Trust

To have data that is worthy of using for decision making, data must be accurate. Remember that data is just information, so making decisions informed by incorrect information will not result in the desired outcomes. To avoid this, data accuracy must be maintained throughout the data lifecycle. Leaders should ensure that every data source and stage of the data lifecycle is owned by a designated team member. If there is a breakdown during any phase of this lifecycle, the owners will lead the effort to correct that breakdown as quickly as possible. The table below outlines some examples of the data lifecycle stages and ways it can break down.

Even if data is accurate, it may not be trusted. I have seen trust in data to be very conditional. Sometimes, data is accepted as the absolute truth and other times, its accuracy is questioned. Oftentimes, this is because folks are looking for data that aligns with their existing beliefs so its trustworthiness depends on what takeaway or position it supports. In these situations, data is being used as a political tool rather than a decision-making one. If this scenario feels familiar in your current organization, data is not trusted. It is a sign that opinions are set before evidence is provided and that members of the team are not using data in discussions. 

In order to build trust in data, practice making sure your opinions are malleable to new evidence and data. Hold off on making decisions until you feel as though you have the appropriate level of information to do so, and be transparent about what data you’re using. If you are a leader, ask your employees for the data that they used to make their decisions. 

Alignment

You knew this was coming. It’s time to talk about metrics. Which ones to use, how many, how often do they change? These are all excellent questions and plenty has been said on this topic. The Data Culture Framework is ultimately agnostic of the metric philosophy, whether that is KPIs, MBOs, OKRs, 4DX, etc. that you would like to use. What is important within the Data Culture Framework is that (1) you have goal metrics, and (2) they are regularly shared with the entire organization. 

Goal metrics have targets and are not just a measure for measurement’s sake. They should tie back to the vision for your organization. Departmental and individual goals should be derived from these goal metrics. Regularly share them in meetings and provide access to dashboards with updated metrics so individuals and teams can reference them whenever they want to. 

So why is this principle called alignment and not metrics? There will be metrics all over your Data Culture. The purpose of goal metrics is to ensure that everyone across the organization knows what they are working towards, aka they are aligned in what they’re doing. An organization is much more powerful when everyone is working towards that same goal with the same success criteria. An organization where each team has their own initiatives that contradict or undermine other teams’ means the organization has failed to create focus around the same goals.

Storytelling

People respond to stories, so make sure you can articulate the story that your data is telling. Building off of Alignment, what really helps your organization is being able to connect changes in goal metrics to what people are experiencing and the story that the data is telling. If numbers are up, why? If they are down, why? If revenue is up but profit is down, why? People want to understand and you want them to understand. 

As you ramp up your Data Culture, storytelling is an especially helpful tool to connect the importance of data for your employees that are less comfortable with data. It gives them a way to understand how what they are doing day to day connects with the set goals for the organization. 

For example, let’s say you run a customer service organization, and a goal metric is to increase customer satisfaction from 77% to 85%. When reviewing your goal metric at an organizational meeting, you see that it dropped from its steady upward trajectory in the last period. You present this to the team as simply,  “here are the results from this last period,” and then you move on to the next agenda item. This is not satisfactory because your people will want to know the “why”; what caused the drop? You may already have an idea of the driver of this, so share that with the entire team. If you don’t know, ask the larger team, especially the customer service representatives. Maybe a new version of the application launched that is confusing people because CSRs were not properly trained on this new version. They can’t satisfactorily support the customers. Your team is now connecting their experiences to the data. 

The same should be done when numbers are good. If things are going well, is it because of the efforts of the team, new projects, or initiatives that have been implemented? Share that and emphasize the wins for the team so they know what works. 

Governance

Data governance is the last category to consider at the organizational level and probably the most obvious. Governance is the set of rules that guide access to the data, how it is to be used, and the security of data. As more of your organization uses data, governance ensures that everyone is clear about what that data represents.

With data access, you want to make sure that data is being provided to folks at their level of comprehension and their level of need. A finance analyst that knows a query programming language like SQL should have a different level of access than a sales development representative that uses spreadsheets only to track their calls. Both should have access to data, but it should be made appropriately accessible for their level of understanding. 

For how it is used, ensure you have clear standards around metrics and calculations. Many organizations track user engagement or conversion to leads but that is going to be different depending on what type of organization you are. An engaged user on a social media app is going to be very different from an engaged user on a banking app. Make sure it is very clear what is being measured and that definitions are readily available. 

Finally, data security ensures that privacy is strictly maintained. This is important to protect your users, employees, and your organization. Besides how data is accessed, what data is available is also a question you should have answered and implemented protocols around. If your organization must track user social security numbers, that is probably not data that should be accessible to everyone in the organization. 

Remember that as organizations mature, data governance should also mature so continue to revisit the data governance protocols regularly. 

Finishing Up

Tackling each of the principles I’ve outlined will require practice and patience so remember that incremental progress is your friend. In my next blog, I'll review the principles to develop in your departments and teams to build a strong Data Culture. 

By Cindy Lin

Edited by Jason Rubinstein