As companies get larger, engineering makes up a larger percentage of R&D headcount. Product management tends to get smaller as a percentage of overall R&D headcount, suggesting that the number of engineers per product manager increases as companies get larger.
What percentage of an R&D department should be engineering? At a 1,000 person company, what is the right ratio of product managers per engineer?
Let’s dive into the data to find out.
We analyzed data from ~233,000 employees in R&D across 4,195 companies in Pave’s dataset, and looked at the relative proportions of engineers to product managers, designers, and data roles.
The data shows a decrease in product management ratios as companies scale—from 13% of R&D headcount in smaller companies (1-100 employees) to 7% in large enterprises (3,000+).
There are a couple hypotheses around why this may be. The compression in product management as a percentage of R&D can be due to:
At smaller companies, teams are smaller—the R&D team likely has a handful of engineers and a product manager and a designer, maybe a data scientist. As the organization grows, the R&D team is built out more to have PMs that manage more engineers individually.
As companies mature, they often shift from product-led growth to enterprise sales models, which can influence PM staffing ratios.
The gross number of product managers is also expected to increase as companies get larger because they are likely releasing more products at scale. A bigger company naturally has more products, and therefore needs more product managers, and more layers of hierarchy within the Product org.
Ultimately, there is no right or wrong answer. The optimal ratio of product managers per engineer depends on your company, industry, and the specifics of your product.
{{mid-cta}}
As companies get larger, engineering headcount makes up a larger percentage of overall R&D.
While early stage startups typically hire engineers with a generalist skill set, larger organizations tend to build more specialized infrastructure, platform, and tooling teams that can support broader engineering efforts efficiently.
On its own, this isn't a surprising insight, but the recent spike in machine learning demand complicates what is classified as engineering.
Companies are struggling with how to classify and compensate AI/ML talent within their existing structures. Some are keeping these engineers within general software engineering job families, while others are creating separate tracks—and this decision has significant implications for compensation benchmarking.
As engineering organizations become more specialized, job architectures need to evolve in pace with company growth. For AI/ML specifically, there's an ongoing debate about whether it should be classified under software engineering or data science.
It’ll be interesting to see where AI/ML lands in the modern R&D org structure. Will it sit within engineering, data, or its own category altogether? Only time will tell.
Data Science maintains a relatively stable representation of 4-6% across company sizes, but perhaps not for long.
There's significant evolution happening where the line between data engineering and software engineering is blurring. Some companies place data engineering within their engineering org, while others keep it within data—and this decision often impacts compensation structures and career paths.
This evolving landscape suggests organizations need real-time decision-making frameworks that can help them accommodate rapid change while maintaining consistent compensation principles.
Ultimately, the right structure for your company depends on numerous specific factors, including:
Companies need current market data to inform their organizational structure and compensation strategies. The key is building flexible frameworks that can adapt to change while maintaining consistent principles for organizational design and talent management.
View cash and equity benchmarks from 8,500+ companies with Pave's free Market Data product.