Leveraging Data to Drive Profitable Growth – A Joint Research Project with INSEAD Business School
Caroline Zimmerman, Director of Data Products & Strategy at Profusion
Data teams have taken a hit in the last couple of years. Since interest rates went up and companies were forced to look at their P&L’s with greater scrutiny, executives have been asking a lot more questions of their data teams around value. What, exactly, are they getting in return for their sometimes eye-watering investments in data lakes, machine learning models, and AI? The answer has often been unclear, and so data teams have seen their budgets slashed. But then, Profusion started doing more work with private equity- (PE)-backed companies and the digital and data folks on the value creation teams at these PE firms. And the trend was different. Markedly different. Investments in data and digital were going UP in this space. Exponentially, it seemed. And they were delivering value. So what were they doing differently? What was their secret sauce and how could other companies learn from their example how to invest in these technologies in ways that went beyond the hype cycle and actually made companies more valuable?
I decided to find out. I teamed up with Theodoros Evgeniou my old Prof. of Decision Sciences & Technology Management at INSEAD Business School, as well as Shahbaz Alam, Head of PE for AWS in EMEA and Oliver Blaydon, Head of Digital at Armstrong Knight, an executive firm focused on financial services, on an INSEAD research project focused on studying how PE firms are leveraging data & AI to increase enterprise value. Here are my favourite highlights from those research findings:
Value is clearly defined. Enterprise value is a function of EBITDA (a measure of profit) and the valuation multiple, which is based on a company’s industry. Part of PE’s secret sauce is that they don’t invest in anything that they don’t think will either affect EBITDA or the multiple. That cuts through all of the “let’s just build a single source of truth and figure out what we do with it later” hogwash. They focus on 2-3 growth initiatives, figure out how data and AI can support those initiatives, and then bang they’re on their way. They build just enough infrastructure to support those initiatives, instead of boiling the ocean with an unnecessarily large tech stack.
Time is on their side. PE typically invests in companies on a 3-5 year time horizon, in which time they expect to sell the company for more than they bought it for. This gives them a couple of years to deploy their value creation plan. This is enough time to do something meaningful without the pressure of quarterly results, but also doesn’t allow you to drag your feet. It’s a great balance between long-term thinking and a sense of urgency.
Regular BI doesn’t increase EV, but growth modelling can. Growth modelling is a quantitative snapshot of how a business makes money. It’s a set of simple, streamlined KPIs that capture your sales engine from a revenue and costs perspective. It can be as simple as capturing these metrics in Excel, or something more automated and scalable. The point is, it’s a tool to help you figure out where you should invest for growth, and how much return you can expect from those investments. You can’t A/B test the value of a growth model, but it helps you with all your strategic decision-making, AND increase the likelihood of a competitive sale at the valuation you are looking to achieve? Why? Because you can actually back up your hypotheses for growth with data. A business with a clearly defined growth model is a company that deeply understands itself – and that alone has value.
If you’re looking for ways to increase EBITDA using data and AI, look no further than commercial team enablement, pricing, and demand forecasting. Commercial team enablement is all about helping sales and marketing teams target high-value prospects and making sure they’re pushing them through the funnel. It’s about using data to determine the characteristics and behaviours of your most profitable customers, and leveraging that understanding to target high-value prospects and make sure they’re converting. Pricing is about determining the elasticity of your pricing, but also rationalising pricing across your products to maximise revenue. Demand forecasting is focused on using machine learning to predict how much you’re really going to sell of a given product or service so you expend the correct amount of resource to maximise sales while controlling costs.
It’s rarer for data & AI to drive the kind of innovation that affects the valuation multiple – but GenAI could change that. So far, POCs for GenAI in PE-backed companies have focused on internal productivity gains like document summarisation or coding assistance. But organisations are looking at how they can incorporate GenAI features into their products and services that create new revenue streams and business models. This space is evolving quickly, so we’ll be updating the paper in 2025 to keep abreast of these changes.
Read the research paper here
