5 tips for building an actionable data story
Let me give you 5 tips to help you ally your great technical skills with engaging storytelling and build a data story that will have people begging for your insight and expertise.
Much is said about the different technologies and skills required to be a good analyst. But the truth is no fancy Excel formula can alone build a clear, insightful data story. The tips below may not showcase all the advanced things you’re capable of doing. However, they do help you meet your ultimate goal as a data analyst: to use data to deliver clear, actionable insight.
Top tips for a clear, actionable data story:
- No need to reinvent the wheel at every analysis — consider frameworks
- Use data visualisation in your favour and not against you
- Show clear action…
- …but don’t try to answer the ultimate question about life, the universe, and everything.
- Bonus tip: how I write data stories
1) No need to reinvent the wheel at every analysis — consider frameworks
In my article about the different ways you can frame the presentation of your findings, I write about how a good data story doesn’t always need to have a never-before-seen methodology or storyline. As data analysts, we sometimes have to put our vanities aside to deliver what our stakeholders really need. And it might as well be that the insight they need is best delivered with a simple storyline of pros and cons, for example.
Besides, have you ever looked at a huge dataset and just didn’t know where to begin? Frameworks can help you filter out the findings that are not important and get you to deliver a powerful, straightforward story that hits the nail on the head for your stakeholders.
2) Use data visualisation in your favour and not against you
I need to confess I secretly envy data analysts who can build impressive visualisations, and this is definitely something I want to get better at. However, the more experienced I become, the more I realise how these skills would be more for personal gain than to be used at my everyday work. If you build an analysis for stakeholders with average data literacy (or no literacy at all), complex visualisations may hinder you from getting your insight across.
I wrote an entire article about how you can build clear and to-the-point data visualisations that elevate your insight with little technical skill. But, whichever visualisation you go for, make sure it’s there primarily to showcase the point you’re trying to make and not your Tableau skills.
3) Show clear action…
Data is expensive. And, if you work for a private business, your stakeholders will expect a return on investment for their data expenditure. So, if your analysis isn’t using data to tell them exactly what their next steps should be, it’s likely that they won’t see much value in it.
This is a mistake I used to make in my first job as a data analyst for a Health & Beauty manufacturer. I used to think that showing the audience what happened was impressive enough. But, in reality, what really makes a data presentation memorable is data being used to clearly justify a set of next steps and focuses.
The biggest challenge to this is to build enough authority and industry expertise that enables you to give good, relevant advice. Perhaps this could be a future article, but the key is to communicate with your stakeholder even before the analysis takes place. Unfortunately, many companies still isolate their data analysts from the important conversations (hey, another future article here), but genuinely actionable data stories can only happen if data analysts are in the know about the concerns, questions, and decisions being made.
4) …but don’t try to answer the ultimate question about life, the universe, and everything!
I know I just told you to go there and let all of your authority and business acumen shine through a very actionable data story. However, now I’m going to pull you back to Earth and remind you that your presentation doesn’t have to provide this unique solution that nobody had thought of and that will revolutionise the future of the company.
A common pitfall is when analysts drive hasty conclusions from the data and end up seeing things that aren’t actually there. For example, you may see two events that always happen simultaneously and jump to a conclusion that you think will blow everyone’s minds. However, correlation isn’t causation and may lead you to go overboard with your insight.
And, actually, when you start your analysis with the mindset that it doesn’t have to change the world, it takes away the pressure and gives you more clarity into what really is an insight and what might be just a coincidence or something to keep an eye on.
5) Bonus tip: how I build an analysis
I’m currently writing an article about a week in my life as a data analyst in the luxury industry, but I wanted to give you a sneak peek at how I go about creating a good data story. For context: I work for a company that offers consultancy to luxury brands on how to deliver a good client experience and my stakeholders are mainly the account managers and, indirectly, our clients (as I don’t present my findings to the clients, instead, the account managers do).
The first step is usually when they request an analysis and send me a brief. I always try to book a call with them so I can dig into the conversations they’ve been having with the client and build an analysis that speaks to them. As I mentioned, trying to build authority and expertise is important.
After that, I go away to analyse the dataset (PS: if I could see the data before the brief call, it’d be even better, but it’s often not the case for me). At this stage, I look at overall performance, and by combining that with the brief, I can come up with my angle, framework, methodology, or whatever you want to call it. I usually have a bunch of interesting stuff I’d like to say, but I often need to filter it down to answer their precise questions.
Then, I build my presentation. However, before I create any charts, I add what I like to call “placeholders”. These are literally blank slides with the insight I want to add in written form. I may even write how I want to display it, if I already have an idea. But this is a very effective way to ensure your story has a flow and arrives at the conclusion you have in mind. It also helps you not to waste time creating full slides just to figure out later that they don’t fit into the story.
Finally, I send this analysis to the account managers and see if there’s anything they’d like to add, change or remove. In an ideal world, you’d like to have multiple touchpoints with whoever requested your analysis. However, unfortunately, this is often not possible in the fast-paced businesses I’ve worked for.
What do you think of these tips? Would you have something to add? Stay tuned as next week I’ll publish an article detailing everything I do in a regular week as a data analyst.
