The unglamorous work of organising data – understanding, cleaning, enriching, labelling and consolidating it – remains a major work-in-progress at most enterprises, a new report emphasises; even as the Chief Data Officer (CDO) role comes under threat in many organisations.
“A lot of companies have made the mistake of thinking their data is ready when it’s not. They’ve invested in new platforms, in popular AI initiatives, but when it comes to that foundational first step – organizing data – many are falling short,” Ataccama’s 2025 “Data Trust” report says.
People, process… some technology too
The vast majority of enterprises are working to become more data-centric; The Stack for example recently interviewed the leader of a data project at a $50 billion revenue insurer that aims to completely rethink how it assesses risk exposure using its own existing, highly fragmented data sets (“spreadsheets in various small group’s VMs; data mapping and transformation logic and all of that; the data’s never good, that’s just the reality!”) along with geospatial and other third-party data.
Putting the technology cart before the organisational horse can be a recipe for failure however. As Ataccama’s CEO Mike McKee put it to The Stack: “Being a technology vendor, perhaps it’s weird to say, but people and process are more important [to fix data trust]; there's no doubt…”
“If the leaders of the organization, the business leaders, aren't bought in to the importance of data culture and muscling through those processes; of not just having the data team work on this, it's not going to work… “
(Banks are among the many organisations grappling with this issue, with Citi for example recently highlighting the challenges it has capturing clean data from 11,000 regulatory reports and “ensuring that the data that we capture at trade entry is the data that's required to ultimately show up.”)
Read this: Succeeding despite the AI hype
Ataccama, a $60 million ARR company where McKee took the helm as CEO in late 2023, has learned this eating its own dog food, he added – whilst working to improve the data it captured from its finance team.
The key lesson?
“Good data is [not] just gonna appear from this island of people over here, no! The business has to be involved in ‘what data do you need?’ ‘What's the problem with it?’ ‘What's the benefit of having better data?’
“It is ‘people, process, technology’ in that order.”
“The technology [sector’s] largely got this figured out,” he added. “The other 95% of companies that brew beer, and do pharmaceuticals, and make cars, and make drills, and offer insurance products, and which have been worried about their products and services not ‘AI’? [Less so..]”
That showed up in Ataccama’s Data Trust survey, he noted – 41% of respondents said they struggle to maintain consistent data quality across their organisation and 68% of CDOs identified data quality as their top challenge for the future, with more than half ranking “improving data quality” as a top priority for 2025. CDOs have to get closer to the business he said: “The pull [has to] come from the business owners who have a specific problem. [CDOs] have to be “realistic in terms of what could help move the needle… don’t try to boil the ocean” McKee added.
