Top 5 Tips on Actioning Insights and Business Intelligence

Top 5 Tips on Actioning Insights and Business Intelligence

There’s a popular phrase that gets thrown around when you start to discuss leveraging data to improve business performance: actionable insights. What is an actionable insight? How do you find one, and more importantly how do you use it once you have it?

There are a few layers to uncover here. Simply put, an actionable insight is an effect, for which the root cause can be directly and inarguably derived with enough specificity that you know how to action a change. Knowing how to identify an actionable insight is only the first layer though. The hardest part of running a successful data-driven business intelligence system is having the strategies in place to act on the insights you find. To help you put your strategy in place, we’ve compiled five tips you can use for better business intelligence.
 

1. Break Down Organisational Silos


One of the biggest obstacles in the way of effective, agile action in an organisation is communication. The larger an organisation, the larger the potential problem. Customers interact with businesses in multiple ways: they shop online, visit a local store, call a call centre if they have a problem, and they comment on social media. If these teams operate in isolation, you won’t be able to gain a clear picture of the way customers interact with your organisation as a whole, let alone solve problems that might be occurring across multiple areas.

2. Look to Unstructured Text Data

A typical NPS survey consists of two questions - the first is an easy 1-10 rating, and the second is a comment box asking why the customer answered the way they did. It’s easy to drop that data into Excel to get an NPS score, but choosing to ignore the comment box data is like chopping the cause off of the effect: You’re unlikely to get actionable insights without understanding why. Analysing your unstructured text data - survey results, comments, feedback forms, CRM data and online reviews - will reveal much richer results than a score between one and ten.

3. Demand Relevance and Specificity

Once you’ve started communicating with your team and analysing unstructured data, you’ll likely be faced with a long list of insights. How do you decide which ones to prioritise? Ask yourself two questions.

Is this insight relevant to my current business goals?


Is this insight specific enough that I can identify a clear plan of action and a measure of success?

Prioritise by acting on insights that meet all of these criteria, and you’ll be affecting changes that will directly improve customer experience as well as improving business performance.

4. Look for Strategic Alignment


Key stakeholders in your organisation have methods to measure their success. These measures are different between different businesses, and they might even be different between different stakeholders in the same business. Think of things like ROI, NPS, or call length, for a call centre. By identifying a concrete improvement to a measure relevant within your business, you’ll be able to choose which insights you need to prioritise to stay aligned with your business goals. Choose an analytics tool that allows you to visualise your data with the metrics you care about and can give you quantitative answers about how you’ll see your data changing as you action insights.

 

5. Perform Regular Analyses


Analysing your data once, or finding one actionable insight (regardless of how juicy it is or how much it improves your ROI) is like taking a snapshot of how your business is performing at one moment in time. Monitoring customer data over time allows you to identify small areas of concern before they blow out into large problems. This trend analysis provides a timeline which can effectively provide context for otherwise unexplained dips (or spikes!) in business performance. This all comes back to identifying the root cause. By performing a regular trend analysis, you’ll be able to identify how events within your business affect your bottom line.

Previous
Next