The Weekly Agency Scorecard

Most agency reporting measures dials and submitted applications. Both are real, both are already known to everyone in the building, and neither tells you anything you can act on — by the time submitted apps move, the cause was three weeks ago.

A weekly number is only worth reviewing if it moves before the outcome does. That is the filter. Five metrics pass it.

1. Speed to first contact attempt

Median minutes from lead delivery to first dial attempt. Median, not mean — one lead worked at 3am drags a mean into fiction.

This is the strongest single leading indicator in phone-based insurance sales, and it is the one an agency owner can actually fix this week. Conversion decays steeply with delay, and the decay is front-loaded: the difference between two minutes and twenty is far larger than between twenty and sixty.

Targets: under 5 minutes for real-time internet leads, under 2 hours for direct mail responses, same day for aged.

Segment it by agent, not just agency-wide. A single agent sitting on leads for four hours is invisible in the aggregate and highly visible in their own results.

2. Contact rate, by lead source and cohort

Right-party contacts ÷ unique leads, per weekly cohort, per source.

Cohort matters more than the aggregate. A source whose week-1 leads contacted at 32% and whose week-9 leads contact at 18% is degrading, and averaging across the quarter hides it entirely — you find out at renewal, three months of spend later.

A falling contact rate has four candidate causes and they are distinguishable in about an hour: caller-ID reputation, list decay, dayparting drift, or dialer pacing. The diagnostic sequence is in connect-rate benchmarks, and the caller-ID test comes first because it is the cheapest — see Spam Likely.

3. Contact-to-application rate

Applications ÷ right-party contacts, by agent.

This is your sales metric, isolated from lead quality and from activity volume. An agent with poor dial numbers but a strong contact-to-application rate has a work-rate problem. An agent with heavy volume and a weak rate has a skill problem. They need opposite interventions, and a combined “closing ratio” measured against dials cannot distinguish them.

Watch the spread across your team more than the average. A three-fold spread between your best and worst on the same lead source is a coaching agenda, and it is worth more than any lead-buying decision.

4. Issue rate

Policies issued ÷ applications submitted, by agent and by lead source.

The metric that separates activity from income. Final expense declines and postpones are common enough that a 15-point spread between sources is ordinary.

Two very different problems show up here:

• Low issue rate by source means the vendor is generating uninsurable prospects. That is a lead-buying decision — run it through the 90-day vendor audit.
• Low issue rate by agent on the same sources means the agent is submitting applications they should have field-underwritten out. That is coaching, and it is expensive to leave alone because every declined application consumed a full sales cycle.

5. At-risk advance

Total unearned advance across policies still inside their chargeback window.

The only balance-sheet number on this list, and the one agencies discover the hard way.

Advanced commissions are a loan against premium not yet paid. If a policy lapses inside the window, the carrier reclaims the unearned portion. An agency writing volume is carrying a rolling balance of that exposure, and in a growth month it grows faster than income.

Track it weekly, and hold a reserve against it — at-risk advance × expected lapse rate. The calculation, the fields your system needs, and the draft-failure alert that turns a chargeback into a save call are in commission and chargeback tracking.

The reason this belongs on a weekly scorecard rather than a monthly one: draft failures are visible weeks before a lapse. A weekly review catches them while a phone call can still fix it. A monthly review catches them after the carrier already clawed back.

What is deliberately not here

Dials. An input, not a result. Track it for capacity planning; do not review it as performance. High dials with a low contact rate is worse than low dials with a high one, and a scorecard that rewards dials produces exactly that.

Talk time. Same problem, and it actively rewards the wrong behaviour on unqualified calls.

Submitted premium. It is the outcome, and it is lagging. You will see it on the commission statement anyway.

Anything that only moves monthly. Persistency is essential and belongs in the monthly review, not this one — persistency and chargebacks is where it lives.

The review itself

Twenty minutes, same time every week, same five numbers, same order.

Agency level: the five numbers, this week versus a four-week trailing average. You are looking for direction, not for absolute values — a contact rate of 14% means nothing on its own and means a great deal if it was 21% a month ago.

By source: contact rate, issue rate, cost per issued policy. One decision comes out of this every week: keep, renegotiate, or cut.

By agent: speed to contact, contact-to-application, issue rate. One coaching action per agent, named.

Then stop. The failure mode of scorecards is expansion. Twenty-five metrics reviewed monthly produce less change than five reviewed weekly, because five can hold an agenda and twenty-five can only hold attention.

What your system has to support

The five metrics need four fields recorded reliably on every record: lead source, lead delivery timestamp, first attempt timestamp, and policy status through to issue and chargeback window.

If any of those is missing or inconsistently populated, the scorecard cannot be computed and you are managing on impressions. That is the practical reason source tracking has to be enforced at intake rather than cleaned up later — and it is the first thing to verify before a migration, per the field map in switching CRMs.

For a solo agent below the software threshold, all five compute fine from one well-designed sheet — the honest threshold test covers when that stops being true.

Related

Contract levels, overrides and releases for what the production actually pays you. Recruiting agents to a downline for the break-even on adding producers. Your first hire versus your first system for which constraint to spend on first.