First-Contact Resolution: What It Demands of the Organisation

An organisational outcome, not an agent's skill#

First-contact resolution (FCR) is often presented as an agent's skill: listening well, diagnosing well, answering well. That is true, but not enough. A competent agent cannot resolve a request if they are not allowed to grant what is needed, if they cannot see the status of the file, or if the answer depends on a department that has not committed to any deadline.

Every unresolved request costs several times over: a second contact to handle, a customer spending time chasing, a manager called in to unblock the situation and, sometimes, a formal complaint. In regulated sectors, the cost also takes a regulatory form, since complaints are monitored by the supervisor. Improving FCR therefore reduces workload, customer effort and risk at the same time.

This is why FCR is one of the best indicators of execution. When it is low, the useful question is not "who is resolving badly?" but "what in our organisation prevents resolution at first contact?". Our guide to contact-centre quality presents the measure alongside a centre's other indicators; this page deals with what has to change to improve it.

What first-contact resolution requires#

These conditions reflect those that determine any team's autonomy. Decision rights are often the first to address: a large share of repeat contacts comes from requests the agent knew how to settle but was not allowed to.

The link with customer effort#

In 2010, Matthew Dixon, Karen Freeman and Nicholas Toman published a study in the Harvard Business Review that left its mark on customer service: reducing the effort customers must make to get an answer does more for their loyalty than trying to exceed their expectations. Having to call back, repeat the problem or be transferred are among the main sources of effort.

FCR and customer effort are therefore linked: every repeat contact is extra effort. How the customer effort score works, and its limits, is explained in our guide to NPS, CSAT and CES. One measure is not enough: they need to be read together.

Measuring honestly#

The most reliable measure combines the last two. The gap between agent-declared resolution and behavioural resolution is itself a useful indicator: it points to the issues where the organisation believes it resolves the request while the customer keeps coming back.

Three rules keep the measure honest:

1. Set the window in advance, issue by issue if necessary, and do not change it to improve the figure.
2. Count every channel. A customer who calls, then writes, then visits a branch has made three contacts, two of them repeat contacts.
3. Exclude explicitly what has nothing to do with resolution, such as an unrelated new request, and document those exclusions.

Common traps#

• Making agent-declared FCR an individual target. Agents quickly learn to code "resolved". It is a textbook case of an incentive that rewards the opposite of what you ask for, described in our guide to incentives.
• Treating a transfer as a resolution. The customer has been redirected, not served.
• Counting a call-back promise as resolved. It is resolved only once the call-back has been made and the request settled.
• Closing a complaint to meet a deadline. In banking, Bank Al-Maghrib Circular No. 10/W/16 sets 40 working days for a substantive response; a quick reply that settles nothing often produces a new complaint, then an appeal. See banking complaints in Morocco.
• Ignoring poorly coded contact reasons. If a third of requests are classed as "other", the behavioural measure is worthless.

What to change#

Start from the five to ten most frequent reasons for repeat contact and, for each one, look for what was missing during the first exchange.

• Decision rights. List the decisions that trigger escalations and delegate the frequent, low-stakes ones, with a threshold, criteria and sample-based checks.
• Tools and information. Show the customer's history and the status of their files in the agent's tool. Access must remain limited to what is needed to handle the request, in line with Law 09-08.
• Knowledge base. One article per frequent case, one owner per article, and a correction within the week when a repeat contact reveals an error.
• Commitments between departments. For every request passed to the back office: a deadline, feedback to the agent or the customer, and an end-to-end owner.
• Routine. Review a few real repeat contacts every week in the KAPER Cadence, and track a lead measure such as the share of the previous day's repeat contacts handled the same day.
• Automation. An AI-based assistant can help the agent find the right answer, provided it is properly governed: see AI in customer service.

A request that keeps coming back despite these changes points to upstream causes in the product, billing or a process: that is the purpose of root-cause complaint management.

Your checklist#

• A written definition of FCR, with the repeat-contact window for each issue.
• A measure that combines customer confirmation and the absence of repeat contact, across all channels.
• The gap between declared and actual resolution, tracked by issue.
• The list of the most frequent reasons for repeat contact and, for each one, the missing condition.
• Written decision rights for frequent decisions.
• Customer history visible in the tool, within the limits of Law 09-08.
• Written deadline commitments with the back office.
• A weekly review of a few real repeat contacts.