A 2021 study found that, while 99% of organisations are investing in big data (with 92% accelerating their investments), just 24% believe they’ve succeeded in creating a data-driven organisation. Companies are generating and capturing far more data than ever but that isn’t necessarily driving better decisions and more efficient operations. This is in part down to the fact that companies have set themselves up to capture a plethora of information but don’t yet have the supporting infrastructure to extract value from it. But what’s most interesting in that same survey is that only 8% of organisations view technology as the primary barrier to true data-driven operations, with a staggering 92% citing internal culture and processes as the main impediment.
There’s no doubt that organisations recognise the value of data and have committed budget commensurately. Data is being analysed, reports are being churned out, and presentations are being delivered. So are organisations still challenged by their own data’s ability to drive action, and how can companies better align their efforts with business objectives?
Here are a few things you need to get right:
Many organisations’ attitude towards data is something along the lines of “let’s collect as much information as we can and then figure out what to do about it”. These organisations find themselves overwhelmed by information and unable to extract meaningful insights, or fall into a pattern of producing regular reports for their own sake, with little connection between data and action.
Prior to undertaking an analytics project, you need to establish two things:
- What are the critical KPIs?
- What are the critical operational decisions we’re looking to improve or focus on?
Addressing these two questions will dictate what data you collect, analyse, and communicate across the business. This (critically) needs to be a collaboration between the data team and their stakeholders, ensuring all analysis efforts are aligned with business objectives. Your goal should be to establish standard metrics and reporting frameworks that both parties agree on to minimise ad-hoc requests from the business and aimless analysis from the data team.
Operate at the right scale for your organisation
If you’re an eCommerce giant taking a million orders a day, changing the colour of your “check out” button from green to blue could be worth thousands. If you’re running an online cake business with one or two orders a day, that same change will have no material impact on your bottom line. Overambition is a trap many organisations fall into at the start of their data journey, attempting to replicate what big tech companies are doing without the supporting infrastructure or revenue volume to justify those activities. Just because you can measure something doesn’t mean you should.
Though it’s admittedly difficult to estimate either the effort or expected return of new analytics projects, some form of rudimentary cost-benefit analysis is required to limit your initiatives to what’s likely to deliver value. Prioritise the most financially consequential decisions you make as a business, and for each consider:
- What information would you need reach a more informed decision?
- How much could better decisions improve your bottom line?
- What’s the effort required to capture, analyse, and present data relevant to that decision?
Make your insights actionable
Never present data to a decision-maker that doesn’t in some way help them make a decision. When determining what to present to whom, consider:
A chart showing a downward trend on your website traffic doesn’t tell you anything about how to remediate the problem. Are your organic search rankings in decline? Are you spending your paid media budget on the wrong channels? Is a new competitor stealing your audience? Do you have site performance issues? Understand the inter-relationships between data points and explore what they mean to your organisation. The advances in data science have allowed many companies to identify trends across products or segments that they never knew had existed previously. Answering these questions in your analysis will paint a complete picture of the problem and provide a clear remediation roadmap.
Indicate how you’re performing year-on-year or against industry benchmarks. Show the impact of a particular metric on your bottom line.
Use visualisations that best represent your data and the point you want to make. Ensure you’re using metrics and speaking in a language that your stakeholder understands.
The reality is that while most organisations are committed to being data-driven, few are entirely satisfied with the value of their analytics projects. For most it’s a leap into the unknown, and mistakes are inevitable. As such, you need to be constantly refining your strategy – don’t be afraid to scrap projects that aren’t delivering value, redesign processes if need be, and always keep the focus on impact.