Why you shouldn’t hire a data analyst

Don’t get me wrong, I love data analysts. I just know that a data analyst should not be the first person your first data hire.

‍

Data analysts are great. They are able to look into your data sets and come up with a hypothesis on what is actually happening in your business and give you recommendations on next steps.

‍

However what happens if you hire a data analyst as your first data hire is that they spend less time focussed on insights and recommendations and more time spent on generating ad hoc reports for the business.

‍
Challenges with Ad Hoc Reporting

Once this starts, it’s really difficult to get out of this mindset. Out of 50 companies I’ve spoken to in the last 3 months nearly all of them fall into this category. This is a shame because at this point these analysts are not really using analysis skills.

‍

What’s worse is that building reports in an ad hoc manner quickly starts to build tech debt in your organisation. Here’s how. Each data report generated creates another data script inside the code repository. Over time you start to have 100’s of scripts laying around and it’s difficult to manage them. Think about what happens if you want to change a business definition. You now need to change it in multiple different places. This itself is a non-trivial task.

‍

Bottlenecks in Manual Reporting

Another bottleneck is that when reporting is done manually, an analyst can only support a fixed number of people. As the organization scales you need to keep hiring more analysts. Not only does this become expensive but what’s worse is that as you scale you start to have separate analysts for different function areas of your business. This means you have an analyst for sales, one for marketing, one for finance. These analysts are focussed on solving problems for their functional area with little standardization. This leads to multiple sources of truth for the business. How often have you been in a situation where sales numbers and finance numbers do not match? This is why.

‍

Importance of Data Foundation

Compare building a data-driven organisation to building a skyscraper. In this analogy an analyst is like a construction worker. An analyst uses data to build an analysis just like a construction worker uses raw materials to add a story to your building.

‍

The first thing you need while building a building is the infrastructure & tooling. cranes, lifts, foundations. I’m not a construction expert but you get the picture ? If you hire analysts at your first data hire, that’s like having your workers have to go all the way to the bottom of the building to pick up materials before adding a new story. The cost of having to go to the bottom of the building quickly builds up as soon as you add a few stories.

‍

This is a big problem for the business which needs to proceed at a steady pace instead of getting slower after each analysis.

‍

Building the Right Data Team

The solution is to build a data foundation for the purpose of data reporting. Getting started with data engineers who can build the correct infrastructure is the better move.

‍

With the right infrastructure and data engineering team, analysts can focus on the analysis without having to worry about transforming raw data, spend time cleaning the data or worrying about running it automatically. Everyone in the organization uses a central business intelligence tool for data reporting which solves the multiple source of truth problem. It also allows you to make updates to existing reports in a seamless manner.

‍

As a rule of thumb when building your first data team, stick to hiring data engineers first, then analysts and then data scientists.