Why you need a Tracking Plan

Before diving into those endless reports on Google Analytics, opening Jupyter Notebooks, or crunching statistics ask yourself, are you tracking what your business need for achieving sustainable data-driven growth?

If you thought too much about the question, the root cause might be your Tracking Plan. Let's take five minutes to go through why having a structured Tracking Plan will save your project manager time, support analysts reports and bring more insights to stakeholders business requirements.

First, what is a Tracking Plan?

Generally presented as a spreadsheet, your Tracking Plan, in a nutshell, is your north star goal determining what, where, and how your data is tracked. As also described by Andy Jiang, in the post What is a Tracking Plan, this document will contain three key pieces of information:

• The events (and their properties) to track,
• Where in your code base and/or app to track them, and
• The business justification for tracking them.

Organize your strategy with the right plan

From a project management point of view, maintaining an organized reference from tracked events, with descriptions of how they connect to the S.M.A.R.T goal of a certain project, for example, will support not only the engineers placing the trackers or tags but will establish a common ground of discussion with your stakeholders. Imagine that in a business decision meeting one good idea is generated regarding a new tracking step on your conversion funnel, Datamindly's perspective on how data-driven processes make us think in terms of funnels is a future content. Opening your organized tracking plan will bring everyone involved to the same technical level without getting deeper on the terms.

As a best practice, setting a naming convention is crucial for both organization and aggregation. Since one of the steps of a data project is analysis, your team will thank you for the correct aggregation of data, enhancing data quality and therefore making analysts life a lot easier. I use and recommend Lunametrics naming convention, you can find it here.

Why not tracking everything?

We agree that storage and cloud prices are getting cheaper every year, and it's intended to become a commodity such as electricity or bandwidth, however, I will need to go against the "Track everything you can" principle. Tomi, from Data36, explains on his Data Collection blog that you should track everything, because you can never know when you would need that data in the future. This is true but from my experience, such data overload might drive to overwhelming reports with analysts lost if they don't have enough experience crunching and segmenting. Datamindly's study case explaining when to not track everything also as a future content, some pictures of Google Analytics disastrous Events > Flow Report below. I still strongly recommend Tomi's content, great materials on his blog, such as Predictive Analytics 101 and his Junior Data Scientist course series.

For me, tracking is about learning and taking action. This is a shallow article regarding the topic's complexity, so for a company getting started and building the data-driven intuition on their teams, focus on the data that will support data-informed actions for driving growth.