95% of AI Investments Are Underperforming. Here’s How to Fix That.

AI is Everywhere - But Are Businesses Getting Value?

AI is everywhere, and development teams are building amazing capabilities. Yet, a 2025 report by Boston Consulting Group (BCG) reveals that only 5% of companies are achieving measurable value from their AI investments - things like revenue growth, cost reductions, and improved productivity through better workflows.

I tend to flip stats on their head. To me, that means 95% of companies aren’t seeing measurable value from AI. If I’m honest, I find it baffling.

Didn’t we learn anything from the 2010–2020 digital marketing tailwind?

We’ve Been Here Before

Back then, I watched teams chase every new ad channel - frustrated that their “random acts of digital” were hammering cashflow without generating money when it was actually needed.

They’d failed to consider the compounding costs that come from working outside of a system:

• Promoting products through high-cost keywords that crushed delivered margins

• Pumping spend into platforms that didn’t pay out for 60 days, while overheads were due Friday

• Running campaigns for products the warehouse hadn’t even received - while promising 24-hour delivery

The Real Lesson

Businesses don’t win on individual tactics or siloed working where everyone’s too busy defending their department’s KPI’s to focus on the fact that there’s only one real goal they need to focus on: keeping the lights on. For that to happen, they need to build a system.

Profit and cashflow don’t come from isolated wins in marketing or sales, but from how every part connects:

• The story that earns attention

• The product that converts it

• The price that holds margin

• The channel that delivers on time

And now, AI sits across all of it - amplifying what works, and exposing what doesn’t.

Every Channel Needs a Job

Back in those early years, I started designing marketing systems that connected sales, marketing, finance, and tech. I pooled the needs of every team to build eCommerce strategies that were intentional about positioning, products, prices, and channels.

Every channel had a job:

• Instagram for storytelling

• TikTok for engagement

• Amazon for cash release

• DTC for fast cash conversion

AI Needs the Same Discipline

Once you’re clear about what you want from it, AI starts giving value back.

Are you aiming for topline growth? Revenue uplift? Per-customer value lift?
Or bottom-line efficiencies - faster order processing, reduced support time?

Decide first.

AI is a tool, not a magic wand - you need to treat it like a strategic lever, not a shiny toy. Make sure every initiative - automating workflows, predicting demand, personalising recommendations - can be mapped to a specific commercial outcome.

Them - and only then - does the ROI become visible and teams finally see the impact of their work.

The Rally Cry

So yes, this is my first post on Substack and it’s kind of setting the scene for what’s to come. Think of it as a rally cry:

• Define the commercial goal first.

• Design the AI initiative second.

• Track impact metrics alongside technical metrics.

• Prioritise projects that create predictable profit and tangible business value.

And before all of that… spend time really understanding what you want.

What’s Next

Over the coming weeks, I’ll share frameworks and approaches that help teams design AI initiatives with measurable commercial impact - so you can turn marketing and AI investments into profit systems - not random spend.