Startup data maturity: When should we…?

Blog
February 12, 2021
Startup data maturity: When should we…?

If we had a list of frequently asked questions from our conversations about data with other early stage teams, the list of questions would be littered with the phrase “When should we…?”. There’s no legal requirement to set up data infrastructure, and there’s rarely a specific, urgent customer need that acts as the trigger point, yet most founding teams gradually feel the need for “better data”. From our conversations and work with hundreds of data teams, both small and large, we’ve seen a relatively predictable evolution of data maturity that we will outline here.

The Stages

We see startups go though roughly 3 stages of data maturity during their early growth. We can bucket these as basic, intermediate, and advanced.

Basic

Teams with basic data maturity are using built-in reports from their SaaS tools or querying their production databases directly. Sometimes, more advanced analysis is done in Google Sheets or Excel, and this is largely a stage that works fine for what it is — basic data sophistication. Usually pre-seed and pre-PMF, teams here are still trying to understand what signals to care about and don’t have a need for more sophisticated tooling yet.

Intermediate

Teams that are a bit more curious about data enter an intermediate phase. Teams in this phase are starting to experiment with more advanced querying of data, and integrating sources of data from across the business into this analysis. This usually looks like a BI tool (often Metabase or Redash) on top of a production database (or a database setup for analytics), and this is usually sufficient to start tracking reliable signals over time, like product usage, or revenue. This works well for companies early in their seed stage, looking for hints of PMF.

Advanced

As teams in the intermediate phase begin to depend on data to guide their decisions, they tend to want a more resilient solution to their data needs that can be accessible to more people and analyses. This stage includes a thoughtful data model that is cleansed, joined across sources, and denormalized for ease of analysis and high reliability. Advanced data maturity infrastructure usually resembles a proper modern data stack, with tooling for extraction of data, loading into a proper data warehouse, custom transformation of that data, and an analysis layer on top. (Charles wrote more about the stack here.) This works great for companies later in their Seed phase of growth and approaching their Series A (or beyond). It is a solid foundation for future growth of the data organization and, with modern infrastructure, scales incredibly well with the company.

3 stages of data maturity: basic, intermediate, and advanced, roughly mapped to funding rounds. Basic maturity ends at pre-seed, intermediate is roughly between pre-seed and seed funding rounds, and advanced starts after seed funding and extends beyond series A.
Stages of data maturity roughly mapped to funding rounds

So when should we…?

Just as setting up an advanced stack too early can feel like overkill, waiting too long to advance along the stages of data maturity can be equally painful. Luckily, there are some helpful heuristics that can guide the evolution across the data maturity continuum that should be easy to understand.

Basic

This is right for you if:

  • You don’t yet feel signs of PMF
  • The product is changing frequently, and any data you might get from your customers is likely going to be written off by your gut
  • You’re not sure what metrics matter (and that’s okay!)

It may be time to upgrade if:

  • You would like a bit more insight into the way your customers are interacting with you business or product
  • You’re repeating the same queries or lookups weekly or even daily
  • There are insights that feel out of reach in your current basic data stack

Intermediate

This is right for you if:

  • You have a few specific insights you’d like to look into
  • You’re iterating on product or GTM enough that data still feels very experimental

It may be time to upgrade if:

  • You’re writing the same queries over and over again
  • Your team does not feel comfortable querying the data on their own (or when they do, they get different results!)
  • Every query feels overly complex because you have to exclude test data, join multiple tables, and remember a number of other data nuances each time
  • There are conflicting definitions of the same type of metrics (what is a customer? what is our revenue? do we count trials/returns in this metric?)
  • You need to parse different types of data (JSON and relational data) across sources

Advanced

This is right for you if:

  • You want more of your company to be aligned on the usage of data
  • You have business specific logic that you’d like to codify and version control
  • You’re experiencing signs of PMF
  • You want a scalable data foundation for your growth

It may be time to upgrade if:

  • Well… you can’t really outgrow this stage! The modern data stack has evolved to the point where this stage actually scales really nicely with the business. There are adjacent capabilities that may arise alongside this stack — say, streaming capabilities for real time predictive insights — but a well built core data and analytics infrastructure should last the lifetime of the business. More investment at this point should be made in better data modeling that evolves with the needs of the business and the growing teams, but the same core stack can be used well through the growth of the company.

Conclusion

There’s a reason data maturity evolves in stages! Each incremental stage of data maturity requires more effort to set up and responsibly maintain, but on the right timeline, this increase in effort is aligned with the incremental value. By following the rough heuristics we’ve outlined in this post, you can be as confident as possible (in a universe of startup uncertainty) that the value you’ll derive will exceed the resources you invest in setting up this infrastructure.

At Prequel, we know how valuable a modern data stack is to startups, and we’re doing our best to lower the energy required to reach advanced maturity and reap the benefits as soon as possible. We’re building the modern data stack for startups, built on a foundation of best practices and modern tooling. Prequel was designed for companies who need a product that just works, and who want more value over more vendors. If you’re interested in what we’re building, or just want to talk about data, we’re always excited to chat. Find us at hello@prequel.co and we’ll be in touch.

Let's set up your data stack_

Bibendum feugiat viverra nisl odio a viverra ullamcorper.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Get Started