Data, Data Everywhere, But Execution is Key


Data is everywhere, and in massive amounts; but what good can all that data be to marketers if the data isn’t usable, scalable, or able to be properly utilised to their advantage? Writing exclusively for ExchangeWire, Dana Hayes (pictured below), president, ShareThis, offers some guidelines for marketers to succeed in the overwhelming world of data.

Almost every marketer has an insatiable appetite for data. Beyond targeting, modern marketers leverage data to holistically understand consumers, validate assumptions, increase their pool of prospects and, ultimately, power their analytics.

That said, not all data is created equal. Common issues are opaque data that lacks transparency into consumer behaviours, data that cannot integrate well into current systems, or duplicate and conflicting data sources.

The rise of the Data Management Platform (DMP) and the influx of hundreds of data sources created even more havoc for marketers; in the centre of the storm is one of their most powerful tools – first-party data. Limited in size and challenging to use at scale, some brands undervalue their own data because of this scarcity, but this is the wrong approach.

Product teams have long claimed that you can either have quality, speed, or price, but rarely all three at the same time. With data, it is all about scale, performance, and transparency. Logic might drive most marketers to focus on scale and performance, but it is the transparency through ownership which is the true game changer.

Marketers excel at acquiring first-party data through website traffic, email, and social media engagement, custom content, and tag management. The real challenge is integrating their data with external sources to further enhance the picture of their customers and prospects.

One of the most common problems even global brands face is the shortage of clean data at scale. Opaque data sources are challenging to apply to a campaign successfully. For example, a marketer might have success with a custom audience segment from a data provider; but with small scale and a lack of insight into the segment membership, it becomes hard to expand the audience size for increased reach and campaign delivery.

To help marketers refocus, I’d like to offer a few guides:

First-party data as the centre

First-party data serves as a powerful beacon for marketers. When marketers append their first-party data to transparent raw data sources, they can model and scale audiences that they know perform best. By focusing on quality over quantity, marketers can increase their scale without impacting the quality of the audiences or targets. This requires a level of data science that is not always available to marketers in-house. Therefore, it’s important for marketers to find a data partner with data science capabilities and/or a toolset to allow the marketer to do this modeling themselves.

Create data ubiquity

Most brands have strong data environments; but the teams who are most involved with campaign and audience planning can’t always access those systems. For example, the Analytics group might have a Hadoop environment that stores purchase data, but the Marketing team might not have access to the Hadoop system and, even if they did, the Hadoop system might not be connected to a buying system (like a DSP) that runs their campaigns. It can be technically challenging to make these connections. However, very savvy marketers work with partners who can enable point-of-care to prove the market performance, which can then lead to more holistic data practices that includes more sets of data.

Look broadly for sources

Agencies are clearly on the frontlines with many of the largest sources of data in the industry, by having multiple interactions with different companies that can provide access to data. For example, an agency can negotiate with a publisher on behalf of a marketer, or buy third-party data on behalf of the marketer, or work with retailer partners on custom marketing programmes to collect purchasing behaviour. Some agency holding companies have made a practice of collecting, aggregating, and providing the data to their marketer clients via various commercial models. For instance, some agencies focus their efforts on collecting the data for individual marketers. This is often determined by the marketers' strategies around data and if they are comfortable with 'data pools' or 'data co-ops'.


Dana Hayes, President, ShareThis

Marketers should push their agency and data partners to share as much insight as possible into where the data is coming from and how it can be used for audience insights and targeting. And when you can find a source of data that is broad, rich, and transparent, it will have a positive impact on the campaigns.

Machine learning to find the new audiences

One of the biggest game changers will be data powered by machine learning. Consider a marketer with a lot of Point of Sale (POS) data on their customers: through matching systems, onramp processes, and machine-learning engines, they find an extended group of users who look almost exactly like the original Point of Sale customers. Once these new modeled audiences are built, they can be tested to improve performance and reach, and iterated further without sacrificing quantity for quality.

Data science is the key to these strategies; and smart marketers with access to either the tools or partners to achieve these results have already increased their customer insights and audience targeting and seen lifts in campaign performance. Marketers who do not deploy these practices are not reaching their entire target audience, will miss out on the most extensive prospecting opportunities, and will likely be reaching the same users more frequently than they would wish.