How to compare SaaS companies

I’d like to introduce a few measures that I and others have touched on to compare SaaS companies. Many of the traditional measures fail because SaaS companies look pretty terrible in their infancy and middle age if you look at them by P/E or similar measures. Many of our companies are in the, “Land and Expand” part of their growth cycle. How can we make sure they are actually creating value and not just growing with no hope of having a feasible business at the end of it all? How do we compare a company growing 60% with no earning, burning cash with a company growing 30% and earning cash? I’ll introduce a few ideas below and what goes into them.

The first is Efficiency of Capital. Simply put you take Revenue/(equity+debt). In english this basically is looking at how much capital has it taken to generate the revenue for that year. For companies that have successfully IPO’d and lived their efficiency of capital ratio is >0.6. So they have created 60 cents of revenue for each dollar of capital invested. Shopify’s efficiency of capital is currently lower than usual because they have raised money but not spent it.

SHOP 0.6 OKTA 0.7 MongoDB 0.37 AYX 0.45

The second is Efficiency Score. Revenue growth + (free cash flow/revenue) . This measure allows us to correct for companies that are growing like mad but burning cash with companies that are growing but actually cash flow positive. Is SHOP’s growth really all that good? (yes it is).

SHOP 0.72 OKTA 0.6 MongoDB 0.21 AYX 0.65

Finally we have the dollar revenue retention rate . Each company calls this something a little different. I like to think of this as how sticky the company’s revenue is. I think DRR can be an indirect proxy for moat and you can follow DRR with some context to see if a companies moat is increasing or decreasing. This is calculated by taking the beginning of year revenue + upgrades from existing customers – churn – downgrades from existing customers divided by beginning of year revenue. So basically are your existing customers in aggregate spending more with you each year (>100% DRR) or less (<100%DRR). Not all companies give us the DRR and each calculates it a bit different and even may call it something differently. For example. SHOP just tells us it is >100%. OKTA gave us 123% one time and then stopped reporting. MongoDB says they have an “attractive” DRR and AYX tells us each quarter which hovers around 131%.

So, why is land and expand so powerful for these companies if they only receive 60 cents for each dollar they spend. The thing is they get that 60 cents every year as most of that revenue is recurring and if their DRR is high then they get more than 60 cents every year. So most of these companies (except MongoDB) are earning a profit on a customer after one year. This business model is doubly powerful in that customers are loath to switch back to a large occasional capital outlay because it screws up their accounting and planning. This is one reason why I’m excited about MongoDB despite their relatively poor numbers. Their atlas program will make switching to another database very very difficult. I’d be interested in Nutanix’s numbers but have run out of energy for today.



So most of these companies (except MongoDB) are earning a profit on a customer after one year.

Will that change with MongoDB Atlas?

Denny Schlesinger

1 Like

I think so. They are running these servers on either Azure, AWS, or Google cloud. All they have to do is price their managed service higher than basic cloud machine use from those companies. As a side note all the above companies also have a nosql service that is a competitor to Atlas. I don’t know enough to know how they compare.

1 Like

It is impossible for us to assess at this time if Mongo has sufficient pricing power to make money per customer, or if that is a matter of scaling, as so many things are in SaaS.

In regard to competitive databases what we know is only one NoSQL makes it into the top 10 on the activities list. That list, that we have posted to on NPI (do not have a link w me) has Oracle owning the top 2 databases, both SQL, and only Mongo making the top 10 otherwise.

Denny cited to another measure taken from a go to developer site, with a survey of 100,000 developers (I believe Mongo also cited to this) the for the second year in a row Mongo was the DB most wanted to be used be developers.

I cited to a Linkedin article, written by a developer (who is also a Mongo consultant), who gives the reasons why he thinks Mongo is an excellent investment opportunity from a developer perspective. This one has to be taken with a grain of salt, but he does go into detail.

We do have Mongo management citing an AWS executive stating that “Mongo is kicking our butt.”

As a lawyer the latter two sources, no so admissible. But I do believe the quote is credible from AWS and goes hand in hand with the other cited data.

What we don’t now is if Amazon will put in the effort necessary to catch up to Mongo, or do something like they did with SHOP. One would think that hosting online merchants would be a core business for Amazon, and yet they let that one go.

We will see. But this is all the evidence we really have at present other than the numbers.

One number that appears excellent, but not so much when you look at it is they have 20 customers with more than $1 million in annual spend. But 20 out of 5700 = .35%. A much more encouraging number is the number of 6 figure customers they have on Atlas. That is money they were otherwise just giving away in the past and Atlas looks like it may be extremely disruptive offering.

Finally, Mongo’s sales force is very inexperienced at present and being put together. Management indicated that they were gaining traction going after bigger deals, going higher up in the organization, and they would be focusing on larger deals. So perhaps that .35% number will go up (almost certainly), and as Atlas becomes more mature, that it continues to grow like mad.

These things will increase Mongo’s cash efficiency.

Well, I hope so anyways :wink:



Look and thou shall find (no cool lithographic font used). This article has some significant and quite relevant statistics about Mongo,including that it indicates that its past customers are profitable, and the land and grab attributes that are quite impressive.

I will let you guys analyze and let me know.

Revenue curve is rather nice, and with such high growth, the ROI numbers going forward may be quite nice. But I will let you guys make any analysis in regard that may be worth the time to print.



ARR had me stumped but the 10-Q came to the rescue

We monitor annualized recurring revenue (“ARR”) to help us measure our subscription performance. We define ARR as the subscription revenue we would contractually expect to receive from customers over the following 12 months assuming no increases or reductions in their subscriptions. ARR excludes self service products, including MongoDB Atlas not sold on a commitment basis, and professional services. For self-service customers, we measure their annualized monthly recurring revenue (“MRR”), which is calculated by annualizing their usage of our self-serve products in the prior 30 days and assuming no increases or reductions in their usage. The number of customers with $100,000 or greater in ARR and annualized MRR was 110, 164, 246 and 320 as of January 31, 2015, 2016 and 2017 and October 31, 2017, respectively. Our ability to increase sales to existing customers will depend on a number of factors, including customers’ satisfaction or dissatisfaction with our products and services, competition, pricing, economic conditions or overall changes in our customers’ spending levels.…

It’s not a secret for sales people that your best prospects are your existing customers (just like your best buys are your best performing stocks). Software, specially enterprise software, is now so complex that you have to deal with it in bite sized chunks, what maybe teams of ten developers can manage at a time. Moving a Fortune 100 company to a new database architecture takes time and if these customers are coming back for more they must like what they see. It is hard to get a better recommendation.

The metric I would watch for is the relation of profit or loss to revenue growth. So far increased revenue has also increased net loss and that needs to change. Even when gross margin is increasing Selling, General and Administrative Expenses (SG&A) can kill profits. This often happens with aggressive expansion plans. But, IIRC, Geoffrey Moore says that during the Tornado just push product out the door. Land grab is designed to keep competitors from reaching critical mass, insurance that you’ll be the one winning most of the market.

Denny Schlesinger


We do have Mongo management citing an AWS executive stating that “Mongo is kicking our butt.”

What we don’t now is if Amazon will put in the effort necessary to catch up to Mongo…

These are different businesses. And, AWS already supports MongoDB:…

AWS enables you to set up the infrastructure to support MongoDB deployment in a flexible, scalable, and cost-effective manner on the AWS Cloud.



I have been looking at the gross margin, rev growth, and SGA. I have also looked at outstanding shares or EPS to see how dilution plays a role on share price or value for shareholders. I am not familiar with the Tornado theory, but have seen you post about it a couple times and am intrigued. The market is blowing in high growth SaaS companies favor atm.

I honestly didn’t understand everything Ethan said, or how to get some of the data points to come to the numbers to compare SaaS companies.

So far I just looked at SeekingAlpha key data and financials to build a sheet on SaaS companies. Then I look at earning calls before I buy/sell or gain more conviction in a company because SA doesn’t have the most recent numbers in their key data or financials usually.


I am not familiar with the Tornado theory, but have seen you post about it a couple times and am intrigued.

Regis McKenna is (or was) the preeminent Silicon Valley marketing guru. After leaving Regis McKenna Geoffrey Moore started The Chasm Group. Eventually he tried to apply his marketing expertise to investing in high tech and wrote The Gorilla Game which relies heavily on his marketing background. The basis for it all is the Technology Adoption Life Cycle (TALC) depicted graphically here…

The Chasm is where most innovation dies. If the technology crosses the chasm it goes through various stages. When the early majority (pragmatists) start buying the product or service flies off the shelves and companies experience 100% annual revenue increases for a couple of years. During the Tornado the market decides which is the winner. There is no telling beforehand who the winner will be (this has been corroborated by the Science of Complexity). Moore advised “buy the basket and sell the losers.” I think it is better to wait until the winner is established before buying, you’ll still have lots of years of great growth and will run less risk.

If you are interested read The Gorilla Game

Denny Schlesinger

Regis McKenna

Geoffrey Moore

Chasm Group

The Gorilla Game, Revised Edition: Picking Winners in High Technology Kindle Edition by Geoffrey Moore…


Is mongoDB qualifies as SaaS?

MongoDB currently gets 10% of its revenues from Atlas (growing from a very small age at 500% YoY).

As such, not very SaaS. Nevertheless it is software. It’s software can be run on premise, off premise, or hybrid. As such, recurring revenue like SaaS, and much more SaaS than Nutanix is, whom has little “recurring” revenue and is instead one time payment for each server/node that goes out the door.

Atlas does appear to be possibly a quite disruptive product that may have been underestimated by Mongo management. Remains to be seen how growth goes as the numbers get larger. Nevertheless, even at now 10% of revenues, they are still seeing many 6 figure annual sales to customers with Atlas. MAny of those are to the enterprise customers expanding their usage without wanting to expand licensing on enterprise (perhaps for development instead of deployment efforts, speculation), but too early to really tell anything. Early results often do not correlate to later results when the low hanging fruit is done, but it does look quite promising at present.

So the answer is, Mongo is not much of a SaaS company per Se, but its recurring revenue business model, nevertheless, does appear to have the same type of attributes, as do so many other software companies.


1 Like

Atlas does appear to be possibly a quite disruptive product that may have been underestimated by Mongo management.

The enterprises have embraced cloud. So offering their database on cloud is a natural extension. I understand many folks don’t differentiate the difference between PaaS and SaaS, but to me Atlas is a PaaS and not SaaS.

but its recurring revenue business model, nevertheless, does appear to have the same type of attributes, as do so many other software companies.

LOL. Windows, Oracle database all had recurring revenue… whatever…

PaaS, SaaS…between the two PaaS does seem superior. That is what SHOP is after all. A platform to build on. Not real sure if there is any difference those investment wise in regard.

In the end, they are software delivered via the cloud.

Nutanix, yes, has a different model, so worth noting.

Whereas, Oracle, Microsoft, et al, have definitely proven the worth of any dominant software model.


At the end the real question is not whether Mongo is a SaaS or PaaS, etc but whether their software is something that enterprises will continue to buy and use. Do they receive sufficient value and for that what are they willing to pay for.

But the title of the post is very misleading and the valuation is even more suspect, IMO.