CX Metrics · Revenue
Your CSAT might be lying to you. Here's the formula that tells the truth
Your CSAT is 88 percent. Your FCR is 54 percent. One of these is lying to you, and it is not the one you'd expect.
I have watched GMs present CSAT in board meetings as if it settles the question of whether support is working. It doesn't. CSAT measures how a customer felt about the last interaction. First Contact Resolution measures whether their actual problem got solved. Those are two different questions, and a business can score well on one while quietly failing the other.
Why a high CSAT can hide a real problem
Here is the mechanism. A customer contacts support with a broken feature. The agent is warm, apologizes well, and closes the ticket with a promise to escalate. The customer, surveyed minutes later, rates the interaction 5 out of 5, because the agent was pleasant and fast. Three days later the feature is still broken, they contact support again, and this second contact never touches the CSAT score at all, because most CSAT surveys fire once, right after the interaction, before anyone knows whether the fix actually held.
FCR asks the harder question: did this contact resolve the issue, so the customer never had to come back for the same problem? When FCR sits low while CSAT sits high, you are usually looking at an agent team that is excellent at being liked and mediocre at fixing things. That combination feels good in a dashboard and costs money in the business.
A friendly agent who does not solve the problem produces a satisfied survey response and an unsatisfied customer. The survey just asked at the wrong moment.
The bridge nobody draws: repeat contacts to revenue
Low FCR shows up operationally as repeat contacts, the same customer reaching out again for the same underlying issue. Repeat contacts are expensive to handle, but the cost that actually matters to a GM is not the extra agent time. It is what repeat contacts signal about churn risk.
A customer who has to explain their problem twice is a customer whose trust in the product just took a hit. Not every one of them churns. But a rising repeat contact rate, sitting quietly underneath a stable CSAT, is one of the more reliable early indicators that Net Revenue Retention is about to move, weeks before the finance team sees it in the renewal numbers.
Most board decks have a slide for NRR. Almost none have a slide explaining why NRR is moving. CX data, read correctly, is often that explanation.
The CSAT-to-Euro bridge
I am not going to hand you a universal multiplier claiming that one CSAT point equals a fixed percentage of NRR. Anyone who tells you that is selling you a number they made up, dressed as an industry benchmark. Every business has a different relationship between service quality and revenue, and pretending otherwise would break the same rule I hold myself to everywhere else on this site: no invented data.
What I can give you is the actual bridge, the four steps that connect a support metric to a number your CFO will recognize. Run it with your own figures.
- Start with your churn. Take your annual gross churn rate and the customer count it applies to.
- Isolate the CX-attributable slice. From exit surveys or churn interviews, find the share of churned customers who cite a service or support issue as a driver. If you don't have this number yet, that gap is itself the finding: you cannot manage what you have never tagged.
- Convert to revenue at risk. Multiply CX-attributable churned customers by your average revenue per account.
- Translate to NRR. Compare revenue at risk to your starting ARR. That percentage is your CX-attributable NRR drag, in language your board already speaks.
This is a directional model, not an audit. It assumes churn reasons are additive and independently reported, which is a simplification. Its purpose is to turn a vague concern about support quality into a specific number worth investigating properly, not to replace that investigation.
Reading the FCR trend, not the FCR benchmark
People ask me what a good FCR number is. I usually decline to answer with a single figure, because FCR definitions vary too much between companies to make external benchmarks meaningful; what one company counts as a single contact, another splits into two. What matters is the trend inside your own operation. A declining FCR next to a flat or rising CSAT is the specific pattern worth chasing, because it means your team is getting better at the survey and no better, or worse, at the actual problem.
The fix is rarely a training issue. It is usually a knowledge base and escalation design issue, which I covered in more depth in AI deflection fails when it is implemented under pressure, and an ownership issue, covered in locus of control and agent disengagement. Agents close tickets fast when they are rewarded for closing, not for resolving. Change what you measure, and the behavior underneath it changes with it.
What to bring to your next leadership meeting
Not "our CSAT is good." Bring this instead: "Here is our CX-attributable churn slice, here is the revenue behind it, and here is what closing even part of that gap would do to NRR." That is a finance conversation, not a support department update, and it is the version of this data that actually gets budget approved.