Bought leads or your own: the math for an insurance agency

Every agency principal we talk to has the same line on the profit and loss statement that keeps growing: bought leads. The question we get is whether to keep paying for them or build demand the agency owns. It is a fair question, and this month there are better numbers to answer it with than usual, because two of the reports agencies plan with came out within weeks of each other.

What a search lead costs in this category

The Finance & Insurance row in WordStream by LocaliQ’s 2026 Google Ads benchmarks, released in May from twelve months of campaigns ending March 2026, reads like this at the median:

• Cost per click: $3.39, against $5.42 across all industries.
• Click-through rate: 9.83%, against 6.64%.
• Conversion rate: 2.64%, against 8.18%.
• Cost per lead: $74.44, against $66.69.

A year earlier, the 2025 edition had the same category at $83.93 per lead and a 2.55% conversion rate. So the cost of a search lead came down, but the conversion rate is still about a third of the all-industry figure.

One caution before anyone does arithmetic on those figures. Each median is calculated separately across thousands of campaigns, so dividing the cost per click by the conversion rate will not give you the cost per lead. Read them as three signals: clicks are cheap, people click readily, and few of them hand over their details. That last one is the shopper comparing several agencies before choosing one, and it is the number your marketing has to move.

What you are buying from a marketplace

EverQuote’s annual report for 2025, filed in February, describes its business plainly: a marketplace connecting insurance providers to “high-intent, pre-validated consumer referrals”, with about 60 carriers and about 6,000 agents in its network. Its two largest customers accounted for 38% and 11% of revenue.

Read that from the agent’s side. The shopper you buy is a shopper carriers also buy from the same marketplace, and the biggest buyers there are far larger than your agency. That does not make marketplace leads worthless. It means three things decide whether they pay:

• Speed. A shared lead called back hours later has usually been quoted by someone else already.
• The phone. Invoca’s 2025 analysis of 60 million calls found 29% of financial services phone leads convert during the call, against 37% across industries.
• Consent. The 11th Circuit vacated the FCC’s one-to-one consent rule in January 2025, but autodialed or prerecorded telemarketing still needs prior express written consent under 47 CFR 64.1200. Ask every vendor for the consent record behind a lead before you dial it. For Medicare leads, CMS allows beneficiary data to pass between marketing organizations only with written consent that lists each recipient.

What your own demand looks like

CallRail’s report covered 1.1 million small-business leads, financial services among them. The top channels driving calls and texts were Google Ads at 37%, the Google Business Profile at 23% and organic search at 22%. For qualified leads, Google Ads led at 47%, with organic search at 23% and the profile at 15%.

Those channels have one thing marketplace leads do not: the shopper asked for you. Nobody else was sold the same request, and your forms collected the consent, so the record is yours.

The independent channel is also gaining ground with the people you would be marketing to. The Big “I” 2026 Market Share Report, released on June 23, puts independent agencies at 62% of all property and casualty premium, 87.7% of commercial lines and 39.5% of personal lines in 2025, up from 36.7% in 2021. Shoppers are already looking for agencies like yours. The work is making sure they find yours.

Your book is the cheapest demand of all

Before comparing paid channels, count the one you already own. The Bureau of Labor Statistics consumer price index for motor vehicle insurance rose 50.9% from August 2021 to August 2024 (our arithmetic on the published index). Increases like that send clients shopping at renewal, and many agencies only hear about it when the cancellation arrives.

The index has started to ease: from August 2025 to May 2026 it fell about 1.9%, again by our arithmetic on the published figures. An easing market is when a renewal review reads as service rather than damage control, and when clients who left over price are worth a call.

Cross-sell belongs in the same plan. CMS lists calling your own auto and home clients about Medicare plans you sell as permitted business, so a P&C book is also a Medicare list, with a recorded Scope of Appointment before the sales conversation.

How to compare the sources fairly

Cost per lead flatters marketplaces and punishes search, or the other way round, depending on the month. The comparison that holds up is cost per bound policy, by source. For each source, track:

1. What you spent in the month.
2. How many leads arrived, and how many you actually reached.
3. How many you quoted.
4. How many bound, and the premium written.

Divide spend by policies bound and you have a number that means something. Do it for three months before cutting any source, because a quiet month for home closings or a busy one for renewals can tilt a single month’s result.

What we are telling clients

• Add a source field to every lead record today: bought, referred, search, profile or book. Without it, none of this math is possible.
• Set a callback target for bought leads in minutes, not hours, and measure it weekly.
• Ask each lead vendor for a sample consent record, and stop buying from any vendor that cannot produce one.
• Run a renewal review campaign to clients with increases in the next 60 days before raising any paid budget.
• Shift budget a step at a time from the worst source per bound policy to the best, and recheck after a full quarter.

We have set out how the channels rank for an agency, and where state and federal rules shape each one, on our insurance agency page. If search is the source you want to grow, here is how we run Google Ads so every lead is tracked from click to bound policy.