Over the past couple of weeks, a number of founders and startup teams have reached out to us to help them think through how they should be setting their 2025 growth and spend targets. As part of this early annual planning work, we collected “whisper numbers” for 2025 where we asked our companies for very rough initial estimates for next year. These whisper numbers signaled a new chapter: growth is reaccelerating, but this time with a sharp focus on maintaining — and in many cases improving — operating efficiency. While both the whisper numbers and our latest Q324 Flash update signal both a reacceleration in growth and increased efficiency, that still leaves companies with the question of how to actually allocate their operational spend.
While we always publish our “Four Vital Signs” benchmarks (growth, burn, churn, and efficiency) on our Scale Studio website, when it comes to annual planning it’s helpful to go a level deeper, especially when it comes to planning OpEx and spend for the year. Operational planning is new territory for most founders, even more experienced ones. How much should you spend on S&M? How do you balance that with product development and R&D? To that end, we thought we’d publish some of the more detailed OpEx benchmarks we’ve been sharing with companies in our portfolio this annual planning season. Hopefully this data is useful to have in your toolkit as you wrap up initial 2025 forecasts and continue to adjust plans throughout the year. Initial signs are optimistic, so we’re hopeful we have a good year ahead of us.
Key takeaways:
- Stage matters: Early-stage companies prioritize R&D to find product-market fit; scaling companies emphasize S&M to drive growth.
- Adjust to context: Median spending benchmarks for OpEx provide a solid starting point, but adjust based on your specific situation.
- A new way forward: Growth is back, but efficiency remains a central theme for 2025 planning.
How do companies split up OpEx across S&M, R&D, and G&A?
One of the clearest trends when we look at the data is how spend in each of the three OpEx categories changes over time. Intuitively, early-stage startups spend a lot more on R&D relative to S&M or G&A, which makes sense given they’re focused on finding product-market fit and the team is typically a few engineers and maybe one business / sales person. As startups begin to scale, however, we see more and more money being poured into S&M relative to R&D. Again, this makes sense. Once a startup has found PMF, the focus turns to scaling and getting the product in the hands of more and more customers. Be it through PLG, marketing, or just good ol’ enterprise sales, a lot of the incremental dollars being spent start going towards S&M to drive growth.
For instance, at the $0-1M ARR range, S&M might account for just 25% of OpEx. But by the time a company surpasses $25M ARR, S&M often represents over 50% of OpEx. At $0-1M ARR your priority is building the product, and that’s where all your spend should go. Once you’re at $25M ARR, you’re going to have a fully-scaled GTM machine, taking up a lot more resources than the product work you’re still doing.
But how much should you actually spend?
The chart above is helpful in better understanding how to allocate your budget across the three OpEx categories, but what does this mean in actual $ amounts you should be spending? That’s obviously going to be relative to a company’s size, but again we can turn to the benchmarks to give us a rough indication of how much the median company spends on OpEx at any given ARR level.
The numbers here show, as a percentage of their revenue, how much the median tech startup spends on S&M, R&D, and G&A by ARR band. You can see the same broad trend we saw earlier, that as companies grow in scale they start allocating more to S&M than R&D (and G&A goes down as well).
For example, a median company projecting $10M GAAP revenue in 2025 might plan to allocate $6.8M to S&M, $4.7M to R&D, and $2.9M to G&A. Remember, these benchmarks are medians—not absolutes. One of our companies building in the healthcare space is spending more than 40% of their OpEx on R&D even as they begin to scale, due to the need for specific regulatory approvals. Likewise, one of our companies building their own foundation model is spending more than half their OpEx on R&D. The benchmarks should be a rough guide. Use them to model scenarios, explore revenue-impact relationships, and adapt as forecasts evolve throughout the year.
One quick note about our benchmarks. The data above and below comes from Scale Studio, our dataset of 1,000+ private companies and 10,000+ quarterly data points from the mid-2010s to present. Our belief is that, in aggregate, this data portrays the profile of a typical tech startup. There are always outliers and other factors that could shift the short-term benchmarks (Covid, tech recession, AI boom, etc.) and not all companies are the same (higher regulatory spend, legal bills, competitive S&M market, etc.), but that should all wash out in aggregate and the medians above prove a great thermometer for if your OpEx allocation is in the right ballpark.
