You Are in the Best Position to Kickstart Your Company’s Data Strategy | by Michael Fichter | Apr, 2023

Don’t complain about data, but learn how to draft a data strategy with five aspects you need to understand and five steps to get started.

Most companies could be better with data.

A lot think they are data-driven. But if you are working with data, you might be more critical about it. And you have much higher expectations.

Not every company is a startup and can start from scratch with data. Not every company has built its core business model on data.

Still, chances are high that you work for one of the SMEs or large corporations that are successful but do not make the most out of data.

If you believe your company’s data potential is greater, you must initiate a data strategy discussion. Here’s how it’s done:

• You must understand five data strategy aspects.
• You must work on five data strategy definition steps.

#1 Data Strategy needs to be driven from the top

Becoming a data-driven company doesn’t scale in a data engineering or data science team.

Data is simply too pervasive and contains too many factors that must be driven from C-level. On a data team level, you most likely do not have influence over things like:

• Commit to investing resources into data-driven business models and data monetization.
• Incentivize salespeople to track their insights in a structured way in the CRM system.
• Encourage departments to share their data with other teams to create more insights.
• Build and maintain data products in every function or business unit you can use as a data scientist.
• Introduce and maintain data catalogs that help non-domain experts to understand, use, and access data easily.

Imagine measures like the above are in place; wouldn’t data science be much more fun? And wouldn’t the benefit for the company be outstanding?

You need commitment from the top level to pull all the right levers in your company to make data an asset.

Easier said than done.

#2 Data Strategy is not a technology strategy

As the previous examples have shown, data strategy is not only about technology.

Data Strategy is about creating benefits for the business through data by establishing processes, enabling people, and (in the final step) introducing technology.

As with every strategy, data strategy is about defining a (long-term) goal and a plan to get there.

The plan to get to the data strategy can consist of elements from the following four areas:

• Business: data-driven offerings, data monetization, improved decision-making, etc.
• People: data ownership, data literacy, data competencies, etc.
• Processes: data governance, cybersecurity, legal & compliance, etc.
• Technology: data platforms, data catalogs, data analytics & automation tools, etc.

However, each organization’s data strategy is unique. Data strategy also depends on your company’s status quo. As every organization has limited resources, you need to prioritize the most important ones to reach the goals you defined.

#3 Data Strategy is required to remove bottlenecks that prevent you from creating value from data

First, a data strategy helps clarify the role of data in a company’s business model. What are the key data assets and how can data help to generate more value for customers?

Data usually live in silos because systems are developed according to organizational structures, as Matthew Skelton describes in Team Topologies.

“Organizations which design systems… are constrained to produce designs which are copies of the communication structures of these organizations.”

— Conway’s Law

Data strategy needs to be driven from the top to break these silos.

Creating and maintaining data assets requires clear responsibilities for data and requirements for data quality. Staffing, assigning, and enforcing these roles and responsibilities is another crucial role of a data strategy.

Finally, data strategy provides the basis for every data scientist and machine learning engineer to work effectively with data. It helps to attract talent that can spend time creating insights and contributing business value instead of wasting time acquiring and cleaning data.

#4 Data Strategy implementation is a change endeavor

Implementing a data strategy will ultimately result in a transformation of the organization. And implementing change is a process, not an event.

Different frameworks help to understand and implement change. One of the most popular ones is John P. Kotter’s change model. It contains eight stages where each stage builds on the previous one:

1. Establish a sense of urgency
2. Form a powerful guiding coalition
3. Create a vision
4. Communicate the vision
5. Empower others to act on the vision
6. Create short-term wins
7. Consolidate improvements
8. Institutionalize new approaches

To successfully implement a transformation, you have to advance through each stage individually. If you miss one step, your change efforts will most likely fail.

#5 Kickstarting a data strategy is about raising urgency and bringing supporters on board

When you do not have a data strategy in place, you must do most work on steps 1 (Establish a sense of urgency) and 2 (Form a powerful guiding coalition).

The first step relates to identifying potential challenges or major opportunities. The company’s management has to feel a sense of urgency to act on it.

“The urgency is high enough when about 75{d0229a57248bc83f80dcf53d285ae037b39e8d57980e4e23347103bb2289e3f9} of a company’s management is honestly convinced that business as usual is totally unacceptable”.

— John P. Kotter

Nothing will happen if no one understands why you need a data strategy. Therefore you need to make transparent why it is “totally unacceptable” not to have a data strategy.

At the latest, with the second step, you need to start to bring other people on board. Assemble a group large enough to lead the change effort. Don’t think in hierarchies but identify potential supporters of the initiative and get one by one more people on board to convince them about the need for a data strategy. Also, ensure that this group has enough power to lead the change effort before jumping to the next step.

From step 3, you work on defining and implementing the data strategy.

Once you understand the most critical aspects of a data strategy, you can start by defining it with the following five steps.

#1 Add context: Get familiar with your company’s strategy or derive it yourself

Data strategy cannot be seen in isolation. For data-driven companies, it is a big part of their business strategy.

As such, it is essential to understand the business strategy. What are the current challenges of your company? What are the current strategic plans?

If you understand this, it will be easier to come up with suggestions on how data could help.

Sometimes the business strategy is not clear or missing. In this case, I suggest putting in some extra work and deriving it yourself.

There are a few frameworks that can facilitate this process.

You do not need an MBA for this, but reading about those frameworks and putting some thoughts on how your company situation is related to them helps you to be able to talk on an eye level:

#2 Top-down data strategy definition: Develop a vision

Now you know clarity about the business strategy. It is time to derive how data can support it and create a data strategy vision.

The Value Chain approach

The framework of Porter’s Value Chain is again beneficial here.

When going through your company’s value chain and the different activities of Logistics, Operations, Marketing and Sales, Service, etc., think about how data can make a difference in each step.

What are the key activities where data gives you an edge over your competition?

Another way to think about it is to derive the 20{d0229a57248bc83f80dcf53d285ae037b39e8d57980e4e23347103bb2289e3f9} of the value chain activities that can make 80{d0229a57248bc83f80dcf53d285ae037b39e8d57980e4e23347103bb2289e3f9} of an impact when leveraging data to its fullest. This should be then part of your vision.

The Offense vs. Defense approach

From a data-related perspective, another way to prioritize your data endeavors is the Offense vs. Defense approach from DalleMulle & Davenport.

According to their article, data strategy can range from offense to defense.

Depending on your industry and the type of business, you might be in the defense or offense area. And consequently, the data-related priorities will be different.

For example, in the offense area, you focus more on monetizing data and creating new products and services. While on the defense side, you care more about meeting industry regulatory requirements, cyber security, and improving data quality.

Creating the vision

The insights you collect from the business strategy, the essential data-related value chain activities, and the offense vs. defense assessment help to define your vision.

#3 Bottom-up data strategy definition: Identify strategic use cases that fit

The vision itself might be pretty challenging to grasp. Therefore it is vital to add a bottom-up perspective.

Brainstorm data use cases and data assets along the value chain through conducting workshops. Those use cases help to make the abstract and sometimes difficult-to-understand topic of data strategy more tangible.

Focusing on the 20/80 rule means that 20{d0229a57248bc83f80dcf53d285ae037b39e8d57980e4e23347103bb2289e3f9} of the value chain activities make up 80{d0229a57248bc83f80dcf53d285ae037b39e8d57980e4e23347103bb2289e3f9} of the impact.

Once you have enough use cases, prioritize the strategic ones based on the fit with your vision.

Those use cases guide you on what is necessary for the next step.

#4 Looking into the future: Develop a roadmap

Now you have a good understanding of where you want to go.

But it is also essential to derive how to get there. This is where you need to define a roadmap.

The roadmap might be more technology and processes heavy to set the infrastructure foundations. But at the same time, you also should implement strategic use cases early on to have quick wins.

And to realize those strategic use cases, you might have to build infrastructure, acquire data sources, build new data assets, train people in data literacy, etc.

Develop the roadmap for the next 5 to 10 years that aligns with achieving your vision.

#5 Getting things done: Define OKRs

The best data strategy is worth nothing if you don’t implement it.

To align everyone on the data strategy, you need to break it down and have a strategy implementation tool handy.

Objectives and Key Results (OKRs) are ideal for that. So break down your data strategy into objectives and key results to make the implementation progress measurable and align everyone on the implementation plan.

Data strategy is a challenging topic, but it is extremely important.

If your company doesn’t have a data strategy in place and you believe data could be leveraged much better, you should look into it.

Therefore, keep in mind the five aspects to understand before starting..
#1 Data Strategy needs to be driven from the top
#2 Data Strategy is not a technology strategy
#3 Data Strategy removes bottlenecks
#4 Data Strategy implementation is a change endeavor
#5 Data Strategy is about raising urgency and bringing supporters on board

.. and start with the five steps to define the data strategy:
#1 Add context: Get familiar with your company’s strategy or derive it yourself
#2 Top-down data strategy definition: Develop a vision
#3 Bottom-up data strategy definition: Identify strategic use cases that fit
#4 Looking into the future: Develop a roadmap
#5 Getting things done: Define OKRs

If you are not doing it, who else will do it? Driving a data strategy forward can only benefit everyone.