Planning the capacity of your Customer Success team

Derek Skaletsky
8 min readDec 13, 2019


I’m writing this post as a follow-up to David Apple’s great post — Nobody knows the ROI of their CSMs. Anyone in charge of building a CS department/team should read it (and follow David, in general).

I love that David confronted the misguided practice of trying to assign an ROI to a Customer Success Manager head-on. He clearly explains why this is a fruitless exercise and one that we need to just stop.

David is spot on when he opens his argument by asking,

What’s the ROI of your Mom?

Haha. Good line (and sounds like the beginning of an odd “Your Mama” joke).

The point David is making is that it’s impossible to calculate ROI on everything under the sun. More tactically, as it applies to your SaaS business, we all need to realize that it’s near impossible to calculate a financial ROI on anyone (or team) that isn’t DIRECTLY tied to rev generation or cost reduction.

  • Sales — sure, pretty easy.
  • Marketing — meh, maybe, but not really (marketing spend easier than marketing people, but still…).
  • Everyone else — nope. Don’t bother.

I mean, what’s the ROI of a Product Manager? Or a backend engineer? Or a designer? Or…this is a good one…a finance director (haha…next time this spreadsheet warrior asks you about the ROI of your CSMs, just flip the script :))

After reading David’s post, it’s easy to see how it’s silly to frame the growth of a CS team in the context of ROI. He came just short of saying, “Let’s just cut the shit here, ok?”…but I’ll say it.

Can we just cut the shit on the whole what’s the ROI of a CSM exercise?

With this said, this does not mean that CS leaders shouldn’t be required to quantify their growth plans. In fact, I would call it negligent for any CS leader not to apply rigorous, quantitative support when it comes to their growth planning.

But it’s not a question of ROI. It’s a question of capacity. The question you need to answer is:

How many CSMs do we need to deliver the service-level we want without overspending?

Sounds like a simple question, but let’s break it down a little. In order to answer this question, you will need to dig into its three main elements. They are:

  1. The capacity of a single CSM. How much time does a CSM have to dedicate to deliver on your service plan?
  2. Your desired service plan. What kind of service do you want to consistently deliver to your accounts?
  3. Without overspending. Ahhh….yeah, that.

Step 1: Figure out the Capacity of a CSM

This is the easy part of the exercise, but not always obvious. If you start your planning exercise thinking that your CSMs have 40 hours a week to dedicate to serving accounts, your calculations are going to be way off. When I budget for CSM time, I generally do it on a monthly basis.

I start with 160 hours of work time in a month. Then I deduct:

  • 20% for internal activities (meetings, training, reporting, outings, etc). Could be more, could be less, but 20% is pretty safe.
  • 12 hours per month to account for vacation time (assuming 4 weeks annual vacation/PTO broken down monthly)
  • 10% for other CS initiatives (writing help docs, processes, supporting product initiatives, etc)

These are the base assumptions. But you should also think about deductions for:

  • Travel — if you have a service model that includes travel to customer sites, you need to account for this (more later).
  • Conferences — if your CSMs will attend industry conferences, account for this time. If your company hosts an annual conference and you expect your CS team to participate in the planning and delivery…definitely account for this.
  • Anything else that will take CSMs time away from direct service (unique to your business/company).

But going off the base assumptions, you should plan on your CSMs to have 100 hrs/month for direct service (160 hours minus 20% internal minus 12 hours for PTO minus 10% for CS initiatives).

Ok, good.

Step 2: Decide on your desired service-level

What kind of service do you want to deliver?

This is very much a unique decision for your business, but it is worthwhile to define — and get buy-in for — the type of service you want to deliver. This desired service level will be very unique to your product and business model. For example, you may want a service model that looks like one of these:

Full Service

The gold standard service plan. Very proactive, hands-on, dedicated service. It would look something like this:

  • Full onboarding for new accounts
  • Full onboarding for new users on existing accounts
  • Regular, proactive account check-in — monthly
  • Quarterly business reviews (QBRs)
  • One face-to-face, onsite visit per year
  • Response time to inbound tickets within 2 hours

Mid-level Service

  • Full onboarding for new accounts
  • Proactive account check-ins — quarterly
  • In-app chat support (24-hour response time)
  • Reactive email support (24-hour response time)

Low-touch Service

  • No onboarding support
  • In-app chat support (24-hour response time)
  • Reactive email support (24-hour response time)

Obviously, these are just some basic examples, but I hope they serve as a good starting point for defining your desired service level. You have can different service plans for different customer segments (start-up vs enterprise accounts), different customer stages (new vs mature), and/or break-up your team to handle different pieces of the plan (proactive vs reactive support, for example), but for simplicity, we’ll assume one team delivering on the same service plan.

Once you have a desired service plan mapped out, the next step is estimating how much time it would take a single CSM to deliver on this plan for a single account. I won’t go through each scenario, but for illustration, the time required for the Full-Service plan would look something like this:

So…if it takes 7.5 hours per month to deliver on your service plan for each account, and there are 100 service hours available for each CSM, this means that one CSM can effectively handle 13–14 accounts.

3. Deliver without Overspending

And here is where it all comes together. In the above example, we found that a single CSM could effectively service 13 accounts with the desired service plan. Now you have to see if this ratio works for your business.

Let’s assume a full-time CSM will cost $100k annually (salary, benefits, etc). How much does each of your accounts need to be worth for this to make sense for you? You need to figure out what economics work.

If your ACV is $100k annually, then one CSM will cover $1.3mm. That makes a lot of sense.

At $50k ACV, one CSM will cover $650k. Starting to feel the scaling crunch here, but may be doable for a bit.

At $20k ACV, one CSM will cover $260k. This clearly doesn’t work.

Important note — as David explains in his post, CSM capacity and allocation plans should not be based on a ARR/CSM ratio (each CSM should manage $2mm in ARR, for example). This is a BACKWARD approach to the question. You need to do the above calculation to make sure your economic model works based on your desired service level, the capacity of your CSMs, and your price point. If the economics doesn’t make sense, then you need to do some iterating.

Iteration: the Service Level or your Price point

If the economics don’t work out the way you’d want them to, then you only really have two levers to adjust — your desired service level or your price point.

If your CSMs can’t effectively manage enough accounts to fit your model, then you need to adjust your desired service level. Maybe you can’t provide full onboarding to every account, or maybe F2F meetings aren’t going to work, or maybe you need to be less proactive.

“But we don’t want to cut our service-level, we’ll lose customers!!”

Ok, then…

Raise your prices. This is the other level you have to make the model work. You can’t invent more hours in a month, so if you want to keep the same service level, but do so more profitably…then raise your prices. A related, but alternative option here is actually charging for services. This would allow you to keep a high service-level without raising the base rates for the product. Moving into a services model is a pretty significant strategic decision, but it is a way to justify the cost of high service levels.

But wait…did I say you only had two levers to make this model work? Actually, that’s not true. As mentioned earlier, you can segment your customer base and deliver different service plans; you could break up the team to handle different types of service; you could create different teams for different stages in the lifecycle; or….you could make your product easier to use. You can improve onboarding, you can reorganize your feature set to decrease complexity, you can provide more training assets, etc. This will allow you to adjust your service plans without lowering success. But, of course, this is a much bigger initiative :)

Testing and improving your Model

Once you have a model and scaling plan that you are happy with, by all means, test it. Like David said in his post, give some of your CSMs more accounts than what your model calls for, others less. Then watch — both your accounts as well as your team. How are the engagement levels of each account group? Are they remaining engaged and seeing full value from your product or are they dropping off? How are the CSMs doing? Are they overly stressed? Bored? Finding new ways to deliver and scale service?

Adjust your thinking and adjust your model as you learn more.

In Conclusion

David is right — spot on — in his post. Regardless of how much your finance team may want it, there is no way to accurately calculate the ROI of a CSM. So don’t even bother. But this doesn’t mean you shouldn’t be able to quantify your growth plans with some smart capacity planning. Be rigorous and deliberate about it. This should keep those nasty spreadsheet warriors off your back.




Derek Skaletsky

Tech founder (mostly SaaS). Latest — Sherlock (; Boston expat; Hollywood escapee; hack photographer; dad (x2)