5 Steps to Becoming a more Data-driven Organisation

A practical guide to close the loop in Business Intelligence

In our previous post, we explored how today’s process of doing business intelligence is missing the most critical part: the decision-making. As a result, businesses often fail to connect analytics into action and can’t extract the real potential of their data. Therefore, the question is:

How to become data-driven?

The bad news is, there is no silver bullet that will magically turn any business into a fully data-driven organisation.
Becoming data-driven is a project that requires strategic planning, executive ownership and a change of culture. It needs to get baked into the DNA of the organisation and at it’s core is a behavioural change.

However, providing guidance, direction and the right tools is the best way to jump start changing behaviour.
You can only do this by adapting how you do Business Intelligence.

Decision Making needs to be baked into the BI process

Taking actions from the available analytics and metrics must be an integral part of Business Intelligence. Your job is not done after you created a shiny new dashboard and published it. You have to go further.

So, here are five things that need to be part of your BI process:

1. You have to manage Reporting

Yes, Reporting. That sounds old-school but self-service, on-demand analytics is simply not enough. With that, you rely on people thinking about it, logging in to your BI tool and knowing what to look for and where to find it.
On-Demand is great for answering ad-hoc questions but not for fulfilling basic data needs.
Instead, make it really easy for people to get the information they need by delivering it to where they already are: Their Inbox, their Slack channel, you name it.
However, that doesn’t mean you should dump everything on a user’s virtual desk. Reporting should cover the basic needs — you have to keep it essential and relevant. It needs to be tailored to the audience, both in frequency and in content.
Generally, people should be able to choose which reports they get and when. But you also need to provide guidance on what could be relevant for them and check in from time to time, to see if what they get is still relevant.

2. Alert people when something is wrong (or good)

Even with the best reporting in place, you still rely on people manually checking and noticing relevant changes and trends in your metrics. That’s not only extremely tedious and repetitive, it’s also really hard.
Even if you narrow your metrics down to the most important KPIs, when taking multiple dimensions (customer, country, sales rep, etc.) for those KPIs into account, you’ll easily end up with hundreds of metrics to keep track of. Nobody got time for that!

That’s why you should set up a monitoring system for your most important metrics that keeps track of changes, trends and anomalies automatically and alerts the right audience in case something happens.
This makes sure relevant incidents are spotted early and can be resolved faster.

3. Gather Everyone around an Insight

When something is up with your metrics, most of the time it’s not for a single person to act on it but it’s a team effort. That means everyone must be on the same page and next steps need to be coordinated.
You should treat these incidents as mini projects and manage them in a structured way. Depending on the incident, some can be resolved in a quick chat while others might need a project team and take weeks to resolve.
In any case, you should have a clear process how to deal with these incidents.
It’s important to loop in the most relevant people, assign clear responsibility to one person and discuss next steps while communicating brief status updates to a larger audience.
It’s equally important to keep track of all incidents, their status and key highlights in one central place that everyone can access.
This not only helps staying on top of your data but also creates accountability and measurability.

4. Take Action and write it down

Your data is only as valuable as the actions you take from it. That sounds obvious but in reality that’s not what’s happening.
A good way to make sure data insights lead to action is by simply keeping a track record of your actions. Write them down. Make it a habit in your organisation to log the response to any incident that came out of your data. By making this a habit, the decision making itself becomes a habit. Soon, one of the first questions about new insights will be: “So what do we do about it? What’s the implication of this datapoint to our business?”
Obviously, it should not be a big hassle to log this — people will hate it otherwise and simply won’t do it. You don’t want to create a bureaucratic filing system — it must integrate smoothly into your process. And even a short note in context with your data can do wonders.

5. Always be learning

Finally, you have to get smarter about how your organisation uses data. If you follow the steps above, you should have gathered enough data points to become more intelligent about the data consumption itself.
Analyse your reporting and how it’s been used. Figure out what people are checking frequently and what gets ignored. Check your Monitoring & Alerting diligently and figure out if it’s delivering relevant insights. It’s a balancing act between alert-fatigue and missing critical events. Measure how many of the incidents result in actions and how many actions resolved the issues, and so on.
As the saying goes “You can’t manage what you can’t measure.”

This will not only make your process and reporting better over time, logging the process in context with your data will also build an internal knowledge base that you never even dreamed of. “Remember that drop 18 months ago? This here looks similar. Good that I can remember exactly what we did last time” said no one ever (yet).

And here’s how to remember it

We believe these five steps are key to becoming a more data-driven organisation.
To sum up and because no such methodology can live without it’s acronym 😉, here it comes:

1. Reporting — Cover the basics and deliver them
2. Alerts — Monitor changes in key metrics and alert users early
3. Incident Management — Align your team and resolve incidents
4. Track Actions — Log data-driven decisions to make it a habit
5. Educate — Build a knowledge base and analyse how data is used

Yes, you heard it RAITE. Sorry about that 😝. But it’s the RAITE way to do Business Intelligence.

I would love to hear your thoughts and ideas about this topic. Leave a comment or shoot me an email at julian@pushmetrics.io.

About the author: I’m the founder of PushMetrics —we build software to help organisations make better decisions from their data. Go, check our website.