How to succeed with data?

According to a 2021 HBR Article more than 90% of companies cited culture as the key block to data success. Here are 10 pointers Chief Data Officers (CDOs) can do, to set a productive culture, to be in the 10% that succeed.

1. Data Success is User Delight: Move to user delight as the guideline for success, instead of data maturity models. Different business units have their own success paths. E.g. some departments may never need prescriptive analytics (the apex in some models) for their success - emailed spreadsheets may be fine for their mission.

2. Use Net Promoter Scores (NPS) to measure Data User Delight: Use simple NPS for measuring data success. 'Give a number between 1-10 on your happiness with data to do your job' is the question. Focus on departments with low scores, start listening to pain points and understand their aspirations. Repeat periodically to track change.

3. Base your priorities on the Urgent and Important: Balance strategic initiatives, with delivering continuous value, so users have a sense of progress. Ensure that you get something, even small, done, for everyone, as you go along with strategic programs. The Eisenhower Matrix is useful to take feedback and potential solutions and prioritize.

4. Take a Platform Product Approach to Data Delivery: Many data programs fail as they put completion of a huge base system build, as a pre-requisite for initial value. Nobody will wait months for anything these days. Using Platform Product thinking is quite useful to solve this. It also aligns well the recent Data Mesh paradigm.

5. Scale with 'Hybrid Decentralized Data Orgs': Organizations with 1,000s of people can consider structuring data professionals in a model that James Densmore put forth his recent article. Smaller companies, and companies where specialized data IP is their competitive edge, can consider having dedicated, Central of Excellence model analytics teams- they also help with providing growth pathways for skilled analysts and data scientists. If you are getting started, check out 3 Best Practices from Barr Moses.

6. Embrace Multiple Versions Of Truth: Different business units may have different goals. So even something seemingly simple as the definition of an active customer may be hard to align. Getting used to clearly labelled, different definitions, is useful. CDOs can think about how to create an organization, systems and processes to effortlessly move along the defense vs offense spectrum (see Two Types of Data Strategies).

7. Seek help from Internal and External Partners: CDOs can seek out help from internal partners such as security, compliance, privacy, governance etc. as it has been established that expectations from the role were too ambitious. Data teams can focus on enabling business outcomes and let partners help out in orthogonal, but important areas, particularly in the 'defense' areas. Partnering with vendors in the rich data ecosystem (See Matt Turck's 2021 MAD Landscape), may be worthwhile, for specific challenges in lieu of building in-house.

8. Become Organizational Connectors: Inter-department communication in complex organizations remains an unsolved problem. It has exacerbated in the remote-first workplace. Arif Wider posited that the Data Mesh is a communication and coordination problem. There is a role that data teams can play in connecting data consumers and producers and breaking down silos, which may have other organizational benefits.

9. Become Force Multiplier with Data Empowerment: 'Data is the new oil' is a common phrase: Data is more like solar power, in that it is infinitely reusable. E.g a single order data table can be provided to sales, marketing and finance and they can use it to realize total different business outcomes, limited only by their creativity. CDOs can think about how to responsibly expand access across the organization, and beyond, to add value to consumers.

10. Aim for Magical Data User Experiences: Getting access to data should become like getting access to email and using it effortlessly. Arthur C. Clarke said 'Any sufficiently advanced technology is indistinguishable from magic'. There is a long way to go before access to usable data becomes simple like email. I hope future data leaders can be inspired by Clarke and aim for nothing less than a magical experience for data users!

---

Photo by Silvan Arnet on Unsplash