Subscribe to DSC Newsletter

Private Equity and Data Science: Due Diligence Stage

An emerging trend in the private equity space is an enhanced focus on data science.

This focus has historically been more on the operations side (post-acquisition) where data scientists have been leveraged to help companies improve performance in many key areas including marketing, business intelligence, financial analysis, and human resources. The advantage of focusing data scientists on the operations side is rather obvious: after an acquisition has occurred, the data challenges and opportunities of the portfolio company are more accessible.

More recently, some target companies have been opening up their books a bit more at the due diligence stage. This allows private equity firms to dig more deeply into the key revenue and profitability assumptions. Key analytics at this stage can include understanding not only the detailed drivers of profitability but also customer segments, factors influencing operations efficiency and direct competitor comparisons. Exploratory insights from mining databases including public data, social media and data vendors can be linked with company data to see if external trends are similar to internal trends. This may help better define any unique value propositions that company has or market segments they potentially can explore. For businesses with physical locations (retail, restaurant, etc.), data scientists can explore the assumptions regarding future site selection, projected number of sites growth, cannibalization, etc. For customers with direct customer contacts, data scientists can help understand the customer value

Leveraged buyout models (often Excel-based financial calculations) tend to have a rather standard structure. The formulas are well understood by financial analysts so many companies are using the same (or very similar) financial calculations. The variation across companies is usually driven by the key input assumptions including the revenue and income growth assumptions.

A key distinguishing factor for private equity firms is how well they understand the key inputs to those leveraged buyout models. Detailed data analytics at the due diligence stage can give them an advantage in better understanding the assumptions and limitations in the projected revenue and projected earnings…resulting in more accurate projections of the target companies value. #privateequity #datascience #data #analytics

Views: 1662

Tags: #analytics, #data, #datascience, #privateequity

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Cristian Vava on July 5, 2018 at 6:23am

Sharing data, models, and assumptions gives analysts the insights and confidence required by the investors. However, my opinion is that sharing too much might be dangerous. On one hand the rate of acquisition / investment following the due diligence is quite low, traditionally in the range of 5-20% depending on the field. On the other hand some analytics based companies have found ways to make profits way above their cost and sharing the strategy might leave them exposed - it is too cheap and easy to copy their product and patent protection is not possible.

To describe the category of algorithms I would keep confidential let me use a very simple reality inspired example. The company offers an app that gets embedded into web pages offering to measure your IQ for free by answering only 10 questions. Here, we can distinguish 3 different types of analytics.

  1. User analytics - the algorithm and data (allegedly) supporting the IQ inference from user's answers.
  2. Operations analytics - supporting - as you mentioned - marketing, business intelligence, financial analysis, etc.
  3. Customer analytics - the algorithm and data implementing the marketing segmentation of users based on their answers. As you might have guessed the result is sold to clients paying for leads. One potential client could be an investment fund looking for customers with a certain level of education and funds to invest.

This last algorithm and its associated data bring the real value of this company. Sharing it can be a mistake. It cannot be legally protected and any decent software developer can implement it in less than a week of work. For this not so hypothetical company the initial investment was less than $25,000 and the annual revenue stayed over $200,000 for about 10 years in a row.

If the example seems too simplistic then we can think of Google's business model (which is by the way not that different albeit much more sophisticated). I am the user benefiting from the free search capability. The paying customer is the company selling ads and buying leads based on the segmentation algorithm applied to users like me. I presume this data and algorithm are out of reach even for most of their senior executives. The tiny segment of data given out to customers doesn't qualify for anything more than marketing communication. Sharing key assumptions like revenue and income growth can be done safely and convincing without exposing the data from which these were derived.

Videos

  • Add Videos
  • View All

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service