At Lantrns Analytics, we mainly work with businesses who own digital products and would like to get insights on performance, growth, onboarding, retention, engagement, etc. They have large amounts of data streaming from multiple sources and a willingness to consume reports through their BI tool.
This usually structures our mandates with clients.
In alignment with their business model and growth strategy, we help them:
- Unify and structure their data in a data warehouse.
- Build processes to automatically refresh their data warehouse.
- Define the questions to be monitored via a suite of reports/dashboards accessible through a BI tool such as Tableau.
- Prepare them for further analytical projects such as the integration of predictive models on top of their data, etc.
When we start work on a new mandate, we follow an onboarding process that provides a clear roadmap to our clients. Below is an overview of that process, structured by gates with associated deliverables.
Gate 1 – Getting To Know You
This step is all about getting to know each other. We want to understand your project and also share how we manage our projects. As a preliminary step, we want to make sure that working together makes sense for both parties.
If it does, we want to further define what we’ll work on and how we’ll work together. That includes a high-level understanding of your project’s main objectives and deliverables, important dates that should be considered in our project planning, etc.
We also want to take care of any house maintenance items, such as non-disclosure agreements, working terms, etc.
- Statement of Work
Gate 2 – Deep Dive
Having decided to work together, we plan a working session to first go deep into your digital product’s business model. We want to answer questions such as the ones that follow:
- What’s the market’s pain and your solution
- What’s the business value chain?
- What’s your strategy to attain product-market fit?
But it’s also a discussion about how the development of an analytical infrastructure is to enforce a data-driven culture within your company. Want want to know how we can get to quickly inject data-driven retroaction in your product’s feedback loop.
This is the longest call we’ll do as we want to understand your business, where you’re at, where you’re heading, the state of your analytical infrastructure and start building a roadmap of where we’re heading.
- Overview of business model (Lean Canvas)
- Growth strategy (Lean Analytics stages)
- Product performance main KPIs (Pirate Metrics)
Gate 3 – First Agile Ceremony
At Lantrns Analytics, we adopt an Agile framework to our projects. This provides values and practices that we’ve found to be very valuable when working with clients.
Our approach is loosely inspired by the awesome “Agile Analytics: A Value-Driven Approach to Business Intelligence and Data Warehousing” book by Ken W. Collier. It is not a rigid approach, more an ever evolving set of principles and practices that guides how we tackle and divide and conquer our projects.
At this point of our onboarding process, we are ready to conceptualize what the first analytical deliverables should be. And that starts with high-level definitions of user roles and their main use cases.
For example, a user role might be a marketer that would like to understand what is the journey taken by a new visitor towards signing up to a free trial. That’s a pretty high-level use case for a user role.
After defining multiple user roles and user cases, we can start planning what would be our priorities for our first release, which could last around 3 months for example.
With our first release plan, we now start building a backlog of stories, which are features associated to a use case and which can be delivered within 1 iteration (iterations usually last 2 weeks).
- User roles and their main use cases
- Release plan
- Backlog of first stories/features
Gate 4 – Iteration 0
We’re now ready to tackle our first iteration, which is actually named Iteration 0. This is preliminary work to set up our architecture.
From the Agile Analytics book by Ken W. Collier:
The goal of “iteration zero [is] to set up the development and testing infrastructure, to install and learn some new development tools, and to develop a high-level system architecture and data model.”
So, not only are we setting the infrastructure that will help us collect, store, transform and expose data, but we also start mapping our sources of data and the final entities (such as transactions, customers, memberships, products, etc.) we’d like to have in our data warehouse.
- Documentation of architecture
Gate 5 – Sprint planning
Now that we have a release plan and our basic infrastructure in place, we can start working on our first iteration. Without getting into details, this will involve a Sprint Planning meeting where we’ll prioritize stories, point them (how do we evaluate complexity, risk and time for this story) and commit to delivering a certain number of story points per iteration.
Our engagement is to complete planned stories within 2-week iterations (our preferred length of time). Those stories should provide immediate business value and should be demoed at the end of the iteration.
- Shared project environment (Trello)
- Sprint planning
- Commitment to first iteration
Gate 6 – Sprint monitoring and review
And we finally get to actual work.
You should be aware that we work in close collaboration with our clients as our approach is anchored in iterative/evolutionary work. Our shared Trello board is always updated so that clients can continuously be involved in the project’s progress. We value communication and the sharing of ideas to provide the maximum business value to your team.
At the end of a sprint, we close our iteration with a showcase/demo of completed stories. The goal is to have final approval from the client to mark stories as Done. This is also a chance to discuss new ideas and improvements for our backlog.
Finally, a quick Sprint Retrospective meeting is conducted to discuss how a sprint went and how we could improve the overall process to make our collaboration more efficient and enjoyable.
- Sprint Showcase
- Sprint Retrospective