So why do the same issues keep surfacing, time after time? In our experience, the answer comes down to four things, and all four are structural.
Gartner's 2025 CIO and Technology Executive Survey, drawing on responses from 3,186 CIOs across 88 countries, found that only 48% of digital initiatives meet or exceed their business outcome targets. The data and the tools are there. The structure to act on them, consistently and at pace, is what separates the 48% from the rest.
Every improvement starts with a finding. An analyst surfaces it, a researcher documents it, a team lead presents it, and the meeting ends with everyone agreeing it matters. What that chain often skips is naming one person as accountable for turning the finding into a delivered improvement, and shared awareness across a team tends to produce discussion rather than delivery as a result.
The practical fix here is straightforward. For every significant finding, assign a named owner before the meeting ends, give that person a clear outcome to deliver rather than a task to complete, and review progress against that outcome on a fixed cadence. Working with clients across financial services and utilities, we see this single structural change accelerate delivery more reliably than any tool or process upgrade.
Prioritisation decisions made through debate tend to favour the most senior voice in the room rather than the strongest evidence, and most digital leaders know this pattern well. A clear commercial opportunity surfaces from the data, a competing priority enters the conversation, and the team negotiates rather than decides. The evidence waits for the next round.
The practical answer is to agree a prioritisation framework before the debate starts, not during it. Define two or three criteria that reflect your commercial and customer priorities, assign weighting to each, and score opportunities against them consistently. When the framework does the deciding, the conversation shifts from persuasion to interpretation, and that is a far more productive use of everyone's time. Our 16-year partnership with United Utilities, which helped drive 70% of their site traffic to self-serve tools, was built on exactly this kind of shared clarity, with agreed criteria and clear decision rights in place from the outset.
When one team generates insight and hands findings across to a separate team who builds the improvements, the commercial context tends to get lost along the way. The reason a change matters, the behavioural evidence behind it, the size of the opportunity, rarely survives translation into a ticket or a brief, and the improvement reaches customers producing a smaller return than the evidence suggested it should.
The practical answer is to keep the people who understand the evidence involved through the design and build, and to build shared KPIs that connect both teams to the same outcome. Working with People's Pension, the KPI framework developed during discovery shaped decisions throughout the build and experimentation infrastructure went in from the start, so that the platform reached launch with data-driven decision-making designed in rather than added afterwards.
A copy change and a platform integration carry completely different levels of risk, but many teams run both through the same approval process and add weeks to the one that could have been in front of customers by Thursday. When governance applies the same level of scrutiny to every change regardless of complexity, the pace of improvement across the whole digital estate gets set by the most cautious interpretation of the highest-risk work.
The practical answer is to tier your change process explicitly. Define what counts as low, medium and high risk, set the approval requirements for each tier in advance, and review that framework regularly as the team's confidence and track record grows. Fast, clear decisions on low-risk changes free up the attention and scrutiny that high-risk ones genuinely deserve.
All four are about structure, and structure is something every team can change. Clear ownership, a shared prioritisation framework, delivery that stays close to the evidence, and governance proportionate to risk, these are the four conditions that allow strong data to produce consistent returns.
This is exactly where our Optimise service focuses. We work with digital leaders on an ongoing retained basis to get strategy, data, operating model and performance working together, because fixing one without the others is what produces results that are promising but inconsistent.
If your team has strong data and sound tools but performance has plateaued, we'd love to talk.