How Does Your Engineering Organization Stack Up? The 2021 ICONIQ Engineering Efficiency Report
Compared to sales and marketing, engineering is a function that often gets overlooked when it comes to reporting. As a result, analytics and benchmarking for the function are lacking. However, R&D is increasingly becoming a bigger line item in spend for software companies and a key differentiator.
In our previous post on engineering reporting, we laid out one framework for tracking and reporting the performance of your engineering organization. But now that you have your metrics, how do you know what “strong performance” looks like?
Earlier this year, we asked CTOs and CFOs across select companies in ICONIQ Growth’s portfolio to complete a survey on a variety of metrics and topics related to their engineering organizations. Below is a summary of the results. For more, please see our full interactive report here.
- R&D Spend
Surveyed organizations indicate that software companies tend to invest significantly in R&D in the early stages of growth, with companies under $50M in revenue spending significantly more on R&D than the revenue they are generating. After reaching ~$100M in revenue, most companies tend to spend around 30% of their top-line on R&D investments — a trend that continues even as companies mature past $300M in revenue.
R&D headcount also tends to make up around 30% of the total organization, with slight spikes in the earlier and later stages of company growth.
Companies across all revenue buckets spend around 75–80% of their R&D budgets on headcount, around 10% on infrastructure, and the remainder on security and other non-people investments. This translates to an implied R&D spend per R&D FTE of around $200K.
2. Developer Productivity
As an engineering organization grows, different types of questions and challenges emerge around the organization’s investments. That’s why it’s critical to have a framework in place that allows the company to discuss and prioritize engineering investments in a way that makes sense for the engineering team and others outside that function. We recommend the below approach to categorize and track engineering investment.
On average, surveyed organizations allocate ~83% of engineering capacity to “elective investments” and ~17% to “Keep the Lights On” (KTLO) activities. Within the “elective investments” bucket, ~60% of engineering capacity goes toward building new capabilities, ~25% to quality improvements, and ~15% to internal productivity.
Just as sales teams measure quotas and ramp time, it’s important for the engineering organization to measure developer productivity. Top efficiency metrics tracked by surveyed companies include service uptime, number of incidents, number of critical defects, PR to release time, % of code delivered versus committed, and many others.
3. Engineering Team Structure
Engineering teams are typically organized by technology, product, a hybrid of both, or in a matrix model.
The majority of the companies ICONIQ Growth surveyed tend to have their engineering teams structured by product, with average team size ranging from 5 to 8 FTEs (implying an average of 25 engineering teams).
One of the key benchmarks we often get asked by our CTOs is about engineer ratios. In this study, we saw the following:
As companies grow, engineering team makeup also tends to shift toward having more back-end and full-stack engineers perhaps due to the increased requirements around scalability and reliability.
4. Organization Health
Historically, software companies have struggled with engineering diversity. Today, the companies we surveyed count on average 20% of their engineers as female . Across the participating companies in this study, teams reported somewhere between 10–50% diversity among their engineers; however, it is important to note that these numbers includes Asian and South Asian engineers and the percentage of Black, Indigenous & People of Color (BIPOC) engineers is likely much lower.
Diverse leadership it appears attracts diverse talent; companies with a diverse founder and/or CEO were also seen to have a significantly greater percentage of BIPOC and female employees.
In terms of tenure, engineers across surveyed companies stayed for around 2 years, with this number increasing slightly for larger companies. Most companies also start tracking developer job satisfaction after reaching $50M in revenue through either custom employee surveys or Culture Amp surveys.
Does this mean one fifth of engineers are female, i.e. 20 females out of 100 total, or the number of female engineers is one fifth of the number of male engineers, i.e. 20 females and 100 males and 120 total?
You can dive into the rest of the report in our interactive version here, and we welcome any feedback or questions on this crucial topic!