“The Sell Sider” is a column written by the sell side of the digital media community.
Today’s column is written by Declan Owens, digital analytics expert at Piano.
As the deprecation of third-party cookies looms, the vision of first-party data as the backbone of advertising is clearer than ever. But before they make the switch, brands and publishers alike must ensure the data they’re collecting meets quality standards for reliability and security.
Here are six ways to audit your organization’s data quality and improve data flow through your organization:
Control the completeness of your data collection
Data sampling is a widespread practice that analyzes only a subset of your data, which is used to estimate overall results. But when it comes to key decisions, an estimate isn’t enough. Your data should also capture all cross-device user behavior, maintaining a full view of your customer and your relative performance. Complete data sets, available with no sampling, are the best way to avoid undermining your organization’s decision-making with skewed information that may not fully reflect reality.
Audit your tagging regularly
Data owners should have access to easy quality control and dependable tagging, including regular procedures such as automatic testing that will allow you to check for the presence of all tags and verify data reliability. Tools to facilitate tag audits are critical, especially when you’re updating your digital platforms.
Ensure the accuracy of your measurements
It’s important to have full transparency over how metrics are calculated by understanding what goes into your provider’s data pipeline and how the metrics are created. To understand your real traffic volume, you’ll also need to identify and exclude traffic from bots that browse your sites. Robot traffic can account for more than half of all web traffic, so this can significantly sway stats if it’s not addressed.
Additionally, many ad-blocking extensions and browsers block some trackers by default, even though the legal context (consent, for example) gives you the legitimacy to collect this data. To address this, it is important to deploy first-party domain measurement to take back control and encourage trust, notably in the relationship between your customers and your business.
Constantly clean your data to enhance its integrity
Make sure your data is accessible, well-formatted and being captured exactly as planned. By testing how your data is retrieved and presented, you ensure your values are correctly displayed on your reports. You should check whether you have fair and verified values of the numbers in your analysis during collection.
With personalized tools for tagging that set custom data-processing rules, you can correct errors due to tagging problems, enrich the data collected and exclude unwanted traffic. Tools that have automated quality control can allow you to independently update your data when needed without relying on technical support to change the code.
Centralize data around a single reference point
Implementing effective data governance promotes consistent use of data by establishing a single point of truth. Common analysis tools facilitate sound decision-making based on data that is collected, calculated and processed in the same way. With common reference indicators for consistency, you identify which indicators, based on their definition and calculation, should be used, for example, as performance benchmarks for each business function or to run macro-level reporting.
Maintain strict regulatory compliance
By aligning your approach to data collection and processing with relevant data protection laws, you guarantee by default that your data is accurate and compliant. You should have total clarity on the data you’re collecting, how it’s processed, where it’s stored and for how long – and be able to modify or delete data when needed.
Data has incredible potential, but so often we see organizations trying to turn lead into gold instead of making pipes. In other words, if data is the oil of tomorrow, we are trying to put it in our engines without refining it first. The result is explosive.
Without the hard work required at all stages of the chain – collection, processing, storage, restitution, distribution – data cannot properly irrigate the decision-making systems that guide a company’s strategy. An executive team that supports a strategic decision based on a questionable indicator without quality control may be worse off than those using good old intuition.
Follow Piano (@piano_io) and AdExchanger (@adexchanger) on Twitter.