Is your data governance strategy ready for syndicated data?
Large retail and consumer product goods (CPG) companies have been buying aggregated market data—known as syndicated data—for decades and using it to glean consumer signals for forecasting and planning. This has traditionally been an expensive practice, but advanced technologies have made it more affordable for middle market companies in a variety of industries to access data analytics solutions and use syndicated data.
Adding syndicated data can broaden the insights machine learning models deliver, expanding the value of existing data, particularly for companies already using analytics. However, using syndicated data comes with some risks, so reviewing your data governance strategy before getting started is a valuable step.
Staying aligned to emerging analytics capabilities
Using syndicated data doesn’t change how you perform analysis, but it can improve the quality of your analysis, either by adding context or leading to the discovery of new knowledge. Simply put, the more verifiable signals you have when forecasting and planning, the better those forecasts and plans will be.
Adjusting your data governance strategy as analytics strategies change helps maintain data accuracy and shareability, which are essential to deriving effective insights from both internal and third-party data. A data governance strategy establishes controls for vetting, aligning, cataloging and maintaining data assets across enterprise systems so that everyone is using the same, singular source of truth. These control points occur where systems ingest the data, when the data is moved to other systems, and where and how the data is stored for use by analytics applications.
Forbes estimates that the amount of data created, captured, copied and consumed in the world grew by 5,000% from 2010 to 2020. Undoubtedly, as more data is collected by all types of organizations, it will become available as aggregated sets of data through syndication. This will present a great opportunity for your company to evolve your analytics strategy, but it also will require making sure that data governance includes controls for syndicated data risks. This could include:
- Sources: Traditionally a small number of leading market research firms, such as Nielsen and IRI, have dominated the syndicated data market, but that’s changing as other businesses and organizations make available the aggregated data they own either through paid subscriptions or open source resources. For example, some commercial food providers are selling syndicated data they’ve collected on food nutrition through online apps, and government agencies, such as NASA and the National Weather Service, freely make data available on their websites.
Syndicated data isn’t regulated, and not every provider is using the same methods. Conducting an assessment of providers and their own data governance strategies before buying or using the data can help you better understand what risks, if any, need to be considered and addressed.
- Quality: A lot of data is compiled automatically through sensors or software, but that doesn’t mean data sets pulled from these streams are always accurate and complete. Sensors, devices, software and networks can malfunction and lead to inaccurate or missing data. These blips could be considered normal outliers, but they affect data quality, nonetheless.
To maintain quality, it’s important to have controls in place to vet all syndicated data for quality issues, especially if the data is coming from a new provider.
- Users: Like shadow IT, shadow data can bring risks to organizations. Having standards for when and how syndicated data can be used will enable chief data officers, chief information security officers and other technology leaders to have visibility into all of the syndicated data that your company is ingesting and using.
And just like when your business is buying technology, using syndicated data works best when there’s a purposeful strategy that aligns with organizational goals. Part of your strategy should be the prioritization of the best opportunities for improving data analytics—this might be where your biggest pain points are. For example, if marketing promotions aren’t working as expected, syndicated data can be used to test the existing model that provides planning signals, or it can be used to identify new and better signals.
Getting help with syndicated data and governance
Syndicated data offers a lot of opportunity for more meaningful data analytics, but making sure that the data is coming from reliable providers, is of high quality and aligns to your strategic needs can be difficult for a midmarket company to do on its own.
