Why Confidence, Not Just Quality, Determines Data Success
Director of Consulting & Project Delivery Phin Smith shares his perspective on what really makes technology projects succeed – and why clean data alone doesn’t guarantee adoption:
In my last article, I spoke about why technology projects often fail – not because of the tech, but because of misalignment across people, process, and leadership. Get those right, and you’ve got the foundation for a successful delivery.
But even when the right people help deliver on time, there’s another, longer-term factor that determines whether a data programme actually succeeds: confidence in the data.
Clean data doesn’t mean trusted data.
You can invest in the best platforms, bring in governance frameworks, and map every lineage path from source to report. But if your teams don’t trust what they see, none of it lands. The dashboard gets questioned. The report gets manually checked. The process gets paused.
And as more firms begin exploring AI and automation more seriously, they’re realising that confidence in the data is non-negotiable. These technologies don’t work without trust in the inputs. If users are already questioning standard reports, they’re not going to put faith in something opaque, algorithmic, and unexplainable. Before any of the benefits of AI or automation can land, the underlying data has to be believed.
This is where many programmes quietly stall – not in the delivery, but in the adoption.
A client recently told me about a new BI reporting tool they’d rolled out… slick, powerful, and expensive. But users weren’t touching it. Instead, they walked over to a developer’s desk and asked for a manual extract from the database. Not because they didn’t understand the tool, but because they trusted the developer to give them the right data. And, as the client pointed out, it also gave them someone to blame if the numbers turned out to be wrong. That small act of bypassing the platform says everything about how confidence works in practice.
We see it again and again: the systems are live, the integrations are working, and the reporting is automated. But the users hesitate. Not because it’s unfamiliar, but because they’re not confident the data reflects reality. So they double-check. They rekey. They slow things down.
That confidence gap is where operational efficiency gets lost and automation fails to take hold. It’s the difference between delivering a system and delivering value.
And the frustrating part is, it’s rarely about actual data quality. Most firms have made major strides in standardisation, data management tooling, and stewardship models. But confidence isn’t built by policy or platform. It’s built by experience – over time, with people, through outcomes.
The programmes that stick aren’t the ones with the most sophisticated architecture. They’re the ones that:
- Embed close to the business
- Involve users early and visibly
- Deliver something small, fast, and useful, then iterate
Crucially, they make people feel part of the process and not subject to it.
Confidence builds when the output reflects what users expect to see, and when it doesn’t, they can understand why. It builds when issues get fixed quickly and transparently, not hidden behind data teams. It builds when the data enables a better decision, not just a faster one.
The problem is, most firms try to solve this top-down. But confidence isn’t imposed. It’s earned.
That means shaping data delivery around real use cases, not just reference models. Sitting with portfolio managers, operations leads, and product teams to understand where trust breaks down. Showing how data is sourced, enriched, and validated. And giving people the ability to challenge it – because being able to question data is often what makes it feel safe to use.
It also means treating confidence as something that needs sustaining. Because data doesn’t sit still. New funds get launched, new systems get integrated, new teams come in with their own expectations. Confidence isn’t a one-off milestone – it’s a continuous relationship between the data and the people who depend on it.
When that relationship is strong, you see the difference everywhere. Faster decision-making. Less rework. More scalable processes. Fewer handoffs and second checks. And critically, a clearer path to unlocking the real value of your data investments.
Without confidence, none of that is possible. With it, change sticks.
If your data programme is struggling to get traction, the issue might not be the quality of the data. It might be the confidence in it. And fixing that could be the most valuable thing you do.
If any of this feels familiar – or if your data programme isn’t landing the way you hoped, feel free to drop me a message.