Nearly all of your PortCos will be investing 6/7 figures in Machine Learning. Those that aren’t, should have a roadmap to build significant capability within the next 12-18 months.
Last week I was turned on to an excellent book, ‘Prediction Machines’. It introduces the concept of Machine Learning simply as an intelligent way to predict future outcomes.
There is considerable unrealised value in predicting future outcomes. For your PE firm, your portfolio companies and their customers. Today, the key to getting access to this value is data.
The example below will resonate with you when working with portco tech leaders on value creation initiatives.
‘Data is often costly to acquire, but prediction machines cannot operate without it. They require data to create, operate and improve. You therefore must make decisions around the scale and scope of data acquisition. How many types of data do you need? How many objects are required for taking? How frequently do you need two collect data? More types, more objects, and more frequency mean high cost but also potentially higher benefit. You must carefully determine what you want to predict’.
Agrawal, A., Gans, J., Goldfarb, Avi. (2018), Prediction Machines, P.47.
Do you sometimes feel that there is never enough data to make decisions when you want it?
How often do you ask your PortCo tech leaders simple questions like:
- At what point do our customers churn and why?
- What are the scaling limits of our platform?
- How much does it cost us to service each user/transaction?
The answer you’ll often get…. We need more data.
The irony is that PortCos have an abundance of data.
Most are often missing the insight to do something intelligent with this data.
- What if you could react to technology-related churn before it happened.
- What if you could proactively measure, assess and reduce your cloud cost profile in line with customer demand?
- What if you knew that increased investment in product A would create increased demand for product B?
Predictions Machines might just be the missing piece of the puzzle.
How would you benchmark your firm’s Machine Learning capabilities?
Machine Learning should be delivering insight to help you make decisions in your PortCos. Similarly, Machine Learning insight should be driving decisions that affect how your PortCos interact with their customers.
Let’s assume that Machine Learning is on your PE firm’s agenda, what do you see as the biggest area for improvement.
Until next time.
Thomas
You can sign up for more ‘unconsidered needs’ below.
For more ‘unconsidered needs’ you can sign up below.