Interview

First of all, what is a data strategy in your opinion?

Shawn Livermore:

Ok so a data strategy basically outlines an organization's plans and goals for managing, utilizing, and safeguarding their data. It includes several key pieces that I can explain a bit. The most fundamental is data governance. Governance defines the roles and responsibilities for data management, as well as policies and procedures for collecting, storing, and using that data. I would say the next part of it would be data architecture, which includes all the technical infrastructure, systems, and tools needed to manage and access data. Then somewhere in this strategy you’re going to want to include a section for analytics which describes how data will be analyzed, visualized, and used to support decision-making and drive business outcomes.

Security of data is another pillar, that is - how data will be protected against nefarious access, theft, etc, and of course how the company will comply with privacy regulations. And I think there would need to be something in there that needs to address the notion of creating a data-driven culture, where data is valued and used to inform decision-making at all levels of the organization. A lot of companies try to go to the data too late. They wait until they are in a bad spot to suddenly declare themselves data-driven, and by that time it’s probably too late. So better to stack things up strategically at the outset. 

What role does strategy play in defining the data footprint of an organization?

Shawn Livermore:

Yes I would say the strategy is absolutely central. It’s 100% necessary in defining the data footprint of a company from the start. The strategy of the data is its lens - how it sees pretty much everything - and the company follows that from the beginning. So your leadership team or founders will depend on that strategy to help determine the types of data that the organization will collect, store, use, promote, share, test with, etc.. And around that strategy you have the goals or objectives of the leadership team, and that will also influence the types of data that are considered valuable and the methods used to acquire and analyze that data and integrate it as well. This are of data strategy is sort of its own separate subgenre or its own separate science. It’s what drives a classic decision we see a lot which is when the leadership team decides to take that step to fund the new data science team… this emphasis on the data strategy. So also - this is going to cascade from there and even influence how you move forward and with which types of technology and infrastructure you select to build and support your apps. So it’s paramount. And yet, you know, even if the data strategy is not defined upfront, the company usually ends up backing-into one and then promoting it. It would be simple usually if that’s the case, and be primarily based on the data they already have and their existing business model and practices. 

How do organizations rework or redefine with existing data strategies that just haven’t worked?

Shawn Livermore:

I think it’s like anything else where you just iterate on it. We’ve noticed when it works and when it doesn't work. It comes with time and focus and a willing set of executives and engineers and even board of directors. They just pound through it in the meetings till it meets the objectives and aligns with the vision.

What does it mean to be deeply vested in the data strategy from an executive perspective?

Shawn Livermore:

I would say being vested means being committed, and committed to what? To the success of the organization's entire suite of data initiatives. This involves a few pieces though, and it’s not all at once - it’s over time. But starting from the outside of the circle and working our way in, it involves having a really formidable understanding the strategic importance of data in the company, and how that refined data fits into the overall business strategy. Many leaders just don’t consciously see it. They speak of it and dance around it, but they don’t see the data as being the key thing. It also means properly allocating adequate resources, including budget and personnel, to support the data strategy. I know one client of ours, they had a data team, invested millions of dollars annually in the data team. But when they went to seek a really foundational piece of software which required paid licenses for their team of - i think 6 or 7 - database engineers, they were told it was not in the budget. So here you have a dedicated, focused, and incentivized team of really amazing database engineers who are asking for the tools they need and being flat-out denied, based on the license costs. And the licenses were not that much, maybe 5-10 grand a year. So it really made no sense to us but we see this happen a lot more than it should. This is really how it goes, unfortunately. So many companies still view data strategy and, transitively, data science as a cost center rather than what we’d call a value driver. They allocate bare minimal resources to their data science teams, and yet, in the same breath, expect them to perform miracles with limited budgets and ever tighter timelines. This approach is absolutely a short-sighted one and it breaks down the trust between the execute team and the data teams. So this is a pitfall - some logical fallacies in there perhaps. Penny wise, pound-foolish, something like that… 

Back to the question - being deeply vested in the data strategy. I think having and stating and reminding the teams of the goals is so critical. Companies need to have a clear vision of what they want to achieve with their data and provide the necessary support to help the data science team reach those goals. This means when the annual planning takes place, setting realistic timelines, providing adequate resources, and then stepping back to ask the right questions in the room where the boots on the ground - the engineers and the support team - so that they can provide their honest answers without any hesitation or fear.  

What about organizations having a chief data officer? What have you seen in that role and how it plays out? 

Shawn Livermore:

So as an enterprise software consultant, I have seen the role of the Chief Data Officer, or CDO, play out in various organizations, both large and small. Surprisingly, some smaller organizations also have these roles, albeit, with more limited budgets of course. So that plays into the overall reach of the department and the role. But from my understanding of the role, and of course, it varies from client to client, but we see the CDO as being responsible for establishing and implementing a company's overall data strategy, and ensuring that data strategy is carried out - effectively and ethically - across the entire organization. The CDO really just becomes the glue or the bridge between the technical and business aspects of the organization, working to align the data and the data strategy to the company's overall goals. 

In organizations where the CDO role is properly implemented, I have seen the following benefits of better alignment of data initiatives with the overall business strategy. I have seen first-hand the improved data quality and better data governance through new policies and procedures. 

We know there’s going to be a higher efficiency and and very real and very measurable ROI from data work being done. 

On the other hand, in organizations where the CDO role is ignored or scoffed at, we’ve seen some challenges. We see a database team that is scattered and disorganized. They are sort of at the whims of the development teams, which can go haywire quickly. It’s either the developers demanding things and making data and architectural decisions, or it’s project managers, some of whom are not technical or not technical enough, who are calling the shots. But either way, things are not being done from the lens of a true data strategist. And that’s the key - the data strategist - normally the senior-most database experienced engineer or executive in the company - that person needs to step forward to the whiteboard and create the plan, drive the plan, and police the data. 

What are some of the common areas of discrepancy or uncertainty that chief data officers or senior database executive IT management grapple with?

Shawn Livermore:

Ok so there are quite a few, but mainly the ones that i can speak to would be - well let’s start at data quality. Because the quality of the organizations data is one of the many measurements of the health of the organization’s technology. And so it’s quite complex actually because ensuring that your data is accurate, complete, and consistent can really be a challenge, particularly in organizations with large and complex and encrypted data sets. And speaking of encryption, another area they fight with would be the overall data privacy and data security bucket. One of their mandates is usually very closely connected to security, so the CDO and CISO - the chief information security officer - are usually connected at the hip for a while. Ok so another one would be governance. They are the institution responsible for driving policy, and their voice should speak the loudest on that topic. Giving users and developers access to the right databases with the right level of access, and then auditing that at certain intervals, and then working on disaster recovery and backup recovery… all of this is key. It can be a lot to bring together so sometimes a team is really built-out under this role quite quickly after the role is defined because it can be a bit overwhelming. 

Ok last one i think would have to be the word alignment. This is really important - that the top-most database executive has to sit down and coalesce the data strategy with the business strategy. And this can be hard, particularly in organizations where the business and technical aspects of the organization are not normally well or easily integrated. We see this all the time. The ops folks are speaking a completely different language than the tech teams, and that might not change for a while.