Data-driven organization. Data-driven insights. Data-driven revenue/processes/operations/systems /marketing.
Regardless of the subgenre, it’s clear that data-driven businesses are the way to go. However, just saying it doesn’t make it so.
Businesses still struggle to make data-driven business decisions, relying instead on all classic strategies — experience, status quo, and “gut feeling” about the right way to do things.
Today, we’re going to look at exactly how you can be better at drawing insights from your data so that you can make better, data-driven decisions.
What is data?
Before we get started, we want to clarify what we’re talking about when we talk about data, and the difference between data, information, and insight (spoiler: insight’s the best).
We found this visual extremely helpful:
Data makes up the base of the pyramid. Think of it as unprocessed information. It’s just the raw numbers, captured or generated by an organization. For example, imagine some survey results. Data is the Excel spreadsheet the holds all the survey respondents’ answers.
Next is information. Information is data that’s been processed a bit, so it’s easier to consume and understand. The best example here is a dashboard pulling from a data lake or a very large database. Both of these are very difficult to understand, but the dashboard makes that enormous amount of data easy for people to consume.
Finally, insights. Insights are when people consume information (and, sometimes, data) and make observations, create hypotheses, and draw conclusions based on that information/data they just consumed.. Insight is also the first step towards action, e.g. doing something based on your hypothesis or conclusion.
The goal of a data-driven organization is to harness their data, turn it into information, and create insight that can be used to make business decisions.
Now that we’re clear on what data is, here’s how to make data-driven decisions.
1. Centralize and integrate your data
The first step is to understand how your organization works. Only then can you find inefficiencies and actually be a position to do something about them.
However, that task is a lot harder if you’re manually combing through multiple systems and databases to export data to a manageable format so that you can do some basic analysis on it. Because the only thing worse than no data is bad data.
So first, work out what data sources you have. List all the different repositories, tools, and systems where data is captured/stored and put it all into a centralized system, or develop a way to pull it together quickly. Then you’ll be much better positioned to use the data effectively.
2. Know what you’re tracking
It’s really easy to get completely lost in the details, especially when you’re looking at operational data that might span dozens of departments and come in all sorts of shapes and sizes.
For example, is it better for a new sales representative to increase their average revenue from $10,000 per sale to $12,000, or is it better for the supply managers to reduce their on-hand inventory by 10%?
It’s an impossible question. And that’s where organizations run into problems.
Our recommendation? Understand what you want to track before you dive into the data. Then, just track it consistently across time to identify problems and come up with solutions. Sometimes, even unearthing one thing that can be optimized opens up a raft of new possibilities and sparks new ideas.
Tip: 5 Metrics You Didn't Know You Were Supposed To Be Watching
3. Know what your end objective is
It’s critical to remember is that you need to know what your end objective is to in order to actually use data to drive insights to improve processes. Basically, if you don’t know what you’re aiming at, you’re unlikely to come up with a clear hypothesis for how to hit it.
It’s also important to note that business objectives do not come from data. Beyond a high-level look at your analytics, business decisions should come from an executive vision, in response to the market, or directly from your customers — not from internal data. Data should be used to deliver insight to improve your operations and product, in turn helping you achieve your stated business goals.
4. Aim for continuous improvement
Data-driven decision making isn’t a one-trick pony. It’s a way of doing businesses where every decision is evaluated against the best/latest data that the organization has on hand and the previous hypothesis to see if a new idea is better.
And, since data has a tendency to evolve as new data comes to light, data-driven organizations tend to operate in continuous improvement environment.
Continuous improvement is a development approach where, instead of making huge overhauls regularly, you make small improvements, all the time. By building an operations machine that can achieve this objective, businesses can continue to optimize and deliver over time, instead of waiting for a “process facelift” every couple of years.
5. Get IT buy-in
We’ll talk about culture change and buy-in in a second, but in before you even get there you need to get your IT sorted.
First, your legacy infrastructure. Legacy data organization will almost certainly cause you major headaches as you try and pull data. You need to work closely with IT stakeholders to pull the data you need and make it manageable and accessible.
Second, you need to work with IT teams to help manage unstructured data. Organizations (especially large ones) have so much unstructured it can take years to untangle if it happens at all. The best way to give teams in your organization the data they need to thrive is is to work with IT to:
- Prioritize connecting and structuring essential data
- Building accessible analytics platforms to make data analysis quick and easy
- Merge overlapping data
- Find and plug any data gaps.
Building a culture of data/continuous improvement
Sadly, pulling the right data, tracking the right data, and working with integrated systems is only half the battle (if that!).
The greater fight is, as always, with the people driving insight at the top. This is true for a few reasons:
- Moving towards data-driven decision making forces people to change how they behave, and no one like to change.
- Data-driven organizations have accountability built into them, more so than organizations making business strategy based on instinct and hunches alone.
- Data-driven organizations are positioned to deliver a faster pace of innovation — a pace that isn’t for everyone.
These challenges mean that for most businesses, moving towards data-driven processes are likely to encounter some pushback. And that’s where building the right culture comes in. Here’s how.
1. Link performance to metrics
OKRs, SMART goals, commission — whatever you want to call it, the single best thing you can do to become a data-driven organization is use data to define success or failure. Aside from sales and (sometimes) marketing, having a clear KPI for a role is reasonably uncommon.
However, by linking performance directly to a number, you encourage teams to both be aware of organizational data and use that awareness to drive behaviour.
One thing to watch for: employees will hit the metric that they’re incentivized to hit. So make sure what you’re measuring aligns with your organizational objectives.
2. Remove barriers to data
Often data is seen as protected within an organization, and large swaths of useful information are kept away from teams. Sometimes, this is done so that nothing is accidentally deleted or changed. Other times, it’s to give specific stakeholders an outsized control within the organization.
Either way, when you really get down to it, not that much organizational data actually needs to be hidden. Plus, by opening up access to data, you can foster transparency and clarity within your organization. In conjunction with performance metrics, this goes a long way towards building a data-driven culture. Suddenly, there’s no room for hidden agendas or ulterior motives — you’re all viewing and tracking the same data, all the time.
3. Use dashboards as meeting agendas
The problem is as old as time — you need to have a regular time to meet for a status update or a weekly/monthly “touch base”.
But inevitably, you spend the first half an hour arguing about some metric or other. This is when bring data-driven can actually cost you time and resources.
To get around this problem, data-driven organizations can use shared dashboards to track key metrics in real time. Pulling up a dashboard keeps everyone on the same page in the meeting, plus gives everyone a portal to check the metrics throughout the week.
We think in the near future, “data-driven” will stop being optional. If software is eating the world, then data is definitely fueling that hunger. But this is a good thing.
Being data-driven improves your organizational metrics, giving you the insights you need to develop and execute business strategies and hit your business goals.
What’s more, the rapid iteration, continuous deployment, transparency, and clear goals and definitions for success all go part and parcel with a data-driven organization. Ultimately, this serves the business perhaps even better than the insights that come from data by making it easier to attract good employees and keep them around.
We’re excited to see how far data can take us.