In today's business landscape, data is more important than ever. Having a strong data foundation is essential for any company that wants to be competitive. But what do you do if you don't have a dedicated data team? Fortunately, there are a few steps you can take to get your data stack up and running without one. In this blog post, we will discuss how to set up your data stack without a data team.
What questions are you trying to answer?
The first step before even collecting data is to sit down and ask your C-level what questions you want to answer about your business. If you are reading this, odds are you already have some users who are hopefully paying you. If you are a B2B company you are probably reaching out to them via e-mail and tracking that in a CRM or you are acquiring them via ads.
Here are some high-level questions you will answer depending on your stage:
Seed - Do users love my product?
- Understand users and what drives them so you can get closer to product-market fit.
Series A - Do users love my product and are my acquisition channels working?
- Weekly and monthly retention
- ARR, MRR and churn if any
- ARR by cohort (join date, acquisition channel, product launch)
- MRR by cohort (join date, acquisition channel, product launch)
Series B - Do users love my product and are my acquisition channels working at scale?
- Net Dollar Retention by cohort
- Gross Dollar Retention by cohort
- Logo Retention by cohort
Series C+ - Do users love my product, are my acquisition channels scaling efficiently?
- New Sales ARR vs S&M Expense
- Customer acquisition costs (CAC) across channels
The second step in setting up your data stack is to pipe your data from various sources, so you can answer the above questions. This means that it should be accurate, complete, and timely. If you are a SaaS company, most of the data that you need is usually in one of many SaaS tools. Here are some common ones:
- CRM - Salesforce / Hubspot
- Product Usage - Segment / Mixpanel / Amplitude / Rudderstack
- Marketing attribution - Facebook ads / Marketo / Google Ads
Perhaps the easiest way to collect this data is to use an ETL tool that provides a web interface to select specific tables and columns that you find useful to create those metrics. In subsequent posts, we will be going over some common ones.
The next step is to store the data. This can be done using a variety of methods, such as relational databases, NoSQL databases, object storage systems, and cloud data warehouses. The most important thing is to choose a tool that will be cost effective, is compatible with your data set and scales with you as a company. Snowflake, Redshift and BigQuery are the best in-class data warehouses that are time tested.
Perhaps the most important part of the data stack is to process the data. This includes cleaning the data and transforming it into analytics ready format. If you have a dedicated data analyst, they would be tasked with cleaning up the data and creating a rolled up view of the data. There are several UI based tools that help with data transformation as well.
The final step is to visualize the data. This can be done using a variety of methods, such as charts, graphs, tables, and maps. The most important thing is to choose a method that will be able to communicate the results of the analysis in an easily understandable way.
It is possible to setup your data stack without a dedicated data team. While having a data team can certainly help, it is not necessary in order for your data stack to be successful. There are many different tools that can help you with each step of the process.
In upcoming posts, we will break down each of these steps in setting up the data stack and go over best practices and pitfalls to avoid in each of them.