How do you measure the value of Customer Success?

Short answer

How do you measure the value of Customer Success?

Measure at four levels: business outcomes (GRR, NRR, churn, LTV), customer outcomes (time to value, goals achieved), operational indicators (customer health, unnecessary enquiries) and activities. Activity alone proves nothing, and retention can rarely be attributed to Customer Success alone. Talk about contribution, and measure it with cohorts or control groups.

Customer Success is meant to create value for both customers and the company. That is easy to say and harder to show. Many Customer Success functions report what is easiest to count – number of meetings, number of emails, number of customers contacted – because it is hard to connect the work directly to revenue.

This article provides a framework for measuring the value more precisely. It explains the most important measures, how they are calculated, and why they must be interpreted with care.

Four levels of measurement#

It is useful to distinguish between four levels. They are connected, but they answer different questions.

Four levels of measurement

Arrives lateCan be steered directly

1. 1. Business outcomes

   GRR, NRR, churn, LTV

   What do the customers mean to the company financially?

2. 2. Customer outcomes

   Time to value, goals achieved

   Are customers achieving what they bought the product for?

3. 3. Operational indicators

   Customer health, adoption, enquiries

   Are we on the right track?

4. 4. Activities

   Meetings, campaigns, contacts

   What have we done?

The most common problem is measuring a lot at level 4 and hoping it leads to level 1. Activities are necessary, but they say nothing about whether the work is effective. Ten customer meetings can be valuable or wasted. A hundred emails can help customers or irritate them.

A good measurement system connects the levels: activities should affect operational indicators, which should lead to better customer outcomes, which in turn should show up in business outcomes.

Business outcomes: retention and expansion#

Customer churn and revenue churn#

Churn means attrition. There are two main variants, and the difference matters.

Customer churn is the share of customers who have left in a period.

Revenue churn is the share of revenue that has disappeared in the same period.

The two can paint completely different pictures:

Hypothetical example

10% / 2%

A company loses ten out of a hundred customers, but only 2% of its revenue – because the customers who left were small.

The opposite can also happen. If one large customer accounting for 20% of revenue leaves, customer churn is 1%, while revenue churn is 20%. Both figures are needed to understand the situation. In fast-growing companies it also matters what you divide by – see How to measure churn in a fast-growing company.

Gross Revenue Retention (GRR)#

GRR shows what share of revenue from existing customers has been retained over a period, without counting expansion. It can never exceed 100%.

GRR = (revenue at the start of the period − revenue lost to cancellations − revenue lost to downgrades) ÷ revenue at the start of the period

GRR is a good measure of how well the company keeps what it already has.

Net Revenue Retention (NRR)#

NRR shows the same thing, but includes expansion from existing customers. NRR can therefore exceed 100%, which means the customer base grows in value even without new customers.

NRR = (revenue at the start of the period − revenue lost to cancellations − revenue lost to downgrades + revenue from expansion) ÷ revenue at the start of the period

Hypothetical example

94% / 101%

The same customer base has a GRR of 94% and an NRR of 101%. The company loses some of what it has, but expansion more than makes up for it.

From GRR to NRR

Show the numbers

Item	Amount
Start	NOK 1,000,000
Cancellations	NOK -40,000
Downgrades	NOK -20,000
GRR 94%	NOK 940,000
Expansion	+NOK 70,000
NRR 101%	NOK 1,010,000

It is important to look at both. A high NRR can hide significant churn if a few customers grow a lot. A high GRR with a low NRR may indicate that customers stay, but do not get enough value to want to expand.

Both measures are usually calculated on a fixed group of customers – those who were customers at the start of the period – and over twelve months. New customers in the period are excluded.

LTV and CAC#

LTV (lifetime value, also called CLV) is an estimate of how much a customer is worth over the whole relationship. A simple formula is:

LTV ≈ average monthly revenue per customer × gross margin ÷ monthly customer churn

With average monthly revenue of NOK 2,000, a gross margin of 70% and monthly churn of 2%, the estimate is 2,000 × 0.7 ÷ 0.02 = NOK 70,000. The figures are hypothetical.

CAC (customer acquisition cost) is what it costs, on average, to acquire a new customer: sales and marketing costs in a period divided by the number of new customers in the same period.

Together they give two useful ratios:

• LTV:CAC shows how many times the customer pays back what it cost to acquire them.
• CAC payback period shows how many months it takes before the gross profit from the customer has covered CAC: CAC ÷ (monthly revenue per customer × gross margin).

Hypothetical example

5.0 : 1

The customer is worth five times what it cost to acquire, and has paid back the acquisition cost after ten months.

This is where the connection to Customer Success becomes clear. Lower churn increases LTV directly, without changing CAC: if monthly churn in the example is halved from 2% to 1%, LTV doubles to NOK 140,000, and LTV:CAC goes from 5 to 10. The work of keeping customers thus makes every krone spent on acquiring them worth more.

LTV at different monthly churn rates

Show the numbers

Monthly churn	LTV
3%	NOK 46,667
2%	NOK 70,000
1.5%	NOK 93,333
1%	NOK 140,000

The payback period also shows why early churn is so costly. A customer who leaves before CAC has been earned back has cost the company money. That is a strong argument for investing in onboarding and time to value.

Both figures are rough. The LTV formula assumes that churn is constant over time, which it rarely is. Many companies have the highest churn in the first months and much lower churn among long-standing customers. CAC depends on which costs are included, and should be calculated the same way every time if the figures are to be compared over time. There are often-cited rules of thumb for what a healthy ratio between LTV and CAC is, but they are based on companies other than yours and should not be used as an answer key.

If you want to calculate churn, LTV and CAC with your own numbers, you can use the churn calculator.

Customer outcomes: are customers achieving what they wanted?#

Business outcomes are what management cares about most, but they arrive late. When a customer cancels, it is often too late to do anything. Customer outcomes say something about where customers are heading.

Time to value is the time from when the customer starts until they achieve the first concrete value. What “value” means must be defined for each product: first payment received, first report delivered, first invoice sent, first employee registered. The shorter the time, the less chance that the customer loses motivation along the way.

Goals achieved is the most precise measure, but also the hardest. It requires knowing what the customer wanted to achieve, and being able to see whether it has happened. For large customers this can be agreed explicitly. For many small customers you have to use indicators in product data.

Use of core features shows whether customers are using the parts of the product that actually create value. That is not the same as general activity. A customer who logs in every day but never uses the feature they bought the product for is not necessarily a healthy customer.

Operational indicators: are we on the right track?#

Operational indicators are early signals. They are used to prioritise work and to spot risk before it becomes visible in revenue.

Customer health#

A health score (customer health score) combines several signals into one measure: usage, change in usage, enquiries, payment history, feedback and possibly contact with an account owner. The point is to be able to sort customers by their need for follow-up.

Health scores are useful, but they are only as good as the signals they are built on. Two common mistakes:

1. The score is never validated. It is assumed that the signals are associated with churn, without checking afterwards. A health score should be tested against actual churn regularly.
2. The score becomes too complicated. When nobody understands why a customer has a low score, it becomes hard to act on.

Customer satisfaction#

CSAT measures satisfaction with a specific interaction. NPS (Net Promoter Score) asks how likely the customer is to recommend the company, and was launched as a measure of loyalty.¹ Both are useful as signals, but neither is the same as value. Customers can give a high score and still leave, and dissatisfied customers can stay because switching is expensive.

Unnecessary friction and unnecessary enquiries#

Two indicators often get too little attention, especially in companies with many customers:

Unnecessary friction is obstacles in the customer journey that do not need to be there: unclear instructions, complicated forms, missing notifications, recurring errors. Friction makes it harder for the customer to get value, and it often affects many customers at once.

Unnecessary enquiries are enquiries that could have been avoided with a better product, information or process. The question is not whether Customer Support answers well, but whether the customer should have needed to ask.

Both can be measured. Examples:

• Number of enquiries per active customer, over time
• Share of enquiries in categories considered preventable
• Share of customers who stall at a particular step in onboarding
• Time spent on recurring cases

When friction is removed, there is often a double gain: customers get a better experience, and the company spends fewer resources handling problems that should not have arisen.

Activities: necessary, but not sufficient#

Activity measures have their place. They help leaders understand capacity, plan staffing and see whether planned work is actually being done. But they should never be the only – or the most important – measure of Customer Success.

Activities say what has been done, not whether it worked.

A team measured on the number of customer meetings will hold many customer meetings. That is not the same as customers getting more value. Activity measures should therefore always be linked to a hypothesis about which indicator they are meant to affect, and followed up by measuring that indicator.

Attribution: what is really down to Customer Success?#

This is the biggest challenge. Retention, expansion and customer satisfaction are affected by almost everything the company does: product quality, price, sales, marketing, competitors, economic conditions and the customers' own situation. It is rarely possible to say that a particular share of retention is due to Customer Success.

It is more honest to talk about contribution than about attribution. Some methods make the contribution clearer:

Cohort comparison. Compare groups of customers who have received different follow-up – for example customers who completed an onboarding programme and customers who did not. Be aware that the groups often differ in other ways too. Customers who choose to take part may be more motivated to begin with.

Control groups. The most precise approach is to keep a randomly selected group of customers out of an intervention and compare the results. This is common in marketing but used less often in Customer Success, partly because it can feel wrong to withhold help. For interventions such as automated emails or in-product messages, it is often unproblematic.

Before and after. Measure an indicator before and after an intervention. The method is simple but weak, because other things may have changed at the same time.

Leading indicators. Show that the intervention affects an indicator documented to be associated with retention. An imagined example: if your own data show that customers who complete onboarding within 14 days have lower churn than others, and an intervention increases the share who do, that is a reasonable argument that the intervention contributes to retention. The 14-day threshold is just an example – the right threshold is found in your own data.

A practical starting point#

For a company that wants to start measuring Customer Success more systematically, this is a realistic place to begin:

1. Calculate GRR and NRR for the last twelve months, and break them down by customer segment.
2. Separate customer churn from revenue churn, and see where in the customer relationship churn happens.
3. Define “first value” for the product, and measure how long it takes.
4. Categorise enquiries and identify the biggest sources of unnecessary friction.
5. Choose two or three operational indicators the team can influence, and test whether they are associated with retention.
6. Report activities as capacity, not as results.

The goal is not perfect numbers. The goal is to be able to answer the question honestly: are our customers getting more value than before – and do we see it in the results?