Opinions expressed by Entrepreneur contributors are their own.
“Data is the new oil” is perhaps the most overused expression in business over the past 15 years — although, it does have some merit as an analogy. Data itself, much like unrefined oil, doesn’t have any underlying value or utility. But when refined through analytics, machine learning and artificial intelligence data have the potential to transform businesses and ultimately the global economy.
Just like oil, however, data has the potential to pollute entire ecosystems via biased modeling, lack of regulatory oversight and the operational overhead required to turn the data into something useful.
All things considered, data can and should be a net positive for organizations that create a strategic plan for how to create value from the data that they are generating natively, as well as data they can access through data collaboration and commerce tools.
Data has a nearly infinite number of use cases that can vary wildly from organization to organization, but there are some core principles that can enable you to enhance your business and create value through data.
Collection and storage
To stick with the oil analogy, data is only valuable if you can get it out of the ground and store it somewhere. Almost every modern business generates untold amounts of data, but often the data itself is ephemeral, meaning that it isn’t seen as important enough to store anywhere. The thought process typically sounds like “we don’t have an immediate need for this data, so let’s not spend the money it costs to save it anywhere.” This logic turns out to be wrong for two reasons:
- Storage is cheap. Using Amazon S3, you can store a gigabyte of data for around two cents. Those costs can be driven even lower if you’re using a less flexible storage tier. For most businesses, the aggregate cost of “storing everything” isn’t going to be a meaningful component of their operating expenditures.
- You can’t go back in time. Even if there isn’t an obvious use case for data today, that doesn’t mean there won’t be one tomorrow. Additionally, the value of data is often driven by a longitudinal analysis of the data, meaning that if you wake up one morning wishing you had saved the data you may have to wait months until you have collected enough to make it useful.
Even the smartest product managers, engineers, and analysts can not predict the future and so businesses should be focused on retaining as much optionality as possible by storing every possible bit that they are generating.
Acquisition and enrichment
First-party data, meaning data that is generated directly by an organization, is always held up as the gold standard for data. And while the provenance and quality of the data have stronger guarantees than non-directly collected data, it often isn’t sufficient in building a data-driven organization.
An interesting anecdote that proves this point can be seen in the form of large technology companies (Facebook, Google, etc.) open sourcing many of the artificial intelligence models they have created over the last decade. This shows that these organizations think that their strength and competitive advantage doesn’t come from the models alone, but rather from the data they feed into those models. Your average company that doesn’t have data assets at the scale of a FAANG company can’t hope to squeeze as much value out of those models.
To combat that, organizations need to look at strategies to acquire net new data and enrich their first-party data assets to build a stockpile of data that can be used downstream to help fuel the business.
Top-down organizational alignment
Data teams have been widely dispersed in many organizations. Each business unit, division or functional area might have its own data group. One working for marketing, another in finance and another in supply-chain management.
This approach often leads to the data also being siloed, overlapping data mandates, and a general lack of best practices across the entire organization. In the last five years, we’ve started to see chief data officers being appointed within organizations to help solve this challenge. Just like a chief human resources officer makes sure that recruiting, hiring and culture practices are not disjointed across an organization, a chief data officer can play a similar role while also making sure the company is adhering to its data governance and security mandates.
In the end, it is fair to say that “data” is not a strategy. Data needs to be seen as a resource that, when collected, organized and enhanced as part of a broader strategy, can transform businesses large and small.