The auto-to-home cross-sell gap is the largest piece of unwritten premium in a personal lines book — and it is not a producer-discipline problem.
A producer takes an inbound auto lead at 9:14 AM on a Tuesday. Name, phone, ZIP, vehicle year, prior coverage. She works it properly — quotes it same day, follows up twice, binds it Thursday morning. Good week.
She never asked about the house.
Not because she was lazy, and not because she needed more training. Because nothing on her screen told her there was one.
The gap has a number now
J.D. Power’s 2026 U.S. Insurance Shopping Study — 12,437 shoppers, fielded January 2025 through January 2026 — found that 45% of people actively shopping an auto policy already carry homeowners coverage. Only 20% received a homeowners quote during that shopping process.
That is a twenty-five point gap, and it is not a rounding error. Read it as an agency owner and it says something uncomfortable: a quarter of your inbound auto volume is a household you were entitled to quote twice and quoted once.
You already paid to acquire these households. The second policy costs nothing in acquisition — only attention you never got the chance to spend.
This is not a producer problem
The instinct is to fix it with a script. Add a line to the intake sheet. "Ask every caller whether they own or rent." Agencies have been running that play for twenty years and it never sticks, for a reason that has nothing to do with discipline.
Asking "do you own or rent?" in the middle of a quote reads as a sales move. The prospect hears upsell and puts a hand up. But a producer who opens the lead already seeing homeowner status and an estimated property value is having a completely different conversation — one that sounds like advice instead of a pitch.
Same information. Opposite reception. The variable is whether the producer discovered it on the call or already knew it before dialing.
The window is narrower than it used to be
The same study found shoppers now pull an average of 3.5 quotes — the highest in the study’s twenty-year history — and that 48% of new auto policies are now purchased digitally, up from 36% five years ago.
J.D. Power’s Stephen Crewdson described the mechanic plainly: most customers are only shopping their auto policy, and if the auto quote is not competitive, they do not stick around to discuss home or anything else.
Turn that around and look at it from the agency side. The bundle is not a follow-up call. It is not a renewal-season review. It surfaces inside the one auto conversation you get, or it does not surface at all — and you are now one of three and a half people that household is talking to this week.
Run the math on your own book
Say you bind 60 personal auto policies a month from leads. Apply the study’s distribution and roughly 27 of those households already own a home. You are quoting home on about 12 of them.
That leaves fifteen homeowners a month you wrote auto for and never quoted. Fifteen a month is 180 a year, sitting inside a book you have already paid to acquire.
You do not need to close all of them. Capture a quarter of the ones you would otherwise have missed and that is roughly 45 additional home policies a year — plus the multi-policy discount that makes the auto policy meaningfully harder to unseat, plus a household the next aggregator buyer has a much worse chance of peeling off at renewal.
Illustrative figures. Substitute your own bind volume and current bundle rate — the shape of the answer does not change much.
Curious what that gap actually looks like in your own book? Get a Free Lead Audit
The same gap runs in reverse
Ask any home producer about renters. A lead arrives tagged as a homeowner's inquiry. The producer preps, pulls comps, dials — and finds out four minutes in that the person rents and always has. That lead cost the same as every other lead in the batch. It just cost four minutes of a producer’s morning on top of the invoice.
Both problems are one problem. The record does not say who owns and who does not. One version of it costs you revenue you never saw. The other costs you time you cannot bill. Neither shows up on a report, which is exactly why neither gets fixed.
Enrichment at intake is a timing argument, not a data argument
Homeowner status is not a field the quote form asked about. Neither is estimated property value, nor whether there is an existing auto policy sitting with another carrier. None of it is self-reported, because nobody filling out a thirty-second form volunteers information that might cost them a better rate.
But that data is appendable, and it always has been. You could buy a list append tomorrow. So the constraint was never availability.
The constraint is when. A homeowner flag that lands in a Tuesday batch report is a fact. The same flag sitting on the record when a producer picks up at 9:14 Monday is a second policy. Identical field, opposite outcome — and the only variable is whether it arrived before the conversation or after it.
That is what enrichment at intake means. The record gets completed on the way in — between the moment a vendor fires the lead and the moment it reaches a producer’s queue. Not on a schedule. Not in a monthly cleanup. Not when somebody remembers to run a report.
How LeadArray does it
Every source lands in the same pipe. EverQuote, MediaAlpha, QuoteWizard, Facebook lead ads, your own web forms, direct-mail response batches — they arrive in different shapes, with different fields, and different degrees of honesty. They get normalized into one record format before anything else happens to them. A homeowner flag that only exists on your best vendor’s leads is not a solution. It is a coverage gap you cannot see from a report.
The append is a waterfall, not a single lookup. No one data source has every household. When the first does not return a match on homeownership or property value, the next is tried, and the next. Coverage varies by signal and by household — prior-carrier detail fills less consistently than homeownership does, and any vendor telling you otherwise is selling. But a waterfall is the difference between a field a producer half-trusts and a field a producer acts on.
It runs upstream of your CRM. Validation, append, and screening all happen before routing — so the producer’s queue is the output of the process, not the input to it. Your CRM or AMS is a destination, not the place where leads get fixed.
It lands as native fields. Homeowner status, estimated value range, prior-carrier signal, and the intent score are written into your existing system as real fields — not a PDF, not an attachment, not a second tab a producer has to go open. If someone has to look it up, they will not.
And for a four-producer agency, the part that decides it: none of that is something you configure. Source integration, enrichment setup, scoring rules, and routing logic are handled for you. There is no Marketing or RevOps hire hiding in this.
See the full intake process on insurance leads specifically: LeadArray for Insurance & Financial Services
Why not one of the other four ways
Agencies do try to close this gap. There are four common approaches, and each of them works up to a point worth naming honestly.
Upgrade to the vendor’s premium tier. Perfectly reasonable, and if you run a single source it may be all you need. The catch is that it only fixes that source. Most agencies run four to six, which means buying up on each one — paying repeatedly for partial coverage, in four different field schemas, all landing in the same CRM.
Run a monthly append against the CRM. Cheap, and genuinely useful for book-level campaigns and renewal targeting. But it is a rear-view mirror. The household you needed the homeowner flag for bound their auto three weeks ago and never once heard the word bundle.
Build it yourself. An enrichment API, an automation layer, and one person who owns the whole thing. This genuinely works — I have watched agencies run it well, with more control over the logic than any vendor will give them. It also requires somebody who wakes up thinking about field mapping and match rates, and the day that person leaves, your intake pipeline leaves with them. Maximum control, permanent staffing dependency. That is the honest trade.
Use your CRM’s built-in enrichment. HubSpot and GoHighLevel both do real work here, and if you are already paying for it you should use it. But both operate on records that are already in the system. That is the wrong side of the line for this particular problem: by the time a record is in the CRM, it has been routed, a producer has been assigned, and the decision about who to call first was made on incomplete information.
Which is the whole argument in one sentence. Every one of those approaches fixes the lead after somebody has already decided what it is worth. Enrichment at intake decides what it is worth before anyone picks up the phone.
Where to start this week
You do not need a platform to find out whether this is your problem. You need one afternoon.
• Pull every personal auto policy you bound in the last 90 days.
• Count how many of those households also carry a homeowners policy with your agency.
• Divide. If the number lands under 40%, the gap in the study is your gap too.
And if it does, resist the urge to make it a producer conversation. Fifteen missed households a month is not a coaching issue. It is a record that showed up incomplete, fifteen times, and nobody on your team had any way of knowing.
The leads were fine. What arrived with them was not.
More on why cascading across sources beats a single provider: What Is Waterfall Enrichment and Why It Beats Single-Source Data.
Send us a sample of your inbound leads, and we will show you exactly what would have been appended, scored, suppressed, or routed before it ever hit your dialer. Get a Free Lead Audit
Sources
J.D. Power, 2026 U.S. Insurance Shopping Study (released 4 June 2026); 12,437 respondents, fielded January 2025 through January 2026.
Comments