#98 Four takeaways by Jarkko

The 100 days of Data Economy series is about to end soon. I have covered a lot of topics in my 45 posts. I decided to pick 3 things to highlight. My focus has not been in technology nor purely in business either. I've lived a great deal of my professional life in between the two in various roles. Thus, my interests are in value creation, customer experience and product management as well as development management.


Picking just 3 items from dozens of important aspects is hard, but also acts as a good exercise to prioritize. Anyway, here's the three items I wish to raise from the series.


1. The customer defines the created value

First I want to emphasize the old golden rule of keeping your focus on the customer needs. It is easy to fall into a trap and "internally invent" features and benefits. That way you will just invent data products and services which might serve the needs of internal customers. If you aim for a partner or public level, you need to move your focus and activities to another ballgame. Have you really thought about your position in the data value chain?


Keep in mind that with data products you are defining the added value, but in the case of data as a service, the customer is driving the value creation which you enable. The further you go into service-dominant logic, the more value is co-creation with the customer. Regardless of which approach (product or service), you take, you should not implement anything before validating the value with the prospective customers. This takes us to the next takeaway.


2. Design first to avoid costly mistakes

Jumping directly into implementation after having "a great idea" is 99% of the time 2nd worst mistake you can do. The worst is not to do anything. Wise takes the design-driven approach and fake as long as possible. This is what I call postponing implementation as late as possible. Ideally utilizing fast mocking tools and nocode platforms offer toolstack to iterate quickly a solution with minimal costs and in hours. The bold approach is to design products and services in the marketplace as fully mocks and start implementing after the first sales. Of course, your team has plans on how to move forward if needed and all are confident and saying "we can do this if needed!" I would encourage you to consider this lean approach.


Don't try to guess what the customer needs, interact constantly with them to nail it and improve your offering over time. Know your customer! You should always mock the solution first, then verify the value and fit with the customer. Only after that proceed to implementation!


3. Apply product thinking to your data

Data monetization requires hard work and a systematic approach. Current hot concepts like Data Mesh and Data Fabric both include data products in the core. One crucial element is to treat your data as products as suggested in the Harvard Business Review.


You should productize data internally to maximize speed and reuse capabilities. Servitize data to maximize partner network and customer value. Hire data product owner(s) and start refining a value-creating portfolio of data products and services.


This is the purpose we created the Data Product Toolkit. The Data Product Owners lead the teams and utilize the toolkit in their everyday work. The toolkit forces the team to approach the problems from customer point of view, take GDPR into account, consider pricing, quality issues, data pipelines, marketing as well as lifecycle of the products.


4. Assume personal data to be present

Currently, the data landscape is changing rapidly and we humans are involved in the process as data creators as well as consumers. Legal frameworks such as GDPR define boundaries for how to deal with personal data. More often a person can be identified from the data and thus privacy issues have arisen more and more. You should prepare processes and development in a way that you always assume personal data and privacy issues to be present next to IPR issues. Winners will have the capability to manage personal and industrial data with the same process without hiccups.