Sometimes the wonkiest changes make the biggest difference.

Nearly 50% of the marketers, publishers and tech developers in North America surveyed by Winterberry Group in 2017 said that centralizing ownership of data would be one of the most important changes that their organization could make to derive value from their data. This tied with dissolving silos between groups as the most popular response.

These responses indicate that there is a lot of friction in accessing and controlling data. As programmatic advertising ballooned, marketers found themselves reliant upon a slew of data-centric vendors such as demand-side platforms (DSPs), data management platforms (DMPs), data lakes and ad verification firms. As more companies began to access a given piece of data, ownership of data became a touchy subject matter for some marketers.

“Centralizing data ownership has been a big focus as advertisers take programmatic and data management contracts in-house to gain a complete view of their consumer,” said David Lee, programmatic group lead at ad agency The Richards Group. “This has allowed clients to see where the gaps in their data are.”

To make things more complicated, different brand marketers prefer different vendors, which means that it isn’t unusual for a single ad agency to use a handful of ad verification vendors. These vendors may operate in similar ways, but the ad buyer is left navigating five different dashboards for doing the same task across campaigns. Opening up additional dashboards sucks time out of the day, which is problematic for ad buyers that are already suffering from a talent shortage in programmatic advertising.

Making data ownership and access more seamless will become increasingly important for marketers since programmatic advertising is expected to account for four out of five US digital display ad dollars by 2019, eMarketer estimates. The growing influence of data in advertising was also reflected in the Winterberry Group survey. Only about 10% of respondents currently classify their organizations as extremely data-centric, but 44% expect that their organizations will become extremely data-centric by 2019.