What do business leaders really need to know about data?

It’s a (very) long held view that businesses should be “data-led” or “data-enabled”; buzz words that mean very different things to different people and don’t translate well between industries. Take a SaaS business and a manufacturer; both highly reliant on tech and efficiency, but both having very different data requirements, data structures, data sources, and, most importantly, both having drastically different strategic and tactical decisions underpinned by the insight good data can help provide.

In this webinar, we discuss 4 fundamental questions with Penelope Bellegarde, founder of The Data Touch, a consultancy working with businesses to turn data into better decision making.

The questions we started out with were:

• We all hear about businesses needing to be data driven, but what does this mean and entail?

• What are the advantages / pitfalls of becoming more data driven?

• How does a business start to become more data enabled?

• How can a company’s culture support a data driven approach?

Webinar recording

Short on time? Here are the key takeaways

We all hear about businesses needing to be data driven, but what does this mean and entail?

Lisa’s thoughts: “it’s all about visibility on key decisions to drive business success. Ask yourself whether, for each key business decision, you can assign data points that help underpin your view of performance – whether good or bad. It doesn’t just need to be digital or ecommerce businesses – it’s equally true of any business.

Penny’s thoughts: “it starts with the leaders at the top of the organisation. They need to recognise that an investment in data doesn’t mean an investment in tech. We all need tech but it doesn’t lead to good insights, which is what we really want. So, start by asking the right questions that will drive performance and then identify the data (and tech) you need to answer them.”

Jon’s view: “no decision is ever made with perfect data or insight. So, there’s always an element of unknown in any decision we make. Being data driven is as much a mindset as it is about the tech and data itself. The degree to which you understand and quantify risk, are prepared to be proven wrong and learn from your mistakes, and be open to there being a better way of doing something – which the data can help you identify”.

What are the advantages / pitfalls of becoming more data driven?

Penny’s view: “Data is becoming a new pre-requisite. Everyone needs to have a certain degree of data fluency. This doesn’t mean everyone needs to become data scientists, rather a sufficient level of data capability is needed so individuals can take charge of their own KPIs (for example). Huge advantages also include greater focus, job role satisfaction and an ability to retain key staff.

As far as pitfalls are concerned, there are a few important ones to note. First, data isn’t perfect. It can be affected by bias and we need to be aware of that before making big decisions.”

Lisa’s view: “It can give people confidence to see metrics moving in the right direction. For example, it can help identify where to spend more money if things are performing well. On the flip side, it can also act as an early warning sign to help avoid costly mistakes.

That said, data is hugely rational and the devil is sometimes in the detail. Having good data and insight doesn’t mean there isn’t a place for gut feel in decision making.”

How does a business start to become more data enabled?

Lisa’s thoughts: “Keep it simple. Before you think about data, a business really needs to know its goals and objectives. Second, make sure the data you start collecting is going to be really meaningful. You can evolve and develop over time.”

Penny’s view: “In my experience, people want to dive straight into the data and the tech. Quite often identifying the key questions to answer is the step that’s missed. Once we know what we really need to measure, then we set our KPIs and work out how to build a framework to deliver that. There’s also a difference between KPIs – which help indicate future performance – and KRIs – the end result (e.g. revenue or sales). We also need to be really clear – KPIs need to be individually owned by someone. This brings accountability.”

How can a company’s culture support a data driven approach?

Jon’s view: “This needs to be bottom up and top down. The small decisions taken every day by the wider team might not be significant in themselves, but over time decisions taken without data-led insight will add up and affect the larger issues being discussed and tracked higher up. Likewise, the leadership needs to demonstrate that there’s an appetite to be data led; for key decisions to be informed by data.

It’s also being open to being wrong or to there being other approaches that might work, and being okay with that. It’s a willingness to challenge the status quo.”

Questions we were asked during the webinar

We didn’t have time to respond to these during the webinar so we’ve added our thoughts to them here…

1) Where in a business (/what businesses) have you found the slowest data analytics uptake, why do you think this is?

Penny’s thoughts: “Generally speaking, there’s now a consensus amongst organisations of all sizes and industries on the necessity to become more data-driven. Having said that, I find that start-ups are often more agile and quicker at making decisions than larger corporations who have multiple data teams and decision makers across all management levels to deal with.”

Lisa’s view: “For me this is about the data-points available to a business, driven by the type of business it is, e.g. e-commerce or fmcg. Digital businesses will have more data that can be tracked and used for performance or insights than a service-based or B2B offering.”

2) How do you detect bias in the data used to inform business decisions? Is it always a good idea to remove/correct for this bias?

Penny’s thoughts: “For the time-being: by being aware of all the cognitive biases that we humans can experience from time to time.

Also, with the rapid progress and expansion of Generative AI in the coming months and years, it’s very likely these algorithms will also be trained on detecting their own biases and will learn to self-correct themselves…”

Lisa’s view: “to know if there’s bias you need to fully understand sources and what you’re actually measuring. The underlying methodology and sample sizes are key – when you’re setting up your measurement really understand what you’re measuring and where the data is coming from. If there is bias then it’s not always easy to correct for. It’s also important to consciously know the limitations of your data and then use it as a guide, rather than as an absolute that must be followed. It comes back to making well-rounded decisions.”