. The assumption that a rise in profit was caused by a system upgrade, without acknowledging things happening in the real world that had an impact on the result.
I recently saw a Business Value Assessment used to justify a spend for a new CRM system, referring to improved profit margins. But what it didn’t mention?
During COVID, material costs soared and the company raised prices to compensate.
When costs came down, prices stayed high and profit rose.
That’s not transformation. That’s margin management.
Correlation is not causation. And in tech, that confusion can be expensive.
Correlation says, “These two things happened together.”
Causation says, “This thing made that thing happen.”
I’ve seen this confusion play out in a few familiar ways with clients across all various industries:
✨ 1. Post-Implementation Hype
“We installed a new CRM and now our customer satisfaction scores are up 20%!”
Maybe.
Or maybe, in implementing the new system the business also introduced better processes and some long-overdue training on handling enquiries. Changes that could have been made without the tech incidentally.
🤖 2. Tool Worship
More often than not now, I hear:
“AI’s the answer!”
And I echo:
“What’s the problem?”
Instead of starting with a real issue they want to solve, organisations see other successful companies using AI and assume they’ll be successful too just by association.
That’s like buying Banksy’s spray cans and expecting to end up with a masterpiece.
🙈 3. Blind Credit
One client automated several processes in a department and a month later, costs dropped. The automation got the credit.
But the real impact came from downstream effects like staff being reshuffled and roles redefined - these played a much bigger role in the savings.
Why it's an issue?
When we assume new tech caused the benefit, we risk:
Overinvesting in tools that don’t deliver ROI
Choosing shiny new systems over improving what already works
Overlooking the need for change management, training and process redesign (why does that only happen when there’s a new system?)
Repeating flawed strategies elsewhere and expecting the same magic
Never validating the investment - because most companies don’t track business benefits post-implementation anyway
If you’re not going to measure the real impact, at least start by asking the right questions.
A smarter way
Set clear baselines: What do you expect to change and why?
Develop MVPs and use pilots where possible
Track multiple variables, not just the ones that make the tech look good (or your IT Manager a hero).
Always ask: “What else changed?”
Define benefits metrics upfront and track them post-implementation to validate success
Don’t let a sexy dashboard or supplier case study fool you.
As I keep saying... it's never just tech.
Correlation might shine a light in your eye but causation needs to be earned with analysis, context and a healthy dose of scepticism.
