Eight Questions on Data Governance: An Interview with Alation Co-Founder and CDAO Aaron Kalb

Today, Alation announced its new approach to data governance, Active Data Governance, and with it, a series of initiatives to embrace data governance as a strategic use-case, all of which are part of Alation’s vision to transform the data catalog into a platform for a broad range of data intelligence solutions. I sat down with Aaron Kalb, co-founder and CDO of Alation, to discuss today’s announcement, the journey to active governance, and what the future holds for the platform.

Nolan Necoechea, Alation:Aaron, thanks for taking the time to connect with me today. What’s your take on today’s announcement?

Aaron Kalb, co-founder and CDO, Alation: I’m really excited about today’s announcement in particular because it demonstrates one of our core values: to measure success through customer impact.

We have been building data governance features into the platform for years.  What’s new is that today we have a lot of exciting data governance stories from our customers and that’s what I’m proud to share.

Nolan:Was data governance something you were thinking about when the company was founded?

Aaron: Early in Alation’s development, even before we started to call our product a “data catalog,” we were focused on analytics organizations and helping people in them use the right data in the right way. Part of that included collaborative curation, ensuring that data was up to date, and making it easy to find accurate definitions — things that many would now consider data governance features. The reality is that no one who’s serious about analytics can ignore things like the provenance or the quality of the data.

At the same time, data governance was becoming a hot topic, and compliance with regulations — like BCBS 239 in the financial services industry and GDPR in the EU — was driving the data governance conversation. Because data governance centered primarily around regulation back then, data governance products were more about restricting what data people could access to mitigate risk. While we were developing features that could be used for data governance, the prevailing definition didn’t fit with our vision of empowering data consumers to do more with data. So, even as we kept building  what people today would call data governance features, we avoided the term “data governance.” Instead, we used terms like “data curation” and “analytics stewardship,” concepts that better aligned with our vision for Alation and our early customers’ philosophies and goals.

Nolan:What changed your and Satyen’s minds on data governance?

Aaron:We observed two phenomena. First, we continued to build data governance features — without necessarily calling them that — because our customers kept asking for features that could serve dual purposes. For example, TrustCheck could tell you that a data asset is out of date and suggest the up-to-date asset to use instead, which is an example that supports analytics. Or, TrustCheck could tell you that a field contains PII and you shouldn’t look at it on a per-row level, which enforces a data privacy policy. We built a lot of features that could support both analytics and data governance into the platform. Without us explicitly saying that we offer data governance, our customers started to use us for data governance, including regulatory compliance.

In the meantime, industry analysts, thought leaders, and data experts, began broadening the definition of “data governance.” The meaning expanded from just “documenting policies” to also include ensuring data is used the right way.

These two phenomena converged to create this stepwise increase in the number of customers who report using Alation for governance: some using Alation for more traditional “defense”-focused data governance as we released more supporting features; some using us for this new notion of data governance for “offense”; and, of course, many now using us for both.

At that point, our thinking on data governance changed. In customer conversations, we heard over and over that data consumers want to do the right thing. No one wants to violate a privacy policy or bring down the data warehouse because they ran the wrong query — but they do want to get their job done.

While we thought that our traditional champions resided only in the analytics organization, we came to realize that many IT and data governance teams were becoming champions, too. It turns out they actually don’t like saying “no” all the time, and get more excited about the chance to empower their organizations and be a part of business success. The head of data governance at one of our customers put it clearly, “we want to be road builders, not road blockers.”

So, there was a philosophical alignment that emerged. We see ourselves as serving the entire enterprise, and building a data culture that spans the whole organization. To drive that data culture requires not only data search and discovery and data literacy, but also data governance.   This is something we are super excited about. It’s a three-legged platform for driving data culture.

From there, we began to build features that were not only dual-purpose but also purpose-built for data governance teams. Our design philosophy is to go beyond just giving someone a powerful tool and letting them figure it out. We do deep user research and provide an experience that is tailored to their persona. So we did that, and came up with interfaces to leverage the intelligence and machine learning technology that has been within Alation for seven years and make it more accessible for data governance teams, through features like our analytics stewardship dashboard.

The final nudge for this announcement came directly from our customers. Customers kept coming to us and saying that not only are their data governance teams using Alation, but that Alation was better for data governance than the other solutions on the market. Why? Partly because of our unique and powerful automation and behavioral intelligence capabilities. But also because Alation supports both analytics and data governance — and those two elements create a virtuous cycle when aligned: data stewards help analysts through curation; analysts help stewards by “telling” them (via usage and the behavioral intelligence) what data is worth curating.

Hearing that feedback from our customers and seeing that about a third of them were leveraging Alation for data governance, we were convinced that it was time to double down on this space, part of which is telling the world our data governance story.

Nolan:We call Alation’s approach “active data governance.” Can you explain how that differs from what people usually think about when they think of data governance?

Aaron:Historically, data governance work has moved slowly. People manually and methodically go through asset after asset and definition after definition. The emphasis is on making sure the rules are written down — but assumes that people are going to look up these rules and definitions as they work, and we find empirically that they just don’t.

Unfortunately, a lot of organizations end up failing with this passive data governance approach, because it introduces a separation between data governance and analytics. The data governance teams can do a lot of work documenting all of the policies and finalizing all of the definitions, but they often don’t have a way to know whether they have been effective. Their job ends before analytics begins.

Active data governance, on the other hand, brings governance and analytics together by using observations of prior activity to inform policy, and then ensuring that policies impact action.

First, active data governance looks at what is already happening to guide data governance efforts. Alation enables data governance teams to prioritize the data that is being used most, measure how well that data is being curated, and assign ownership of that data to the people who are expert at using it. Most important, Alation goes the last mile by surfacing recommendations for how to use the data at the point of access — i.e., right as the data consumers are working with it. It’s like the difference between dropping someone in the middle of nowhere and handing them an atlas or providing them with a GPS and turn-by-turn directions. With active data governance, enterprises are much more likely to see accurate and compliant analytics. In short, another way to see active data governance is just as data governance that works.

Nolan:How does the data catalog support this approach?

Aaron:When we invented the data catalog we were thinking broadly about how to bring machine learning and intelligence to metadata. As discussed above, in the early days we gravitated towards analytics use cases, like data search and discovery or guided SQL composition. But we always envisioned that we were building something bigger. So while customers first used the software we built for analytics, the software itself was built to extend to a broad range of use cases in metadata management, or to use the more modern term coined by IDC, “data intelligence.” That’s why it was fairly easy for us to organically extend into data governance in response to customer demand.

In short, we’ve never seen the data catalog as a standalone analytic application. We see the data catalog as a platform to support a broad range of solutions including analytics, data governance and stewardship, data privacy, cloud data migration, data democratization, digital transformation, data ops, data forensics, and so on. That’s why I dig our new Marshall McLuhan-esque tagline, “The Catalog is the Platform™.” It pithily captures what we believe: that the data catalog is the ideal platform for all these applications, since they all benefit from intelligent metadata, applied at the point-of-use.

Nolan:Thanks for sharing that vision. Let’s switch gears back to customers. Can you give an example of how customers use Alation for  data governance?

Aaron:Sure, I’d love to. I had a wonderful chance to sit down in Chicago with Brad Burke, Chief Data Officer at American Family Insurance. As an insurance company, their job is to constantly think about and mitigate risk. At the same time, AmFam is doing some really cutting-edge data science and machine learning. And that makes American Family Insurance a great example of active data governance because their data governance programs drive both risk mitigation and innovation.

The way AmFam leverages Alation both for defensive governance and to further data literacy efforts and create a competitive advantage really represents the strength of our active data governance approach.

Nolan:What does the future look like for data governance on the Alation platform?

Aaron:We want to collaborate with both our customers and our growing ecosystem of partners to prioritize the next data governance features to include, and the next partner integrations to build,  in our product roadmap. We haven’t spoken too much about them thus far in this interview, but partners are key to our strategy at Alation. We neither want to nor are capable of doing everything, so we want to work with leaders in their areas to deliver solutions for customers. , For example, we’ve recently announced strategic partnerships with BigID, a leader in data privacy, and Databricks, a leader in data analytics and artificial intelligence. The goal is to make all of the intelligence within the catalog easier to use and apply to a wider set of data governance use cases.

Nolan: Do you have a closing thought about today’s announcement?

Aaron: It’s been delightful to see our success in data governance emerge through our close collaboration with our customers. Now that we have made a commitment to data governance, the opportunity we have is to accelerate the frequency and the amplitude of customer outcomes. I’m excited to see even greater customer success. And, as I said at the start, one of our core values is measuring success through customer impact.