First contact resolution is the rare support metric that customers and finance both care about, which is precisely why it gets distorted. Set a target too high and agents start closing conversations optimistically. Set none at all and nobody owns the repeat contact.

Cost out a second contact honestly

The direct cost is easy: another agent-handled interaction. Take fully loaded agent cost per hour, divide by conversations handled per hour, and you have a unit cost. The indirect costs are larger and usually omitted:

  • Context reconstruction: two to four minutes per repeat contact spent reading history, even with a good timeline.
  • Escalation probability: repeat contacts escalate at several times the base rate, and an escalation typically costs an order of magnitude more than a first-line reply.
  • Satisfaction drag: satisfaction falls sharply between the first and second contact on an issue, and the effect persists into the next interaction.
  • Retention exposure: the accounts with the highest repeat contact rates are consistently over-represented in non-renewals.

Model these together and a repeat contact usually costs somewhere between three and six times a first contact. That multiple is the number that unlocks the investment conversation, because it converts a support quality problem into a straightforward payback calculation on product and documentation fixes.

Why very high FCR targets backfire

Above roughly 85%, the marginal gains stop coming from better answers and start coming from redefinition. Conversations get closed while the customer is still unsure. Genuinely complex issues get force-fit into a single interaction. And agents stop referring cases that should be referred. A target in the high seventies to low eighties, paired with a low repeat contact rate, describes a healthier operation than a 92% FCR with a 20% reopen rate.

Attribute failures to causes you can fix

Tag every repeat contact with a cause: missing documentation, product defect, policy ambiguity, agent knowledge gap, or customer-side dependency. Report the distribution monthly. In most organisations two causes account for the majority of repeats, and at least one of them is owned outside support, which is exactly why the cost model matters. It is the artefact that turns "support is struggling" into a prioritised backlog item with a number attached.

A simple monthly review

  • Repeat contact rate overall and for the top ten accounts by value.
  • Cause distribution, with the change from last month.
  • Cost of repeats, expressed in agent-days, next to the engineering effort needed to remove the top cause.
  • One committed fix, with an owner outside the support team.

Run that review for two quarters and the argument for investing upstream stops needing to be made. The numbers make it.