Churn is probably one of the most documented SaaS metrics. The aim of this article is not to define and speak about the general concept of churn (you can find plenty of outstanding resources about that) but to analyse 4 SaaS benchmarks and see what they say about real churn figures.

The 4 benchmarks we’ll analyse are: Pacific Crest Survey, Open startups, Totango annual SaaS metrics report and Groove SaaS small business survey (if you know more of them please => @clemnt)

But before digging into the actual numbers I would like to stress on a very important point: before you compare churn numbers you have to clearly define which churn you’re talking about.

User or revenue churn? Annual or monthly? Gross or net?

When you see a churn figure you have to ask yourself whether:

• It’s revenue or user churn
• It’s annual or monthly churn
• It’s gross or net churn

Revenue or user: pretty straightforward, in one case you measure the revenue lost from a cohort during a specific time range and in the other the number of paying customers lost from a cohort during that period of time.

Annual or monthly : meaning the churn is measured for a year or a month period. Since both are very often used here is a table with monthly churn and its annualised equivalent. I just use the formula: Annual Churn = 1 — ( 1 — monthly churn )^12.

Be careful though as you shouldn’t mix monthly plans churn with annual plans churn (different beasts).

Gross or Net churn: probably the trickiest of the 3. Gross revenue churn doesn’t take into account revenue expansion (= some users of your cohort can actually end up paying more at the end than at the beginning of the period considered, because they have upgraded their account) while net churn does.

So it’s totally possible to have a 6% monthly gross churn and a 2% monthly net churn. So if you see a 2% revenue churn alone, without context, it’s hard to compare it with other figures. Here is a good post on Chartmogul explaining it. And yes with net churn it’s possible to have a negative churn rate.

Now that we’ve made that clear let’s start.

TL;DR

The “general observations about customer account churn rates by segment” that Tomasz Tunguz shared are confirmed by ¾ of the benchmarks analysed. We cannot draw any meaningful conclusion from Totango survey, hence the ¾.

Couple of comments after receiving some messages / DM on Twitter:

• of course the industry you are in can play a role. Especially in industries where the competition is not as fierce as in others and where customers have less choice and as consequence cannot change easily your churn rate can be lower.
• that said there is still a very strong correlation between the segment you’re going after and churn. Higher ACV generally means real sales / field sales, account managers, longer term contracts etc… and logically lower churn.
• I’ve received a couple of messages from startups saying that they are on the SMB segment (with ACV <$1k) and have monthly user churn between 1% and 1.5%. So yes it’s totally possible, they are doing an awesome job :-) The 3% - 7% range is where most of the companies targeting this segment are, but you can be far lower (and that’s awesome).

Pacific Crest Survey

About this benchmark

Conducted by David Skok and Matrix Partners the Pacific Crest Survey is the most complete and detailed benchmark in the SaaS industry. You won’t find better than this currently (available publicly at least). A must read.

Survey participants

The benchmark is based on the answers from 306 companies:

• $4MM median revenues, but nearly 50 companies with >$25MM and 80 with <$1MM
• 46 median full-time employees
• 284 median customers, with 25% having >1K customers
• $21K median annual contract value (ACV), with 30% below $5K and 20% above $100K
• Good mix of sales channels including field sales, inside sales and mixed distribution models
• Participation from around the world, though primarily U.S.

So we’re speaking mostly about companies selling software to the enterprise and mid-market segments here and probably fewer to SMBs customers (especially that in the results underneath the companies with less than $2.5M of revenue are excluded).

Results

The 2 graphs above are very interesting to get a bit more details:

• the churn rate from customers with an ACV (annual contract value) < $1K is much higher than customers with an ACV > $1K
• ACV < $1K generally corresponds to SaaS companies with a “self serve - low price” model where there’s generally no direct sales / few inside sales.

Comments

Churn figures:

• Median annual gross dollar churn = 6% ~ 0.51 monthly
• Median annual customer churn = 8% ~ 0.69 monthly

Knowing that:

• it excludes companies with revenue less than $2.5M
• the median annual contract value (ACV) is $21K (in the )
• the median number of customers per company is 284

We can infer that these churn results are representative of SaaS companies targeting mid-size / enterprise customers with generally sales / inside sales

Open startups

About this benchmark

I’ve basically collected the churn data from the “open” SaaS companies (except Outreach Signals which is more an online agency than a SaaS).

Benchmark Participants

• median MRR = $18.9k, average MRR = $72.8k ( completely distorted by Buffer revenue)
• median ACV = $552, average ACV = $941

These are mainly SaaS targeting SMBs, with a much lower price point, ACV and also more early stage companies compared to the Pacific Crest survey ones.

Results

• median monthly gross revenue churn = 11.1% ~ 75% annual (huge)
• median monthly user churn = 5.4% ~ 46.% annual

We can also notice that revenue churn is higher than customer churn for these companies, it’s the opposite for the Pacific Crest Survey ones.

Conclusion

These numbers are interesting because they confirm that at early stage and for a lower price point, SaaS companies tend to have higher churn rates (user and revenue) compared to Enterprise — mid market ones. Which was expected.

Totango Annual SaaS Survey

About this benchmark

Totango is a customer success software (one of their aim is to lower churn) which runs every year a survey among SaaS companies to benchmark several metrics.

Survey Participants

“500 SaaS professionals” which revenue distribution looks like this:

60% of participants have an ARR < 10M.

Results

A first problem is that the survey doesn’t say clearly whether we’re talking about gross or net revenue churn. Since they don’t have negative churn on the graph I infer that it’s gross churn.

So basically:

• 34% of companies have an annual gross revenue churn < 5% (= 0.43% monthly)
• 31% of companies have an annual gross revenue churn between 5% and 10% (= 0.43% — 0.87% monthly)
• 31% of the companies have an annual gross revenue churn > 15% (> 1.3% monthly)

Conclusion

We have unfortunately not enough information about the precise profile of these companies or the distribution of churn by ACV to draw real conclusions.

Groove Small Business Conversion Survey

About this benchmark

In 2013 Groove, a customer support software, run a survey to benchmark several SaaS metrics among small business saas.

Survey participants

“712 respondents who have reached Product/Market Fit, have been in business for at least 6 months and have at least $1,000 (but less than $500K) in monthly recurring revenue.”

• Average MRR = $10.5K
• average ACV (for B2B participants) = 1680$

Results

Again the problem is that Groove doesn’t precise which churn they talk about (revenue, net, gross or user churn). Especially that in their survey the question asked was quite open “What is your churn rate?”.

That being said we can infer that it’s monthly user churn (from the context).

• 3.2% monthly user churn (= 32% annual)

Conclusion

The participants of this survey are very similar to the ones from the open startup (around the same ACV and MRR) with a lower monthly user churn though, 3.3%, compared to the 5.4% of the open startups.

But again we can only assume that it’s user churn and we don’t really know the methodology used by each company to calculate their churn. So these results should be taken cautiously.

So finally, what should we think of these numbers?

Tomasz Tunguz shared a table that shows his “general observations about customer account churn rates by segment”

This is also what I regularly see among SaaS startups, so let’s compare that to our 4 benchmarks.

Pacific Crest Survey:

• Mid-Market / Enterprise segment
• Median annual customer churn = 8% ~ 0.69 monthly

That’s in line with the 0.5–1% monthly and 6% — 10% customer churn from the table above.

Open Startups

• SMB segment
• Median annual customer churn 46.% ~ 5.4% ~ monthly

That’s in line with the 3–7% monthly and 31% — 58% customer churn from the table above.

Totango

We cannot draw meaningful conclusion here.

Groove Survey

Big warning as we don’t have enough context and info on what and how the churn rate was calculated. If we pretend it’s coherent and concerns monthly user churn then:

• SMB segment
• Median annual customer churn 32% ~ 3.2% monthly

That’s in line with the 3–7% monthly and 31% — 58% customer churn from the table above.

So if you are a pre- P/M fit and early stage startup (less than 1-2 years with paying customers):

• it’s normal if your churn fluctuates a lot (even from month to month)
• it’s likely that your churn rates are higher than the ones above

If you have reached P/M fit and that you are running your business for more than 2 years and:

• if your churn rates are higher than those indicated (in your category) you should work hard on lowering them