What it means
The metric answers a simple question: when a customer got in touch, did the problem actually go away? A contact only counts as resolved if the customer does not come back about the same issue within an agreed window, commonly seven days.
It matters commercially because repeat contacts are pure waste. Every second call about the same fault consumes agent time, adds queue length for other customers and pushes up the cost to serve without adding any revenue.
High resolution rates therefore cut cost and lift satisfaction at the same time, which is rare among service metrics. Measurement is where most of the argument happens.
Some teams use agent self-reporting, some use a customer survey question, and the most reliable approach matches contact records by customer and issue type to spot genuine repeats. Different methods can produce readings twenty percentage points apart for the same team.
Typical rates sit somewhere between 70% and 80% for general consumer support, though a technical helpdesk handling complex faults will sit lower and a simple order-status line will sit higher. Because of that spread, comparing the raw number across industries is close to meaningless, and the useful comparison is against the team's own trend.
There is one important trap: pushing the metric too hard encourages agents to declare things resolved when they are not, or to keep customers on the line while they improvise a fix. Sensible operations pair the measure with customer satisfaction and average handling time so that gaming one number shows up in another.
In practice
Real-world examples
Example 1
A retail bank finds that card dispute calls resolve first time only 52% of the time because agents lack authority to issue provisional refunds. Raising the agent refund limit to $250 lifts resolution to 74% within two months and removes roughly 900 repeat calls a month.
Example 2
An online furniture retailer notices that damaged-delivery contacts almost always need a second call while photos are reviewed. Adding a photo upload step to the first contact lets agents approve replacements immediately, and the resolution rate for that contact type climbs from 41% to 83%.
Example 3
A software firm splits its measure by product line and discovers that one recently launched module drags the whole centre's figure down. The support data becomes the strongest evidence the product team receives, and three documentation fixes remove a fifth of the module's contacts entirely.
Think of it
FCR shows how often you solve the customer's problem on the first try-support efficiency and effectiveness.
Formula
Calculation
First Call Resolution = Contacts resolved on first attempt / Total contacts handled
Worked example: a broadband provider's support centre handles 12,000 contacts in a month. Matching the records shows that 8,400 of those customers did not get back in touch about the same issue within seven days.
First call resolution is 8,400 / 12,000 = 0.70, or 70%. That leaves 3,600 contacts that generated at least one repeat.
If the fully loaded cost of handling a contact is $6, those repeats cost 3,600 x $6 = $21,600 a month. Lifting first call resolution to 80% would leave 12,000 x 20% = 2,400 repeats, costing 2,400 x $6 = $14,400. The saving is $21,600 - $14,400 = $7,200 a month, or $86,400 a year, before counting any benefit from happier customers.Case study
Seen in the real world
Larkspur Energy is a fictional utility retailer created purely to illustrate the point. Its contact centre reported first call resolution of 88%, a figure the operations director had been presenting to the board for two years, and yet complaints to the ombudsman kept rising. The measure was based on agents ticking a resolved box at the end of each call.
A review matched customer identifiers across a seven-day window and found the true rate was 61%. Billing queries were the worst area, because agents could explain a bill but could not correct one, so almost every explanation call was followed by a correction call a few days later.
Larkspur gave billing agents direct adjustment rights up to $150 and rebuilt the measure on matched records rather than self-reporting. Within six months the honest figure reached 79%, contact volumes fell by about a tenth, and the illustrative lesson was that the fix only became possible once the measurement stopped flattering the team.
Watch out
Common mistakes
- Relying on agents to mark their own contacts as resolved, which almost always overstates the true rate.
- Setting an aggressive target without giving agents the authority or system access needed to actually finish the job.
- Comparing the percentage against another company's published figure when the two are measured over different windows and contact types.
Questions
People also ask
What counts as a repeat contact?
Any further contact from the same customer about the same underlying issue inside the chosen window, which is most commonly seven days but can be anything from 24 hours to 30 days.
Does the measure apply to email and chat as well as phone?
Yes, and most teams now describe it as first contact resolution so the name reflects every channel they operate.
Can the rate ever be too high?
A reading close to 100% usually signals a measurement problem or an unusually simple contact mix rather than outstanding service.