Research Case Study: PE Strategic Portfolio Growth

The situation

A private equity firm’s investment thesis was that the service industry was fragmented. Therefore they were interested in multiple acquisitions to create a “dominant player.” They identified several sub-verticals (HVAC, roofing, landscape/arbor, electrical, etc.), analyzed companies in relation to the micro-trends within each sub-vertical, and identified potential acquisition targets to build dominance in multiple sub-verticals.

They were building a master warehouse to capture a diverse array of data points, and get a multi-dimensional view of these targets. They wanted to go beyond revenue, profitability and growth rate, and believed that fleet information could provide indicators, or validate traditional financial metrics.

Fleet Data Research Case Study

What we provided

To support this, first we identified and prepared a large volume of data (150K companies and 300K locations) for them to acquire. This was provided as an anonymous dataset with several detailed fleet fields that could otherwise not be provided.

Second, we enriched our firmographic data by adding dynamic variables like year-over-year growth in revenue and fleets, private or publicly traded, years in business, credit ratings. Further, we were tasked to reconcile and calibrate firmographic data with fleet data. For example, maintaining the ratio between employees and fleet count within acceptable deviation norms.

Then we expanded our API gateway to accept more parameters, provide faster throughput and other measures to increase the utility of their analysis.

Finally, we provided a mechanism to exchange the anonymous token for deanonymized, identified information. This preserves the confidentiality of vital fleet data while allowing quantitative methods to leverage them in getting the best outcomes.

How were our advanced datasets used

The customer’s data science team ran dozens of simulations to build scoring models to rank order the various subverticals.

They were able to pull in the anonymized data and fuse it with known data through aggregation techniques, including geography.

Fleet data provided an additional dimension (or factors) that were different than those provided by financial, base firmographics and raw data feeds by themselves. This provided additional lift in identifying targets that would have otherwise been considered undesirable.

Combining with identified data, they compared score groups and further refined the criteria and thresholds.

The conclusion

The customer created a robust data framework that was unique in the industry, especially for anyone trying to penetrate the $395 billion service industry.

While the initial data warehouse set-up took time beyond standard deployments, they were able to iterate and come to conclusions much faster over the longer term, and that provided superior returns.

Key takeaways:

  • Seek and acquire related information that no one else is thinking about
  • Test several hypotheses and scenarios rapidly to arrive at a much more optimal solution
  • Create multi-dimensional views and models to find incremental lists and hidden winners
  • Acquiring fleet data can be a fraction of the overall cost of the operation and pay rich dividends

What are your research goals? How can you build massive datasets, and create derivative data to provide deep insights? Contact us if you have a project to discuss.