Speed to lead is the time between a lead arriving and the first attempt to reach them. In most insurance agencies, leads go cold because of how the work is set up: more leads arrive than licensed agents can call, alerts never reach the agent, or the lead lands with someone who doesn't write that product. You can measure your real response time in a week by mystery-shopping your own agency, then fix most of it with pacing, routing checks and a written handoff rule.
Say you spend $3,000 a month on leads. If a third of them wait until tomorrow for a first call, you've paid for a lot of people who already talked to someone else. The fix starts with knowing which of the three causes you have, because each one needs a different answer.
How fast do insurance leads go cold?
Faster than most owners assume. Researchers writing in Harvard Business Review's "The Short Life of Online Sales Leads" sent a test lead to 2,241 US companies. Only 37% responded within an hour, 23% never responded at all, and the companies that did respond took 42 hours on average. Firms that tried to contact a lead within an hour were nearly seven times as likely to qualify it as firms that waited longer, and more than 60 times as likely as firms that waited 24 hours or more.
That study covers online leads in general, not insurance, and it dates from 2011. We still see the same pattern in agencies we work with. Someone who fills in a form at lunch is talking to whoever called first by dinner.
Before you measure anything, agree on the definition. At Kevlex we count speed to lead from the moment the lead lands to the first outreach attempt: a call, a text or a voicemail. A first conversation can come later. If you wait to count until someone picks up, you're measuring the prospect's schedule along with your agents'.
Reason 1: more leads arrive than your agents can call
Speed to lead is a capacity question before it's a discipline question. Most lead vendors push a lead to your agent the moment it lands in their system, so you can't ask them to stagger delivery. Your levers are which sources and states you switch on, and when your agents work.
Say a vendor drops 25 leads into one agent's queue between 10 and 11 a.m. A first attempt (dial, leave a voicemail, send a text, write a note) takes about six minutes. The agent can make ten first attempts an hour, so the 25th lead waits about two and a half hours, even if the agent does nothing else. Add one booked appointment in the middle of that hour and the wait goes past three.
Check your own numbers this way:
- First attempts per agent per hour = 60 ÷ minutes per attempt.
- Peak leads per hour, per source. Use the busiest hour of the week, not the daily average.
- If the peak is higher than what your licensed agents can call, the wait grows with every lead in the queue.
- Note which states the peak comes from, and whether those agents are licensed and working at that hour.
If one source floods the queue, switch it off for a few weeks and watch what the other sources do. That costs you some volume, and it shows you what one agent can handle. Time zones belong in the same check. An agent in Mountain time whose leads come from the East Coast should start the day earlier, because those leads arrive two hours before the agent's usual start time.
The source you pick also decides how much of your volume can be worked at all. We compare the trade-offs in insurance lead sources compared.
Reason 2: the lead never reached the agent
Some leads go cold because nobody knows they exist. The form works, the vendor sends the lead, and somewhere between that system and your agent's phone the data drops out.
"We send the data, and then it's basically like, putting an envelope through the thing, and then nobody's there to take it from them." — Kevin Beeftink, founder of Kevlex Digital
In one case we looked at, leads passed to a partner office hit an integration that created a contact with a name and nothing else. No source, no notes, and no alert to the agent. The leads sat in the insurance CRM as bare records, and every report counted them as assigned.
Ask these questions about each lead path:
- Does the full record arrive (name, phone, email, state, source, what they asked for)?
- Does the assigned agent get a text or push alert they'll see away from the desk?
- Does anyone get told when an assigned lead has had no activity after a set time?
- Who checks the integration after a form, vendor or GoHighLevel workflow changes?
You answer all four by sending a test lead and watching what happens to it.
Reason 3: the lead went to someone who won't work it
The third cause is the hardest to see in a report. An agent who doesn't write Medicare gets a Medicare lead. An agent with a large book and a full calendar gets the same lead volume as one with 20 clients. Neither ignores the lead on purpose. It drops down the list behind work they know how to do and get paid for.
"there should be a system in place that says, hey, if there's a Medicare lead that comes in, this is how we get it written by whoever is going to write it." — an agency principal we work with
Picture the agent's side. A lead for a product they don't sell comes in. There's no rule for who takes it, so they have to work out who to send it to, copy the details into a text, log it somewhere, and remember to claim any referral credit. Then the phone rings with one of their own clients, and the lead waits until tomorrow.
The fix is a written rule per line of business, plus a handoff that takes one click. In GoHighLevel that can be a button or a pipeline stage that reassigns the contact, alerts the receiving agent and records who referred it. If passing a lead on takes five steps, expect it to happen on slow days only.
How to mystery-shop your own speed to lead
A mystery shop tells you what a prospect experiences, which a CRM report can't. The HBR researchers used the same method: they submitted test leads and timed the replies. You can run it across your whole team in a week.
-
Decide what good looks like first. Write down the response time and number of attempts you expect, per agent and per lead type, before you see any results.
"Let me ask you this. What would you have liked to have seen from this test? What would have been the perfect picture?" — Kevin Beeftink, founder of Kevlex Digital
-
Pick a shopper the agents don't know. Give them a phone number and email address that aren't in your CRM, and a realistic prospect profile. Tell the team that tests will happen this quarter, without saying when.
-
Submit through the real lead path. Use the same web form, vendor route or integration a real lead takes. Send one lead per agent, at different times of day, and include your busiest hour.
-
Log every touch with a timestamp. Calls, texts, voicemails and emails, plus who made them.
-
Test the callback. Have the shopper miss the first call, then call back within two minutes. Note whether anyone picks up. Try both the main line and the agent's direct number.
-
Score each agent against the targets from step 1.
-
Repeat monthly with a new profile, so you can see whether changes stuck.
A simple log is enough:
| Field | Example |
|---|---|
| Agent | Agent A |
| Lead path | Website form, Medicare quote |
| Lead submitted | Tue 10:14 a.m. |
| First attempt | Tue 10:51 a.m., call, voicemail |
| Attempts in first 48 hours | 4 (3 calls, 1 text) |
| Callback answered? | No, went to voicemail |
| CRM record complete? | Yes |
| Minutes to first attempt | 37 |
When you review results, look past the agent. Kevin's view from these tests is that the logistics deserve as much attention as the agent's performance. A slow agent may be sitting behind a broken alert, and a fast agent may be ignoring a whole lead type.
Check the same numbers on your real leads
The mystery shop gives you a sample. Your CRM gives you the full picture, if it records the right times. For the last 30 days, pull the time each lead was created and the time of the first outbound call or message. Then work out:
- The median minutes to first attempt, per agent and per source
- The slowest 10% of leads, and what hour of the day they arrived
- How many assigned leads have no activity at all
- How many callbacks went unanswered
Watch for one data gap. Some CRMs, GoHighLevel among them, don't record when a tag was added. If your process marks "contacted" with a tag, you lose the timing. Log the first attempt as a call, a message or a dated field that a workflow fills in, or keep a separate event log. We do the latter for clients who need clean reporting.
How to fix speed to lead, step by step
Once you know which cause you have, work through the fixes in this order:
- Write the definition and the targets down. Lead arrival to first attempt, with a target per lead type. A warm web lead and a lead that's two years old don't need the same target.
- Match intake to capacity. Turn off or cap sources whose peak hour outruns your licensed agents. Bring them back when you add people.
- Match agent hours to the lead's time zone.
- Test every lead path weekly. Send a test lead, confirm the full record and the alert arrive, and set an escalation: if an assigned lead has no activity after, say, 30 minutes during working hours, the manager gets pinged or the lead moves to the next agent.
- Write a handoff rule per line of business and make the handoff one click.
- Automate the first reply, then call. A text or email within a minute tells the prospect you got their request. Only text people who gave TCPA consent. A person still makes the call.
- Make callbacks ring someone. Route missed-call returns to whoever is free if the assigned agent doesn't answer.
- Review it every week with a short scorecard.
Automation buys you minutes. Kevin's rule for clients running social or web leads: after the first automated response, a person has to take over, and fast. If the lead calls in, someone has to pick up the phone.
What happens after the first call?
Speed gets you the first touch. The lead nurture sequence and the agent's persistence decide whether the lead becomes an appointment.
"Because if you buy 30 leads, she needs to exhaust that list until they scream no at her, right? So how do we make sure the number of touch points is adequate?" — Kevin Beeftink, founder of Kevlex Digital
Most agencies call once a day, if that. A cadence we'd start from:
| When | What the agent does |
|---|---|
| Day 1 | First attempt within your target, text after a missed call, second call in the afternoon |
| Days 2 to 4 | Morning and afternoon calls at different times, one text a day |
| Days 5 to 14 | One call a day, alternating morning and late afternoon, plus email |
| Day 15 onward | Automated nurture, with a call when the contact opens or clicks |
Two rules sit on top of the table. A callback is the warmest moment that lead will ever have, so drop what you're doing and take it. And "I didn't fill that out" isn't a request to stop. That person hasn't told you no, so call again another day. A clear "don't contact me" goes on your internal Do Not Call list the same day.
Leads that go quiet after a few weeks still have value. Move them to a lead reactivation campaign instead of deleting them. The same follow-up rules apply to event registrants, which we cover in seminar follow-up for Medicare agents.
How do you keep response times fast after month three?
Most agencies fix speed to lead once, in a burst of attention, and then it drifts. Kevin's rule of thumb is that motivation fades about three months after you start something, so the habit needs a weekly management cycle to last.
Put these on a weekly scorecard, per agent:
- Median minutes to first attempt
- Share of leads reached within target
- Attempts made in the first 48 hours
- Callbacks missed
- Assigned leads with no activity after 24 hours
- Appointments set from new leads
Add a short call review. Each agent picks their best and worst call of the week and talks the manager through both against your call guide. Run a new mystery shop each month. When a number slips, check the plumbing first (sources, alerts, handoffs) before you talk to the agent about effort.
What to do this week
- Send one test lead to every agent through your real lead paths, and log every touch.
- Pull the last 30 days of leads and find the median time to first attempt, per agent and per source.
- Find your busiest lead hour and compare it with how many first attempts your agents can make in that hour.
- Write a one-line handoff rule for each product your agents don't all write.
- Call your own main line after a missed call and see who picks up.
If you'd like help setting up the routing, the alerts or the mystery shop itself, book a 15-minute call with Kevin. Bring one lead source, and we'll walk through where its leads wait.
