Data Recipes: Benchmarking Middle Management Quality
December 20, 2024
Table Of Contents
A systematic, data driven investment strategy for evaluating leadership quality and operational efficiency across companies
For investors, assessing leadership quality is critical for evaluating the growth and performance potential across portfolio companies and other potential investments. Middle management, in particular, has a significant impact on a company’s executional ability and operational efficiency.
In this recipe, we’ll walk through a systematic, data-driven approach to extract, analyze and score middle management quality across key attributes including role expertise, industry experience and tenure stability.
Ingredients
PDL Datasets:
Inputs:
List of target companies for evaluation
Role definition for middle management employees
Company definition for benchmark companies
Outputs:
Target Experience Arrays: Arrays for each target company (containing the work experience history for each middle manager in your target company list)
Benchmark Experience Array: An aggregated array to serve as a reference (containing the work experience for middle managers across comparable benchmark companies)
Middle Management Quality Scores: A combination of quantitative and qualitative metrics for each target company (calculated from the Target Experience Arrays and Benchmark Experience Array)
Key Fields Used
Company: Name, Website, LinkedIn URL, Industry, Revenue, Funding, Location, Headcount
Person: Job Title, Role, Seniority, Location, Work History
Directions
Step 1: Prepare your target company list
Create a dataset of all the target companies you would like to evaluate and score. For each company ensure you have key identifying information including some combination of: Name, Website, LinkedIn URL, and Location.
Step 2: Access the Company and Person datasets
Setup your PDL account and access the Company and Person datasets through our APIs or bulk data files. For this recipe, we recommend the Company Data License and Person Data License, which allow you to efficiently process person and company profiles at scale.
Step 3: Match your target company list to PDL Company records
Use the identifying information you defined in Step 1 to find matching records in the Company Dataset.
Step 4: Define your criteria for Middle Management roles
Define the specific person-level attributes that represent the most relevant middle management employees to track across your target companies. We recommend considering attributes like: Job Title, Seniority, Role and Industry.
Step 5: Define your criteria for benchmark companies
Define the specific company-level attributes that represent the companies to benchmark your target companies against. We recommend considering attributes like: Industry, Revenue, Funding, Location, and Headcount.
Step 6: Build a list of Benchmark Companies
Using your company definition from Step 3, filter the PDL Company Dataset to generate a list of benchmark company records. These records will serve as the set of companies you will evaluate your target companies against.
Note: If you already have a set of benchmark companies, repeat Step 3 above to directly match your list of benchmark companies to PDL company records instead of filtering down the Company Dataset.
Step 7: Build a Target Experience Array for each target company
Using your role definition from Step 4 above and the target company records from Step 3, filter the Person Dataset to identify the set of person records representing middle management employees at each of your target companies.
For each profile, extract the Work Experience history from the record and add it to the Target Experience Array for the appropriate target company.
Step 8: Build the Benchmark Experience Array
Repeat Step 7 above, using your role definition for Step 4 above and the benchmark company records from Step 6.
For each profile, extract the Work Experience and aggregate them in a single Benchmark Experience Array.
Step 9: Calculate middle management quality scores for your target companies
Using the Target Experience Arrays from Step 7 and the Benchmark Experience Array from Step 8, score the middle management quality at each of your target companies. See the next section for some example scoring methodologies you can use.
Scoring Middle Management Quality
In this section, we’ll share three examples of the types of signals and quality scores you can generate using the experience arrays built in the recipe above:
Role / Industry Expertise Scores
Comparable Company Experience Score
“Stick Around Score” / Tenure Stability Score
These scoring methodologies are based on examples from workflows we’ve seen our customers use in the past. Feel free to use these as inspiration for building and customizing your own approaches.
Role / Industry Expertise Score
What it does: Evaluates the current role of middle managers against their cumulative experience in similar roles and industries
Key Insights: Identify if middle managers possess deep, relevant experience in their specific roles and within the company’s history
Example Calculations:
Compare overlap of key attributes from previous roles compared to current role (evaluated for targets and benchmarks). Suggested attributes: job title, role, seniority, location, years of experience, and salary
Compare similarity of previous work experience descriptions to current role summary. Suggested attributes: work experience summaries, leveraging LLMs/natural language techniques to automate and scale processing
Comparable Company Experience Score
What it does: Compares middle managers’ past experiences across key company attributes like size, industry and funding stage
Key Insights: Evaluate whether managers are accustomed to working in similar environments to estimate the risk of cultural or operational misalignment
Example Calculations:
Compare overlap of previous employers across key attributes compared to current employer (evaluated for targets and benchmarks). Suggested attributes: previous employer headcount, industry, funding stage, and revenue
Search for growth or impact signals at previous employers during tenures. Suggested attributes: fundraising events or historical headcount growth / churn events
“Stick Around Score” / Tenure Stability
What it does: Measures a middle manager’s average tenure at past companies and compares it to their current company’s employee tenure.
Key Insights: Assess likelihood that managers will stay long enough to drive success
Example Calculations:
Compare statistics on historical tenure at target companies vs benchmark companies. Suggested attributes: use start and end dates to estimate tenure, and current role start date to estimate flight risk
Compare headcount and growth/churn rates at target companies vs benchmark companies. Suggested attributes: PDL’s precalculated headcount metrics or build your own custom insights signals using monthly role and seniority breakdowns
Chef's Recommendations
PDL Databricks Integration
For this recipe, we highly recommend using a tool like Databricks, which allows you to run analytics, build custom dashboards and evaluate data at scale without managing any data or cloud infrastructure. For PDL customers, we offer a direct integration through the Databricks Marketplace, allowing us to directly deliver monthly data refreshes right within your Databricks Environment.
Our integration is available through Delta Share where you can also access supporting data assets including dashboards, notebooks and ML models. Learn more at: Integrations - Databricks
Custom Insights Professional Service
When working with headcount data at scale, we recommend taking advantage of our Custom Insights Professional Service. This service allows you to extend your PDL company data with tailored, precalculated headcount breakdowns and insights, based on customized personas you can define. For example, you can use your middle management role definition (from Step 4 the recipe above), which we will then use to precompute headcounts and growth churn metrics for these employees across all your company records. Because we build these headcounts from the ground up using our Person dataset, this also means you have access to every single record behind each metric.
We take our headcount data very seriously, and we even wrote a deep dive whitepaper on our methodology, which you can read here: Exploring Employee Headcount Accuracy: Pitfalls and Solutions.
Still Need Help?
Check out our developer documentation for details on data schemas, stats, code examples and more!
Check out our help desk / reach out to us at support@peopledatalabs.com
Reach out to your customer success representative (Enterprise customers)