Getting the right data, getting that data right, and then using it effectively in decision-making is key to unlocking the full potential of businesses. Inflexion Digital Director Alex Mathers led a panel to discuss this at the 2021 Inflexion Portfolio Exchange.
Knowledge is power. This has been said many times – however ensuring that insight is available at the right place and right time is easier said than done. Even if within companies renowned for data, information silos can still exist (often in finance), holding the business back.
“Good information needs to be available in all parts of a company, not just finance. If finance guys can come out of their silo and share the information that they are holding, we’ll have much more intelligent conversations going on in the wider business,” says Simon McMurtrie, serial Chair and adviser to a number of companies in the consumer space.
The benefits of harnessing data are evidenced from many businesses in the Inflexion portfolio. Take the case of Comparison Technologies (CT), a multi-channel price comparison technology enabler for home services which was created from two smaller businesses with different areas of expertise. Decision-making in the firm had been driven largely by gut feel, and CEO Julie Harris recalls, “Gut doesn’t scale a business, we needed data to optimise. Once we took a more data-led approach, we realised there was a lot more optimisation we could do.”
They did, and it bore fruit: in the last two years, the changes drove CT’s gross profits up by over 20%. “Gut feel is good but needs to be validated and challenged, with decisions rooted in data and tracked accordingly,” she said, saying the data-led approach made a “substantial difference”.
Data can also help free humans from certain tasks and focus on those which are more rewarding – for the business as well as themselves as people. John Kerr, Chair of CMSPI says: “Data lies at the heart of automating processes, and this will make us better able to scale as we won’t be as headcount-dependent.” This in turn could help firms focus (hu)manpower on enhancing customer experience in the business rather than more repetitive activities. For CMSPI, it may also enable the introduction of new products and services so the firm can access new markets and customers.
Getting the data right is only half the battle however – building trust in both the data and the models is crucial to adoption. Data outputs are only ever as good as the inputs, meaning quality control is a key element. “If you can make data useful to the operational inputters as well as strategic users, you drive good behaviour as the people capturing the data can utilise its benefits as well,” points out Marc Warner of Faculty, a leading applied artificial intelligence company that applies AI to solve real-world business problems.
Trust is equally built by transparency – at two of the businesses Simon McMurtrie chairs, Scott Dunn and Virgin Experience Days, the teams have worked extensively to both ensure common definitions of core metrics and ensure ‘democratic’ access to data regardless of team. VED saw profits improve after investing in data-driven marketing, with the return on this investment shared company-wide to drive engagement. Scott Dunn is utilising data to focus on better communication with new and existing guests (customers) to maximise lifetime value. Free tools such as Power Query and PowerBI make it straightforward for team members to visualise impact and enable swift data queries. “We now have proper insight that helps us change what we do, and these tools help less data-oriented people to understand the power of data – and inform their decisions.”
Finally, with increasing use of AI and predictive analytics, trust and transparency in how the ‘model gets to the answer’ is increasingly important too.
Explainability is key to building trust in ML models – users have to understand why a prediction or decision is being made, and evaluate whether this is ‘fair’, for example to avoid replicating biases that exist in historical data sets. We therefore put a high degree of emphasis on being able to explain things like the relative importance of different features in a model, or the decisions at each point in a decision tree for example, to ensure this trust is established – which is often just as important as the prediction.
CEO & Co-Founder, Faculty
Putting it into practice
Data skills are highly sought-after in today’s talent market – and therefore many businesses face a question of whether they can and should build this capability in-house or instead partner / outsource.
John has acted as both a provider and a client in his career and feels partnering is very helpful for getting results fast. “You always have to think about what you bring, where are the gaps, and how do you plug them. Your strategy may be partnering and then building up the capability in-house as well.”
Comparison Technologies on the other hand has data specialists in different areas in-house and then looped finance in to be guardians of data integrity. “It’s helped us to democratise data and move from data literate to data evangelists,” Julie says. Having reporting tools enabled this: not everyone loves spreadsheets, but being able to see and understand the data and the difference it makes helps more people to get on board. Enthuses Julie, “I can’t imagine us having a meeting now without data at the heart of it.”