Talks Tech #37: Building an Effective Data Strategy
In almost all the companies that I've worked at, data was the product, which means even if customers didn't realize it, our product was essentially data aggregated in a way that was interesting and or it was useful to our customers in some way and that is inherently what they came to use the product for. What is data strategy? A data strategy is essentially a plan on how your company will use, maintain and store data. It is the data management plan that your company can commit to so that data can be utilized effectively to power good business outcomes. A data strategy has a few key components. It has the business case, the technology, the people, and the processes. The first component of the data strategy is the business case for leveraging data. Clarify why your business benefits from data insights and how you can improve your product and customer experience through data. This is very important as it sets almost the why behind the investment that you're making into data topics.
The next component is the technologies that you will or plan to use. This can include tools that you'll use for storing, mining, analyzing and visualizing the data. There are actually many different options for the technologies that you can possibly use. There's a lot to process on this front, coming up with the right technologies that don't just suit your organization's unique needs, but a plan for that effective strategy is crucial. The data strategy should include a plan for the people. What functions do you hope will be part of your core data team? How big does your data team need to be? There are a lot of inconsistencies in the industry around terminologies and what each function is supposed to represent. It's good to clarify that. Data scientist, data engineer, data analyst, machine learning engineer, analytics engineer BI or visualization engineer, there's a lot of different terms out there. They can all do slightly different things.
Who you need depends on how you define the work and what you feel needs to be included in the data strategy. For the people aspect, it's important to define roles and responsibilities even outside the data organization. This helps ensure that there is clarity around those crucial cross-functional partnerships that can either amplify or detract from the data strategy. One of the most crucial pieces of the data strategy is the actual process or the structure that you will put in place to manage and use your data assets. How will you maintain your data ecosystem? Who is actually going to maintain it? What is the intentional governance strategy? What is the cadence of audits that you will run to build and grow your company's data-informed decision-making, having some data processes that you can at least start with but build on and iterate over can help you in evolving your overall data strategy.
Those are the four pieces that form the core of your data strategy, the business case, the technology, the people and the processes. These are the pieces that help you effectively leverage your data in support of your business. Why do we need a data strategy in the first place? Ultimately, companies are defined by their ability to quickly and efficiently execute on their product vision. The product vision in order to be successful has to provide real tangible value to their customers. When companies leverage data in a timely, relevant and efficient manner, they're able to build better roadmaps and that brings them closer to business success. If product decisions were made purely based on gut or intuition, success will also be luck-based and that is simply not sustainable. This is one of the biggest reasons for investing in a data strategy.
Faster and higher quality decision-making within companies can be that difference in achieving business and revenue goals or getting left behind. A data strategy also ensures operational efficiency. After all, many companies are now data companies. Think of some of the largest consumer tech companies. Pretty much all of them have huge data components, Google, Meta, Amazon, Netflix, they're all sitting on massive amounts of data and that product is essentially how you interact with that data or maybe how you generate more data that in turn provides you with something useful. If these companies did not invest in aiming for maximum efficiency when it comes to using, maintaining and storing all that data, they would just not be successful. There is also a lot of operational overhead when your data architecture isn't clean or cohesive. There is a very real cost of wasted money on things like compute and storage when you're dealing with big data that isn't streamlined.
All of those reasons ensure market competitiveness through effective leverage of data, faster decision-making, ability to hit revenue goals and operational efficiency and reducing wasted money. These are just some of the very high-level reasons for investing in a company-wide data strategy. What makes for an effective data strategy? For a data strategy to be considered effective, it has to nail it on three different aspects, relevancy to the business, cross-functional alignment and flawless planning and execution. A data strategy is created for the business. It has to solve a real business pain point. You don't create a data team for a business that is not building or selling anything. You don't buy data technologies if you have no data and there's no need to create a process for data if your business doesn't really deal with much data.
You have to audit your business and see what it needs. How much time and resources can you afford to spend on data aspects and come up with the best data strategy that meets those boundary conditions? If this business relevance criteria isn't met, nothing else in your data strategy really matters. The next aspect of an effective data strategy is cross-functional alignment. Nothing in the data world can be achieved without it. For data initiatives to be successful, the company at large has to be aligned and it has to be invested. There is almost no part of the data process that can be successful without that cross-functional support. To ensure cross-functional alignment, I would suggest that you start with education. Does your organization understand what is required for successful data initiatives? Everyone loves a pretty dashboard or cool experimentation results that tells you what to build and what not to build.
Focusing on education early and often and also customizing it to the maturity of your specific organization will yield many results. Part of that education is also showcasing that vision on shared goals. Make sure that cross-functionally everybody understands what's in it for them with each data initiative. You want to get more attribution tracking for maybe some marketing campaigns. Proactively explain how marketing budgets can have better ROI through that attribution information. People are way more willing to help if they understand how their own goals can be met through the initiatives that you are focusing on. The third and final aspect of an effective data strategy is great planning and execution. It takes time to audit and figure out what your organization's data maturity is.
There can be many parts of the business that differ in that data maturity. Some parts may be a little ahead and any plan that you create has to incorporate those differences in the different parts of the business. There can also be some inherent differences in the skills and the experiences of the people who are on your team and who you can hire. In addition to that, your tooling and technology choices need to merge your current and growing team skills as well as evolving industry standards and preferences. You have to find that sweet spot, that little balance between everybody having to learn something completely new, maybe that ends up taking away precious time from their data objectives versus picking something that everybody already knows how to use, but maybe that just won't fulfill tomorrow's use case. Another aspect of great planning and execution is the intentional data governance strategy.
Make sure your data assets are accurate and reliable. That is the foundation of good data practices. If your stakeholders don't trust the data that they use to make decisions, it'll have cascading negative repercussions throughout your organization. The final piece of executing and planning flawlessly is to account for the cultural change that needs to happen for you all to succeed. People can hear your logic, they can look at your charts and they can share your vision. Until you give them a blueprint, which clearly outlines how the changes are going to be sequenced, how to shift the culture and not just the people processes and technology, you will inevitably see yourself getting stuck. You have to commit to that cultural mind shift change that needs to happen around the data strategy and the importance of various pieces within it.
Cultural change is hard, but it is doable. Every data practitioner knows that they have to get that organizational, something that cuts across all levels. They have to get that entire organization to speak the same language and to go after the same goal. How do you get buy-in for your data strategy? There are four pieces. One is to pitch your data properly. Two, represent the needs of your product and business. Three is to understand your stakeholders. Four is to know what's actually important. You have to know your data deeply to be able to clearly present a vision on data strategy. You can leverage that understanding to lay out a clear plan that presents the data roadmap. It presents what's possible by when and what specific value it will unlock. Showing that vision for the data roadmap is a key piece for getting buy-in for the larger data strategy. What does your business actually need? What are the highest value problems today that data needs to help you solve? Where exactly will you get the highest ROI from your data investments? What do those add up to? Answering those questions and diving deep to figure this out will go a really long way in your ability to effectively pitch your data strategy. You'll be able to better tie in the tangible value of that data strategy, and what it can create for your business.
You have to understand your business stakeholders and their motivations. This part is critical. Framing the data strategy to specific stakeholders and their motivations and success metrics in mind is a very smart plan. Always keep your data champions close and amplify them and their needs when you can. This is a good way to get buy-in for your own data strategy. The final piece of getting buy-in is being smart and knowing what's important. Once you have a deep understanding of your data, your products and your stakeholders, it becomes much easier to triage the many different things that you can focus on. The aim should be to ruthlessly stack rank every possible thing that you can do from a data perspective. Paint that vision of what should come first. A data roadmap can be effective in which it's clear to others that you're not going to get lost in trying to do everything, but rather you'll be focusing on very realistic goals and you'll be focusing on the goals that bring in the maximum value.
Who can contribute to a data strategy? That's a trick question because anybody can contribute. A good straight data strategy is led by data leaders, but it is supported by all cross-functional partners. Data leaders can drive the decision-making, they can identify that racy ownership of who's consulting responsible for delivering, approving the strategy. Who is responsible for the execution? This hopefully helps create that engagement and that sense of shared responsibility. When you have an engaged group, they can help shape powerful data strategies. Those things then incorporate the nuances of the entire organization and ensure everyone's success. If you're not part of the data team, but have viewpoints on your company's data strategy, make sure to be proactive and reach out to your data partners and see how you can help shape it.