The lofty goal of being ‘data-driven’ is common in high-level strategic discussions at all manner of businesses. But what does it really mean and how can you actually achieve it? Carsten Weidmann, Technical Alliance Manager at Exasol, explains all, including:
– Defining self-serve analytics
– The importance of a data strategy
– Where data literacy comes in
– Tech stack requirements
Whenever anyone discusses the importance of data, there’s one thing you can be sure of: nobody is going to tell you that you need less of it, or that our insatiable demand to understand it is slowing down.
We live in a hyper-connected, always-on world where we rely on a mountain of varied data sources to make business critical decisions – and it stands to reason the faster you can get insights from this information, the quicker you can act. But the key is not just making faster decisions, it’s all about making the right decisions, faster. And to do that you need to build the solid foundations of a data-driven business.
A data-driven business is one which draws on the data it’s generated – along with external sources – and uses this as the basis of all major decisions it makes.
That’s why self-service analytics has become so important – as in theory, the more people you have working with data in your organization, the more you can enable truly informed decisions throughout the business hierarchy.
How self-service analytics helps
Self-service analytics can help by giving everyone who makes and influences decisions access to the information they need to properly do their job. It also removes bottlenecks to information – often eliminating the need to rely on IT departments or data scientists.
For example, in the past if you needed a report, the IT department extracted the data, put it in a format you required, and ran the report. But with data coming in from more sources than ever, this approach doesn’t scale – especially across an entire organization. Existing reports might not have all the information needed to make a profound decision, and users can’t afford to wait for their IT teams to create custom reports.
Start with strategy, rather than tools
You might be thinking about the tools you can use to do all of this. But for me, this comes last. It’s vital you start by creating a clear strategy.
Ask yourself what kind of decision do you need to make to achieve your goals? And what kind of information you need to make that decision properly?
Then you can think of the tools which best serve you. After all, if you have a cutting-edge data analytics tool but you have no knowledge in the organization of how to use it, that’s probably not the most important thing you need.
Data literacy – context is everything
Training is important to building this missing knowledge, or data literacy. But this doesn’t necessarily mean you need everyone in your organization to be a data scientist.
It’s all about making it possible for your teams to read and understand the data they need to access in a way that matches their learning style – and the goals they have for that data.
For example, your revenue-focused teams, such as those involved with sales, should be able to discover the various income streams that make up the organization’s bottom line. On the other hand, your marketing team needs to know about conversion rates from web traffic and the return on investment of any advertising campaigns.
Everyone who works with your data needs to have a common understanding of its meaning. This is where leadership is important. You need someone – perhaps a chief data officer (CDO) – who can identify the gaps and prioritize for the entire business.
Questions to ask of your technology stack
When you do look at your technology tools it always comes back to your priorities and strategy. Ask what you have in your current stack, then ask what you can accomplish with that. For example, you might find the front end is good enough and it’s the database that needs improving.
One key consideration for self-service analytics is performance. As you encourage teams to use more data you need to make sure they can do it easily, without being slowed down or losing their train of thought.
We all know how frustrating it is when you click on a web page looking for answers and it takes more than a few seconds to load the page. Unless it’s near instantaneous, you’ll likely look elsewhere. Or scream at inanimate objects.
And that’s exactly the same when you’re doing self-service analytics. You will have a hypothesis and you’ll want to validate it. So, you’ll start digging into the data. And if it takes ages to run you’re wasting time and potentially getting distracted from the task. That’s why it’s crucial to have tools at hand which keep you in dialogue with your data to get the answers you need.
It’s also important to not have to rely on an IT department for that. You want access to everything so you can work at your own pace – not be dictated by the stack.
Cloud or hybrid?
If you go for a cloud solution you can use pay-as-you-go models. These could save you some money as it’s generally cheaper and more scalable than on-premise.
On the other hand, you may need a hybrid solution. With a hybrid architecture, you can leverage the cloud benefits of flexibility and potential cost savings, while keeping your sensitive or highly-regulated data on premises – and still provide your data consumers with one single access to all that data.
What gives you the most return on investment?
It’s notoriously tricky to pin-point a ROI on different tech stack combinations. But one of our customers, Piedmont Healthcare, reported that once they had a clear strategy for their self-service analytics goals, using our analytics database and Tableau really helped. With greater analytics performance, giving more people greater access to data, implementing more metrics and by working with a wider variety of data sources (which wasn’t previously possible), they estimated they reduced the rate of harm for their patients by 40%.
So, as you can see, self-service analytics really does help build a strong foundation for any data-driven business – and the key to doing this comes down to having the right strategy, a data literate team and tools which match your goals.
Carsten Weidmann is Exasol’s Technical Alliance Manager. He has worked in software development, consultancy and project management in the finance software space. Carsten has also founded his own tech business, which helped its customers identify bottlenecks and inefficiencies. Today, he helps our partners to leverage the adoption of Exasol with their clients and to maximize the success of their applications through integrating Exasol in their software stack.