Follow the "data as a product" model to help the company make the best decisions as fast as possible
- Customers = follow colleagues {journey with data is frictionless and constantly iterated}
- Product = company data {product and its features are trustworthy, reliable and helpful}
"Good" self-service analytics
"It's the experience that matters, not the functionality. As long as people are comfortable with that experience and trust the result it produces, we can call it self-service"
⭐ Indicators of good data culture
- Outside team members should be able to independently and painlessly answer their questions in the BI tool
- Thorough data documentation (the structure is intuitive, easy to navigate)
- Async resources for support (especially important for distributed companies)
- Eg. How to setup data, an overview of the analytics ecosystem, how to ask for help
- When they build a visualization, they trust the answer they’re given
- "They trust data and have confidence in their abilities to use that data"
- Colleagues post company updates
- Investigative questions or infrastructure tasks - data questions
😍 Benefits of a good self-service model
- "It is more effective to scale knowledge than to scale bodies"
- The team spends more time improving the data experience than hiring new people
- Reduces time spent on reactive work (eg pulling numbers/building dashboards)
- Empowers peers; colleagues can answer questions on their own and hence faster decisions
- The data team can spend more time on more complex, high-value tasks (insights, long-term strategy)
❓ Who is "self-service" for?
For younger companies, if organizations can build this early on, it will be easy to scale
For large scale organizations, start on a small scale to see what it could be like across teams and the organization
dbt Labs started a dedicated data team in August 2020 > company wide-initiative in spring of 2021 (when the team was at 100 people)
<aside>
📌 It helps when non-data teams have at least one member with a data background. It empowers the team to explore data and draw their own insights.
</aside>
🛣️ Road to "Self-Service"
Some questions to ask at the planning stage:
- “What is the current vs. ideal journey when interfacing with data?”
- “How do we set them up for success?”
- “Where are the knowledge gaps? What resources are we missing?”
- “How do we communicate these new expectations to the greater org? To incoming team members?”
- “How do we measure if this is working? What metrics should we establish?”
- [ ] Planning → Current data journey
- Establish personas
- Prioritize a list of resources
- Communicate expectations
- Set success metrics
- [ ] Executing
- Data Documentation
- Record demo training materials
- Record and evaluate every data request
- [ ] Setting Expectations
- Announcing to the team; goals, and roadmap
- Onboarding new team members
- A walkthrough of the Data resource page
- Intro of team structure, responsibilities, and operations
- Explain the two personas
- Set expectations upfront
- [ ] Empathy, Patience, Reiteration
- Make it easy for your colleagues to ask questions
- Make them feel empowered via affirmations -- they are the subject matter experts
- Writing resources != default to linking docs
- Reiterate on your self-service strategy
- [ ] Celebrate
- Vocally celebrate wins
- Tell them thank you
“So often, as data teams, we chase the self-serve experience that we think we’re supposed to build. We should be more critical of that, and chase the self-serve experience that makes us and our customers feel most at home.” self-serve is a feeling
P.S. These notes are from Leena. 💙 I work at Atlan.
