How can Data Teams use Product Thinking?

''Wrong projects are ones that defy conventional business rationale, creating outputs that either few people want, that add little to no real value, or that undershoot the desired benefits because they are so difficult to achieve."

This is the primary risk for project failure as per a 2023 HBR article. Doing useful things fast is the key to success was the main point of my 2022 US Army presentation. That is more relevant now than ever. The end of low interest rates eliminated free money and has led to change in engineering practices, trajectory of data vendors and tech layoffs. How do you determine what is useful? Here are some ideas:

1. Internal data teams need to operate like entrepreneurial start-ups

If you are a Data Engineering, Data Science or Data Governance team, servicing internal customers, think like a Start-up that focuses on Value Creation. Scrappy start-ups work hard to get Product-Market-Fit. No clear value, or reasonable time to value creation, have been a reason for failures of central data teams, and why Data Governance initiatives, routinely fail to succeed. If you are a central data platform/shared services data science team, keep reflecting on these:

1. Will other internal teams want to work with you, if they had a choice not to?

2. Are your internal customers so delighted with you that they will never want to do what you do?

3. If you were not permanent employees but contractors, will they extend your contract next month?

These are hard questions. They keep you grounded. Being grounded makes you see reality for what it is, be humble, open to learning, and proactively change, so you will always remain relevant and valued.

2. Leads need to connect team strengths to company goals

As a data team lead, your job is to understand the following

1. Your company's goals

2. Your manager's goals

3. Your team members' goals
   

..and craft a strategy that gets your team focused on the most valuable thing, they are uniquely placed to solve, connecting what the company goals are, with what your manager has been entrusted to deliver. It is foundational for you to realize what exactly your job is, before you can help your team, or the company.

3. Start with customers not with the technology

Who are your potential customers? Look at the org chart and start making a list. Shortlist departments that may have the problem you think is worth solving. Identify a range of people, at various levels, within the shortlisted departments, so you cover a spectrum of potential customers. Have 2-3 people max per department.

As someone who has been a technologist since 2000, I have been guilty many times of starting with technology and not with the customers, I was supposed to serve. Steve Jobs said this best in 1997, You've got to start with the customer experience and work backwards to the technology.

4. Do a survey and interview people

An internal survey is quite useful to get data from your shortlisted customers. When you create the survey, ask open ended questions about the problem area you are interested in. Then go and talk to them and make sure they care about it. Don't craft your survey or interview questions to push the tool/solution you are passionate about. Keep an open mind. Learn Product Discovery techniques.

5. Listen to their pain points intently and test your understanding

Simply joining Slack/Teams channels of teams you are interested in serving and engaging in conversations, is a powerful way to understand their challenges. Develop active listening skills that Shreyas Doshi spoke about recently and use them when interviewing. Write down what you understood and review it back with them. If you have examples of problems that your stakeholders have raised in email/Slack etc., connect those to the work you are proposing, so it is easy for them to relate.

6. Define the problem concisely in a sentence

Once you have the pain points from the customer, you need to define the problem crisply in a sentence. If you need a paragraph, you probably haven't understood it enough. When you define the problem ensure that you are defining a problem and not trying to sell a solution. A simple tactic from Lant Pritchett (Harvard Kennedy School), is 'Write down three other solutions to the problem'. If you are not able to write down other solutions, you are most likely selling a solution. This is particularly the case with tech tools. Questions for a new technology is a useful checklist before considering new technology.

Ideally, your first solution is a pain killer and not a vitamin. For radical transformations such as moving to a Data Mesh operating model, start with a widely recognized organizational pain point. Getting people aligned is more important than technology in data.

7. Define goals using OKRs to ensure alignment and eliminate the 4 risks

OKRs or Objectives and Key Results is a simple framework to bring alignment on what you are going to do to solve the problem and how you will measure progress. It is the answer to two questions:

• What is the goal you are trying to get to?

• How do you know if are getting closer to it, along the way?

OKRs are deceptively hard. Invest time in Andy Grove's OKR framework. John Doerr's What Matters site is useful. How Google sets goals: OKRs is an excellent deep-dive. As Gokul Rajaram explained in a recent interview on Product Thinking and Innovation, customer adoption as evidenced by change in behavior, that leads to commercial success, is the only real goal. This step of coming up with clear OKRs is often missed. In the case of data teams, it could be usage of tables or tools, for analytics teams, it could be adoption of Business Intelligence Dashboards and Reports. An easy-to-understand OKR aiming to solve one small thing, is far better than massive, but comprehensive, utopian ones that are confusing and not measurable.

Another technique that may be useful to bring clarity and alignment, particularly for multi-quarter initiatives is the Amazon PR/FAQ . What you are trying to avoid with all these tools are unpleasant surprises that come up with when what has been worked on for months, is not what your consumers thought they were getting. Remember, customers are hiring you to solve their problem: success is not if you have all the company's data in a Data Lake, it is they can easily solve specific business needs with the data.

Have the SVPG Four Big Risks always on your mind. Learn ways to tackle them from their classic book Inspired

1. Value

2. Usability

3. Feasibility

4. Viability.
   

Summary - Your time is short: Focus on creating products that bring customer delight

Your time in any organization is finite. Make the most of it by doing things that honestly matter with simple steps such as talking to people, focusing on things that play to your strengths, help grow colleagues in other teams so your organization wins by providing exceptional service their customers.

Useful Resources

• HBR - Approach Your Data with a Product Mindset

• Thoughtworks - Shifting mindsets: why you should treat data as a product

• McKinsey - How to unlock the full value of data? Manage it like a product

• Fivetran Blog - Product thinking for data teams

• 
  

---

Photo by Tanner Boriack on Unsplash