Ground Rules for Data Usage: Governance & Data Usefulness Series Part 1
How to balance company agility with data quality & governance
There is a constant push and pull in big organizations. You generate a ton of data, and you want to be able to use that to gain a competitive advantage. Yet, you also are beholden to boards of directors, shareholders, and the law.
Naughty things happen at big companies; we won’t lie and say they don’t. Everyone has seen a slide deck that caused them to narrow their eyes as the numbers looked too good to be true. Everyone has seen flat-out wrong data shown at a meeting where it should have been fact-checked before.
Unsurprisingly, while we have more data than ever, we still don’t trust what we see. There are so many issues that pile up and contribute to the problem:
Lack of data literacy
Resistance from marketing and sales leaders towards more advanced measurement methods (sometimes we show them things they don’t want to see)
Non-analysts creating data storytelling decks and misrepresenting the data
Analysts creating storytelling decks and not explaining the context
Data quality issues in one area that reduce trust in all others
Lack of transparency into data sources and quality
The list could go on and on. It’s not surprising that, as a result, many companies respond by:
Restricting access to data so that they cannot use any of it to make business decisions, effectively neutering any analysis teams and projects
Not prioritizing data collection and not using it in decision making
Reducing data sharing to highly restricted reports that get shared with the board of directors and few others - those reports have very little room for interpretation or deep-dive analysis.
These options sound less than appealing when put on paper (or a screen in this case); however, you’d be surprised at how many orgs have at least one of those options in play because of a history and fear around data that led to reduced data sharing or other issues. Many out there can seem like a barrier to getting impactful insights out of data.
The good news is that there are a lot of strategies to fix this, and many other companies are using their data in pragmatic ways that make major growth impacts for their business and provide a competitive advantage - think Google, Walmart, or Amazon. Do you think any of those companies let the list above stop them from finding ways to get an edge with data? (We won’t get into the morality of some of the choices those companies have made in their pursuit - I promise that you can still be an ethical company and find your edge!)
Data Still Requires Experts
Sometimes the term data democratization gets grossly misinterpreted. It’s not about turning every employee into an analyst (I promise you, the idea gives data leaders nightmares and cold sweats), although some companies interpret it that way.
Data democratization is about giving everyone in a company a working knowledge of data - from how to interpret visualizations to informing their decisions from data. It’s also about ensuring that your analyst teams are empowered and funded enough to get digestible data to all teams and have access to everything they need to do their jobs.
Few things are more dangerous to a company than an ambitious marketer with access to data and over-inflated confidence about their ability to interpret data. There is a reason why analyst teams should act as checks and balances with decisions.
Set Some Ground Rules for Data Usage
One result from above is a lack of trust or control over data - we have people who shouldn’t be reporting on numbers doing so and others who should be, not having the access they need.
This is where data governance comes into play in a different way than you may originally think about the term - yes, we also need to be looking at our overall compliance with data privacy laws. Still, another incredibly important function of data governance teams is maintaining knowledge management and processes that avoid erosion of trust around data in companies.
(If you don’t have a data governance team, the first step to this process is finding a good place in the org for this team and putting money towards hiring them - it’s a critical part of modern companies that can’t be overlooked.)
(If you feel like you need help figuring out where this team should live in your org, how much it will cost, and other logistics, many firms like us can help you)
The Core Things You’ll Need
There are a few tools that you will need to be able to set ground rules by department or area of responsibility. If you have some and are missing others, it’s time to start rounding out your stack of tools in partnership with your governance, IT, finance, and data teams.
A company-wide data dictionary
A data dictionary is a shared, easy-to-access document that includes all of the metrics that are core to your business. It should include what data source they should come from, their full definitions, and who owns the data source. This can be a hefty but worthwhile undertaking to collect all of these metrics in one place for the following reasons:
Know exactly who is responsible for data-source quality if anything ever goes wrong with a metric
More easily onboard new team members and risk losing less institutional knowledge when team members leave
Reduce risks of metrics being reported to the wrong people from the wrong sources
Stop disagreements and turf wars over data sources by defining and agreeing as a group
2. A company-wide data strategy
Using data will never be easy and effective if your company doesn’t align on an overall data strategy first. A data strategy helps you identify the gaps within your skill sets, helps you identify how you will collect and use data as a company, and defines the projects and budgets needed to execute those goals.
3. Documented review processes for different levels of reporting
Many companies (even incredibly large ones) are still missing this critical step. Having processes that review the quality of data presented in decks & reports ensures that information is correct and is in the best interest of the company as a whole (not the individual making the report only).
Below we’ll work through some ground rules by function for the documented reporting processes and the overall ground rules on how data should be shared/distributed at those levels.
4. (Recommended) Data discrepancy one-sheets
This isn’t as much of a standard, but something that we highly recommend at Insight Lime. A data discrepancy one-sheet is a document that shows the differences between data sources at a company - for example, the difference between finance revenue numbers and Google Analytics revenue numbers.
The one-sheet should be updated regularly to check for any major changes in the difference between sources. There will always be discrepancies between your data sources, and having an easy-to-understand reference sheet for your company will help increase trust in data.
Conclusion
Hopefully, this list gives you ideas of where you may have gaps in your organization's current governance (and data usage). We will continue this series in later months, and cover topics like the level of governance recommended for each level of the organization and how to maintain agility while keeping proper control over data sources.
(If you’re interested in learning more about Insight Lime’s data strategy packages or just have a question that you want answered for free, contact us)
Let’s get squeezing!
Join our founder on Linkedin and hear thought-leadership on marketing analytics and analytics consulting.
“[Insight Lime] are great at identifying the issues and providing a solution.”
Search for more blogs and content you’ll love
Need more help calculating your marketing budgets? We’re here for you 💚
Let’s get squeezing!
Join our founder on Linkedin and hear thought-leadership on marketing analytics and analytics consulting.
