Impact thinking: solving the return on data and AI investment challenge

For more than a decade, organisations have poured time, talent and millions into data and AI. New platforms, bigger teams and smarter dashboards. All built on the promise of better, faster decisions and ‘becoming data driven’.

Yet many leaders we speak to still ask the same question:

“We’ve invested so much, but what’s the return?”

It’s not that data and AI don’t work. It’s that we’ve spent too long in the build phase focused on platforms, pipelines and building capabilities, rather than focusing on outcomes and real impact we can touch and see.

So, now comes the impact phase. It has to. The phase where data and AI deliver measurable business performance improvement.

Big investment, small impact

The issue here isn’t lack of intent, it’s structural. Data and AI capabilities are often built as back-office functions, disconnected from operations and the decisions that drive profit and performance.

When they’re treated as technical assets in this way, rather than business enablers, they rarely change what actually happens in the business.

Equally, data has often been built for availability, not accountability. It’s judged by delivery, not by measurable business results.

And you see it everywhere:

Dashboards that don’t drive action
Models that never make it past the testing stage
Capabilities that make good PR but fail to move the numbers

The result of all this is a divide between the investments that are made, often well intentioned and in good faith, and the tangible and intangible results that get delivered.

From data projects, to business impact

At Cynozure, we have always focused on bridging this divide by translating data and AI ambition into measurable business outcomes.

We believe data and AI shouldn’t just support strategy, they are strategy. Important partners in delivering business results. That starts with changing how we think about what gets built and how.

Rather than treating data as a collection of dashboards or pipelines, organisations should design data and decision products – solutions that directly influence the operations and decisions that shape business performance.

A pricing tool that helps improve rates dynamically and increase margins
A forecasting model that optimises stock levels to reduce wastage
A triage product that speeds up claims decisions, reducing the cost of processing
A customer data set that enables call centre agents to resolve issues faster and improve customer experience

Each one designed around a real action or decision, owned by a business leader, with measurable impact built in. This is because when operations and decisions get faster, smarter and tied to P&L outcomes, then data starts to pay.

But moving from investment to impact isn’t just a technical challenge; it’s a cultural one.

Many organisations already have capable data teams, but they’re stuck building and proving outputs, not outcomes. Meanwhile, business leaders are hungry for results but unsure how to translate data capability into return on data and AI investment.

It’s where data and AI stop being technical investments and start becoming strategic assets.

Bridging the gap between data and results – impact thinking

Most data programmes stall between delivery and impact, in what we call the impact gap. That gap between what gets invested and what gets recognised as value.

Bridging that gap takes both cultural and operational change focused on building around outcomes, with agility and in collaboration:

A product mindset: data teams build for outcomes, not outputs
Co-ownership: business and data leaders share accountability for results and KPIs
Measurement rigour: financial and non-financial returns are tracked and validated to capture the value story
Adoption by design: tools are built and integrated into how people already work

When aligned with a clear process, this ensures impact is built in from the start, carried through and not a surprise at the end. The journey goes like this:

Define – Identify which business inputs and outcomes to focus on
Deliver – Build and embed the data and decision products into real processes
Adopt – Drive consistent use through intuitive design and incentives
Quantify – Measure outcomes against business KPIs
Scale – Replicate what works, retire what doesn’t

Measuring what matters – return on data and AI investment

To truly understand impact, it’s important to have a consistent way to measure and communicate it. Return on data and AI investment quantifies the value an organisation achieves from its investments in data and AI.

It is a structured approach to understand how data initiatives translate into tangible business outcomes, operational efficiencies and strategic advantages.

Return on data and AI investment can be applied across multiple levels, from individual use cases to portfolios and overall capability performance, helping leaders determine whether data investments are delivering real-world impact.

At its core, return on data and AI investment helps answer three critical questions:

Which initiatives deliver the highest return and where should we double down?
Are we getting enough value from our data and AI investments?
Is our data capability evolving to support strategic objectives over time?

Measuring return on data and AI investment is inherently challenging because data investments span a wide spectrum, from direct use-case driven initiatives with known outcomes to foundational enabling capabilities.

The key is to focus on the work and activities that deliver the greatest impact and return on data and AI investment in a way that is clear, consistent and communicated from start to finish.

When data starts to pay back

Our learning from all this is that transformational gains can be made by approaching data and AI in this way – focusing on return on data and AI investment, building data and decision products and applying impact thinking.

When you get this right, it moves you from a world that looks like this:

Thousands spent on data and AI, but returns fall short
Platforms, tools and dashboards don’t improve decisions
Slow, costly processes weighed down by technical debt
Data teams stuck in the back office, not shaping the business
Poor or unclear return on data and AI investment

To an organisation that looks like this:

Portfolios of data and decision products delivering return on data and AI investmen
Data and AI integrated into real business processes and workflows
Faster, smarter business actions tied to P&L impact
A sustainable execution engine that scales and evolves
Positive return on data and AI investment

This is the transition every modern organisation now needs to lean into, and when it truly takes hold, change becomes visible rapidly. At Cynozure, we’ve supported clients to make this transition, for example:

Logistics company: 7.9% increase in revenue through optimised customer renewal price modelling
Charity organisation: £15m external funding from the NHS secured through transition from legacy systems to secure, scalable cloud platform
Entertainment business: 120% increase in ticket sales through targeted marketing that shifted customers from third-party sellers to own sites

These aren’t isolated wins, they’re examples of what happens when data is designed for outcomes, impact and repeatability.