Note: this post assumes a working understanding of basic SaaS metrics/ratios/etc. If you need a refresher, check out SaaS Metrics 2.0 & SaaS P&L Refresher.
In light of slowing growth rates, cratering valuation multiples, and a steep decline in venture funding, startups are now prioritizing “efficient growth” over 2020-2021’s prevailing mantra of “growth at all costs”. Consequently, metrics/ratios that quantify what constitutes “efficient growth” (and what doesn’t) have become first class citizens in the board room and in investor pitches/fundraising conversations.
By now, most investors/operators have heard of the Efficiency Score (or Burn Multiple, if you flip the numerator and denominator) and the Magic Number as measures of capital & sales efficiency respectively (note: if you need a reminder, the equations are included in the Appendix of this piece). These are objectively great metrics that have helped many founders/operators better understand and operate their businesses, but I don’t think they paint the full picture when it comes to growth efficiency. I’d like to introduce another metric that deserves a place in the SaaS efficiency benchmarking lexicon: Operating Yield.
Operating Yield compares a company’s production of Net New ARR (NNARR) to its total expenses (cost of goods sold & operating expenses) for a given time period (monthly / quarterly / annually). In other words, the output states “for every dollar of cost to run the business, we generated x% of that figure in Net New ARR”.
If Operating Yield is high, it means that a business is adding a lot of Net New ARR/growing quickly relative to the costs of running the business (good)
If Operating Yield is low, it means that growth/Net New ARR is low relative to the costs of running the business (bad)
Here’s an example of calculating Operating Yield using a made-up SaaS company:
I like this metric for several reasons:
You can use it on any/every SaaS business — A key limitation of the Efficiency Score/Burn Multiple is that you can’t use it on profitable companies, and therefore can't compare profitable companies to unprofitable ones. It also becomes less relevant/useful as companies scale beyond early-stage. Operating Yield can be used on any SaaS business, regardless of its profitability, and remains a valuable measure of growth efficiency throughout the company’s lifetime.
It forces apples-to-apples standardization and is hard to game — Measures like the Magic Number can be calculated in various ways that bias/skew its output. A couple specific examples of common calculation disparities in the Magic Number include: gross margin weightingvs. not gross margin weighting Net New ARR in the numerator, and including all S&M costs vs. excluding "non-direct" S&M costs in the denominator. In contrast, by comparing NNARR to Total Expenses, Operating Yield incorporates all costs regardless of where/how they're categorized on the P&L and eliminates most types of gaming/bias.
It’s the best all-around abstraction for growth efficiency — One helpful framework for understanding SaaS businesses is to think of them as factories that produce boxes (customer contracts) that print money at regular intervals. In this metaphor, Operating Yield measures the relationship between how many boxes the factory produces & how much money those boxes are printing annually (NNARR), against the overall cost to run the factory (Total Expenses). In my opinion, those are the two best variables to compare to understand growth efficiency. The Burn Multiple helps answer the question “am I burning too much?”, and the Magic Number helps answer “is my S&M spend efficient?” — but neither answer “is my business as a whole efficiently growing (i.e. producing Net New ARR)?” Operating Yield does just that!
For a deeper dive into the “SaaS-business-as-annuity-factory” framework and why it fits so well with Operating Yield, please scroll down to the “Appendix — Annuity Factories” section.
Public Comps & Benchmarks
Now that we’ve established what Operating Yield is & why its valuable, the question remains: what is “good” operating yield vs. “bad”? To answer that question, I asked my friends over at Meritech — who maintain a high quality set of public comps & benchmarking data here — if they could provide an Operating Yield benchmarking dataset for the publicly traded cloud businesses they track. They graciously obliged:
The above data is as of Calendar Q4 2022 and measures the Last 12 Months (LTM) NNARR and Total Expenses for each company. We used LTM instead of measuring a single quarter in order to filter out potential seasonality and/or one-time spikes/irregularities in any of the businesses.
As you can see, the data is pretty clear-cut, and using it we can segment the results into the following buckets:
As of 5/10/2023, public companies with >30% LTM Operating Yield traded at an average of 10.0x EV/ARR, companies between 15-30% traded at 6.3x EV/ARR, and companies <15% traded at 4.2x EV/ARR.
I’ll caveat by saying that the above benchmarks are based off of one point in time (LTM as of Q4 2022), so consider these loosely held rules of thumb. I encourage other firms with lots of data to create their own Operating Yield benchmarks to compare to this and I can update this piece to include those here or link to your own post.
Update: After initially posting this piece, my friend Arye Lifshitz provided me some incremental analysis illustrating that at least as of 5/10/2023, Operating Yield has a higher correlation to valuation multiple for public SaaS businesses than Y/Y ARR growth" or “Rule of X” (Growth + FCF Margin).
The 3-legged Stool of Efficiency
With Magic Number, Efficiency Score, and now Operating Yield at our disposal, our growth efficiency toolkit is complete.
Magic Number: measures Net New ARRproduction relative to S&M
Efficiency Score: measures Net New ARR production relative to Burn
Operating Yield: measures Net New ARR production relative to Total Expenses
Magic Number measures sales efficiency, Efficiency Score measures capital efficiency, and Operating Yield measures operating efficiency. Together they create an analysis trifecta that gives investors/operators a holistic view into a company’s efficiency from $5M of ARR up to and through $1B+ of ARR.
I encourage investors, founders, and operators to try the measure out on their own historical dataand see what you find. In practice, I think founders/operators can use Operating Yield as a sort of governor / sanity check on the overall size & cost structure of their businesses. If you’re consistently running at a <15% Operating Yield, either 1) you have growth problems (PMF issues, running out of TAM, Act 2 isn't working, competitor is beating you H2H, etc.), or 2) your cost structure is bloated relative to your stage and/or the normalized output of your growth engine. And if you’re running a business at a >40% Operating Yield, please email me at firstname.lastname@example.org so that I can invest in your business.
Appendix — Annuity Factories
One way to think of SaaS companies is as factories that build/produce money-printing boxes.
Factory line workers (sales & marketing) use tools/machinery (the product) to build boxes that print a certain amount of money on a regular cadence (customer contracts). Once built, the boxes require fuel/power to operate (direct COGS). Some maintenance workers (customer support) keep the boxes running & fix issues while others (customer success/account management) look for ways to increase the amount of money coming out of each box. Also in the factory resides scientists looking for ways to make machinery that can produce bigger & better boxes (research & development), as well as all the other staff that help maintain the factory itself (general & administrative). If the factory is well-run, over time the stockpile of money-printing boxes will grow and provide an increasing & recurring stream of cash that the factory owners can use to pay the factory’s expenses & reinvest in the factory’s underlying components.
It’s not a perfect or exhaustive metaphor (e.g. it excludes the idea of competition), but I think the factory framework does a good job of making the functions of a software business tangible, and importantly emphasizes Net New ARR as the core unit of value for SaaS companies.
Bringing this back to Operating Yield — if you were to invest in or buy a factory like this, how would you evaluate it & what questions would you ask?
At first you’d want to learn the basics: how much money is the current money-printer stockpile producing each year? how much money does the average box print? how long does the average box last before breaking down? how high quality is the tooling/machinery used to build the boxes? etc. etc.
Ultimately though, if you were investing with a long-term time horizon, your evaluation of the the factory would rest on the two most important variables which are:
The quality, quantity, & size of money-printing boxes it can produce (Net New ARR), as over time the size of the money-printing-box-stockpile will be determined more by the rate of new box production than the number currently operational boxes.
The overall cost to run the factory (Total Expenses), as this will determine what kind of profits (and subsequently ROI) you can expect from your investment.
These two variables are, of course, the variables that make up Operating Yield.
Appendix — Burn Multiple & Magic Number
P.S. — Thank you to Buck
I also want to state clearly in this post that I can’t take credit for creating the Operating Yield metric — I first saw the SaaS mensch Buck mention it on Twitter, and for all I know he could’ve learned it from someone / somewhere else too.
So thanks Buck for introducing me to this awesome measure that hopefully becomes commonplace in the coming months/years.
Thank you to Lucas Oliveira for the help thinking through this piece and to the Meritech team (Anthony DeCamillo, Alex Clayton, etc.) for providing awesome data for this analysis.
You can also offset costs to compare Q-1 total expenses to Q0 NNARR, as many do with the net magic number. I find it simpler to keep costs & NNARR aligned in the same quarter.
Hence why you don’t see it on Iconiq’s Enterprise Five benchmarking report of even Bessemer’s own Scaling to $100M ARR benchmarking report
Or at least until it stops growing for good, in which case all “growth efficiency” analysis is useless anyways
This means multiplying Net New ARR by Gross Margins. This is the correct way to calculate Magic Number in order to normalize for different gross margin profiles.
Another common way to game this measure (and many others) is by moving expenses that should be in COGS into Operating Expenses. The most popular ways to do this are over-allocating customer support & customer success into Opex.
Non-direct costs typically mean things not directly related to a sale, e.g. brand marketing, sales executive (CMO, VP of Sales, etc.) compensation, etc.
One big area that Operating Yield doesn’t account for is cash conversion and capital intensity. We’re assuming that we’re comparing businesses with similar (low) Capex intensity like most software companies have, and if you’re dealing with companies with significant Capex + rapid growth, you should caveat/adjust accordingly.
Technically Gross Margin weighted NNARR
Enterprising investors may want to track incremental operating yield (calculated the same way you would incremental margins) to understand the trending operating yield for a business and have it be less of a backward looking metric
Or, more likely, some combination of the two
Or ya know, buy a yacht or something
It also illustrates how interconnected these functions are and how just a couple variables going wrong can screw up the whole system. For example: if the product (box-building machinery) sucks, it will be harder to sign customer contracts (build boxes), and the customers will be more likely to churn regardless of the quality of the support team (boxes will have short operational lifetimes before they stop working).
Many great growth investors focus on the second derivative of ARR growth, i.e. the sequential growth in net new ARR, as the #1 / most important measure when evaluating growth-stage software businesses
using this to evaluate a portfolio of African fintechs to help them avoid hard landings! Thanks