Those of us who work in data, analytics and insight are numbers driven. Quantifying, classifying, and providing “proof” is our purpose. But it’s one of the great ironies of our profession, that we often struggle with quantifying and proving what we do.
We are not talking about the quantity of work we do. Given enough time you’d find it easy to give a number on the amount of code you’ve written or analysis you’ve run. We’re talking about the impact your work has had on the business. The change you’ve enabled.
Some data & insight teams rely on stats on user activity. For example, queries run on their tables, or the number of users accessing a dashboard. These are somewhat useful metrics which help provide context to the impact we are having. But they don’t tell us anything on their own when;
Data, analytics and insight are only valuable when they help affect change.
And that change is most often made because of the insights influence on a decision, or action taken by people.
Don’t take our word for it. ‘Locally Optimistic’, a popular slack community full of the brightest minds in data have discussed this at length.
Take these two comments during one of their recent webinars;
“Knowing the ROI of analytics is hard because it’s usually a third order affect. It is only realised through someone else.”
“The ROI story for a data team comes down to the business saying these guys are great and I’m getting a lot of leverage.”
In a nutshell what is being said here is that;
The value of any data team is largely defined by its stakeholder’s opinion.
How do you gain that opinion? Well, you must ask. Here are some tips to make this easier.
- Collect and record the stakeholders view on the quality and utility of every piece of work.
Just because a model is perfect, doesn’t mean it’s useful. Therefore, we recommend that you should ask your stakeholders if they think your work was of high quality. Did it answer the question, was it delivered on time, did we meet spec etc.
And most importantly was it of high utility? Did the work give you your desired outcome? Did something useful happen?
- Make post project follow-up part of the required process.
Our research shows that many data teams don’t systemise the process of project follow-up. This is a mistake. Knowing how you’re helping provide value over time and for different areas of the business is essential and helps you to learn and improve. Make sure you are doing this consistently by making it part of the agreed process.
By doing this, you’ll change your data culture to one which is more data literate and effective.
- Limit the amount of work you start which can’t link to a desired action, decision or outcome.
If you can’t associate a piece of work with its ultimate purpose, then why are you doing it? Once you’ve delivered, knowing if you’ve enabled that outcome, gives you a simple measure of the value you’ve created, and how to provide more value the next time.
The above principles apply to projects which are easy to quantify the impact of. For instance, marketing and sales often have measurable KPI’s which can link to your work. But not all work is so easy to associate with measurable outcomes. It helps to have a simple system to collect data we can use to prove the huge amount of value data teams provide.