There’s little debate about how important data is in fueling startup growth. Good data can guide strategic product development, increase marketing ROI, improve customer retention and LTV – the list goes on and on.
However, the challenge for startups is deciding how much (and when) to invest in their data team and technologies while juggling competing needs for dollars and resources.
Knowing where to start can feel overwhelming.
This blog post explores five key questions startups should ask to determine their best path to data success. These insights are from the ExponentialX GTM Live Session “How to Leverage Data & Techstack to Fuel Startup Growth” with growth advisors Bei Zhang, Chloe Liu, and Holly Chen.
The first step in determining a path forward begins with understanding where you are from a data perspective. Think of this like looking at a map of Disney World to find the “you are here” icon.
Start with an honest assessment of your company’s data maturity and understanding the stage of your startup.
Different startup stages require a different balance between quality and speed, so it’s important to be aware of the level of rigor needed.
We hate to disappoint you if you’re looking for a quick tech stack list here. The reality is that which tools make sense depends on your data maturity (see above) and business model. That said, there are some key things to consider along the way.
You don’t have to go big in your initial foray into data tech. But make intentional decisions about data investments early on.
Whenever you bring on a new data tool, there’s the cost of that tool, implementation costs, etc. Often, those are the only costs people focus on. Those factors are important, but it can be equally costly, or more so, to move away from those platforms later when your business requires a more robust solution. So keep those hidden costs in mind.
In the early stages, the focus can be on basic tracking – what information do we need today to answer the fundamental business questions we need to move forward? That’s where free or low-cost tools can be invaluable, such as Google Analytics for marketing and Mixpanel or Amplitude for product data.
But getting a centralized data framework in place early can pay dividends later, which brings us to…
The short answer is as early as possible to give yourself the right footing to grow in the future.
As you move to Series A stage, have 1,000+ users, and need to understand more about segmentation, it is an excellent time to bring on an internal data hire. This person can help you centralize information in one place and avoid the pitfalls that many heads of data in Series B and C companies face – having a lot of tools but no confidence in which one is the source of truth.
At the same time, it can be hard to prioritize a data engineer over a product engineer during early- to mid-stages.
That’s particularly true because, early on, the data engineer typically doesn’t have a playbook—the architecture changes, the foundation changes.
Many companies successfully rely on Google Analytics 4 as a primary reporting tool vs. investing in relational databases, including a Series A company with $10 million in annual recurring revenue. These companies invested heavily in product development to drive growth, and there wasn’t a strong sense of urgency that something needed to be fixed.
But waiting can be costly. For example, Grammarly was about 12 years into the business before migrating from an internal database to Databricks. It was an arduous and expensive endeavor.
The good news is that with the help of tools like Snowflake, AWS, and DBT, which does the data modeling, it’s much easier to build a solid foundation. It’s far less expensive than even five years ago. The clarity is better, and the cost is a lot better.
We highly suggest that founding teams talk to a data advisor to guide early decisions about data setup and tools because it’s doable at a way lower price.
When building a data team, deciding who to hire can be confusing because the roles are conflated. Many people with the same or similar titles are doing different things at different companies.
Early-stage startups should focus on answering fundamental business questions. There’s rarely a need for complex data science and analysis.
Begin with a data generalist with a technical bent. Consider hiring a data engineer interested in analysis or a data analyst with strong tech skills. This person can lay the foundation needed, including tool selection, data centralization, and basic reporting and insights about business performance.
Shy away from a data scientist as the first hire unless your unique business differentiator depends on complex analysis.
I worked with a company that used dynamic pricing as a competitive differentiator. It was a no-brainer, in that situation, to invest upfront in a data scientist.
CapitalOne is another example. It targeted an audience traditionally considered too high-risk in the credit card realm, so the company needed to build complex risk models to identify and assess prospective customers for their business model to succeed.
But for most businesses, hiring staff with the technical skills to lay the data groundwork is most important – because you can’t do data science without a core of good data. As your business grows, you can add additional roles, including a head of data, channel analytics, experimentation, or business intelligence to fill gaps along the value chain.
Instead of thinking about it in terms of titles, given the ambiguity, we recommend focusing on capabilities. That is, transforming data from a raw stage to an insights stage to a predictive stage:
Many businesses start with a data person housed within a business unit demanded by a department head (e.g., marketing analyst, product analyst, sales analyst) and then build outwards. Other companies centralize data operations under one team (i.e., a horizontal vs. a vertical approach).
The centralized data person/team should focus on supporting the entire company to:
Later, you can embed people into specific business units (e.g., sales, marketing, finance) to address the unique needs of those departments. These dedicated data analysts or data scientists will become experts in those disciplines but still can pull from the same “single source of truth” data, call on their peers for best practices, and elevate their insights in the context of broader company goals.
It may be obvious, but the reporting structure for your first data expert and your data team can affect the type of information you glean:
The result can lead to inadvertent conflicts of interest depending on who is presenting the information.
The ultimate goal should be to provide an objective view unskewed by individual department priorities.
If knowing the exact steps to take at every stage of a business was easy, there’d be many more business successes than failures in the startup world. We hope the questions posed and insights provided here will guide SaaS startup founders as you consider how to leverage data and tech to fuel growth.
