I hear this quite often, from friends that work at companies that say they are data driven, and while i am sure they believe it, in many cases upon chatting for some more time I realize that many don’t fully understand what being driven by data truly means and entails, here are a few points on how an organisation can be really driven by data.
I also realized while exploring this topic with these friends, some of which own tech startups, that the value of making decisions based on insights and data is often missed, so before I detail a few aspects of how to be data driven, let me take a step back and explain why taking decisions based on data is actually important.
What differentiates successful tech companies from the others?
I would argue that its how they collect their data, validate it and use it in their decision making process, usually at the product team level and all the way up to their leadership team, often in these companies, relying on data to make decisions is part of the company’s DNA.
The ability to derive insights from the data you collect is what really differentiates these companies, they have mastered tools and processes that facilitate access across all layers of the business to the data they collect, PM’s, engineers, UX/UI, copy writers, customer service and commercial teams have access to all of the data that is available (with proper regulatory and permissions management to ensure compliance with things like GDPR), and they usually have dedicated team members or a complete department to validate and store this data and ensure stability and accessibility.
A recent Forrester research indicates that advanced insights-driven businesses are 2.8 times more likely to report double-digit year-over-year growth. And 31% of advanced businesses say that their ability to drive new revenue streams is a primary benefit of using data, versus 17% of beginners.
So here are a few main points you need to take into consideration and reflect upon to see if you are a data driven organisation
Interpretation of changes in user behavior
Lets give a simple example to elaborate, lets say you’re using google analytics to track behavioural user metrics across the check out process in your product, often this is seen as one of the basic steps of data collection, a product manager of this product or area of the product would normally check this data often to see if anything changes, others in the business such as LT and sales might check it as well.
At some point, a change is seen in this data and reasons as to why this change might have happened are made, and sometimes solutions or changes based on this data are generated or implemented, and because these changes were done based on this change in the data, this team or company thinks they are basing that decision on data.
In reality this scenario of using data is just scratching the surface, product teams here fall into the trap of assuming their reasons for explaining this change are valid, and often don’t validate these assumptions, approaching every assumption as a hypothesis needs to be part of the DNA across the business and all assumptions should be validated either by other quant. data if available or by qual. data or both (even better).
Do you trust the data you have?
Many focus on collecting as much data as possible without thinking of how do we make use of this data? how do we store it and maintain its accessibility and what are the definitions that we need to have within the business to be able to use this data. The end of objective is to become an insight driven business, many business collect allot of data but continue to make decisions based on everything other than the data.
So first things first, you need to trust the data that you have, and define metrics/KPI’s that matter to your business based on this data, make this data accessible to every layer in your organisation and make sure tools to visualise and see this data are available for product teams, skip beyond vanity metrics and focus on metrics that drive your business. Grafana is a powerful open source data visualisation platform.
Data is subjective
Someone will always decide what data to collect, how to collect it and how to store it, all of these decisions play a factor in determining how useful that data is, so you always need to ask yourself, where does this data come from? and why was it gathered and presented.
Empowering everyone to take decisions
Now comes the harder part, once you’ve collected data that you trust, made is accessible and monitor-able, the hardest part is actually empowering your product teams and individuals to make decisions based off of this data, empowering these teams to innovate without having to go allot of red tape while addressing risks that come with such empowerment.
This empowerment is all about fostering innovation and making independent decisions, and is less about pushing down ideas and decisions from leadership & management on to teams to just implement.
The decisions product teams make should be visible and accessible by the organisation, to foster accountability and collaboration.
Validation doesn’t limit you
Some consider making decision based on data in the form of an A/B test limiting to their freedom or their creativity, which is quite strange when you think about it, how is validating your idea with data limiting? to me I don’t see how you would do product development in any other way…
A/B testing should be at the core of developing features
Empowering your teams with the ability to experiment (with A/B tests) and test their hypothesis should be the heart of your strategy, validating new features and ideas through A/B tests should be the only way to roll out a new feature, something that is much easier now than it was 10 or 15 years ago, there is an array of tools that allow you to get started with a relatively small investment.
For the sake of keeping this post short I will skip going into detail about A/B testing and leave that for another post and another day.