Mapping the quote-to-bind funnel

Two auto carriers report the same quote-to-bind rate: 7%. At the first, a quarter of applicants never see a price at all, because screen two asks for a VIN. At the second, almost everyone gets a price and then walks away from it. Identical headline number, two completely different meetings to book: one with engineering, one with pricing. The blended rate cannot tell you which carrier you are.

This lesson splits the path from quote start to policy in force into stages you can measure one at a time, and separates the drop-off that tells you something from the drop-off that is the system doing its job.

The stages

Most personal lines funnels have five measurable points, not two:

1. Quote start: a visitor submits enough to trigger a price calculation (name, postcode or ZIP, vehicle or health details).
2. Quote completion: the rating engine returns a price. This sounds automatic. It isn't.
3. Referral: a share of completed quotes cannot be auto-priced and route to a human underwriter, or bounce back as a decline. Prior claims, a modified or high-value vehicle, an address the geocoder dislikes, a non-standard occupation, an applicant outside the appetite for that line.
4. Quote viewed: the prospect opens or re-engages with the price (clicks in, scrolls, opens the email or SMS link).
5. Bind: payment is set up and cover is in force.

Each transition is a conversion rateconversion rateThe percentage of visitors or prospects who complete a desired action (purchase, sign-up, contact form), calculated as conversions divided by total opportunities.View full definition →, and their product is the quote-to-bind rate.

Quote-to-bind rate = Binds ÷ Quotes started

One warning before the diagnostics: bind is not the terminal state. Policies cancel inside the cooling-off period, fail first direct debit, or get repriced when documents (licence, no-claims proof) contradict what was declared. Measuring at day 30 rather than at bind is the honest denominator, and the two differ by enough to change a channel decision.

Where the drop-off actually happens

Quote start → quote completion

Industry estimates (2024-2025) put 15-25% of started US auto quote forms as never reaching a delivered price, mostly form length, partly data the applicant cannot supply cleanly. Progressive, which sells the very product under discussion here, has attacked this from the price end rather than the form end: its Name Your Price tool inverts the sequence by asking what the prospect wants to spend and assembling coverage to fit. That moves a price rejection from the last stage of the funnelfunnelThe customer journey from awareness to purchase, typically Awareness, Interest, Consideration, Decision, Action, with prospects narrowing at each stage.View full definition → to the first, where it costs nothing.

Fix pattern: progressive disclosure (three or four fields, an indicative price, refinement afterwards) beats a single long form on completion. The second-order cost is real, though. Shorter forms mean thinner data at rating, more corrections later, and a higher share of binds that unwind at document check.

Referral and decline

This stage is invisible in most marketing dashboards because it belongs to underwriting. It shouldn't be. A referral that takes 48 hours to come back has already lost the customer to whoever quoted instantly.

Cuvva, the UK app insurer selling short-term and subscription motor cover, handles this by gating hard at the top: the app tells a driver it cannot cover the vehicle before they have spent five minutes entering details. That looks like a worse funnel (fewer quote starts converted) and is a better business, because the alternative is a warm applicant left waiting on a manual review for cover they wanted to start this afternoon.

Treat the referral rate and the referral turnaround time as marketing metrics. If referrals run above roughly one in ten completed quotes on a line meant to be straight-through, the rules engine, not the media plan, is capping your growth.

Quote completion → quote viewed

A quote can be technically delivered (price computed, email sent) and never opened. In life insurance the gap is wide, because the quote often is not a final price: a range, pending a health questionnaire or a medical exam. Distribution platform estimates suggest only around 40-60% of delivered life quotes are ever opened, against a much higher rate for instant-bind auto.

Fix pattern: simplified issue products with algorithmic underwriting, plus SMS rather than email for the reminder. Both work by removing delay, not by adding persuasion.

Quote viewed → bind

Benchmarks for US direct-to-consumer auto (estimates, varying by source and carrier, circa 2023-2025) sit somewhere in the 8-15% range, with 20%+ for agent-assisted or warm referral, because a human closes objections a self-serve flow cannot. Online term life runs lower, often cited at 3-8%.

Diagnostic drop-off versus cosmetic drop-off

Not every leak deserves a fix. A decline at referral for a driver with three at-fault claims is margin protection; retargetingretargetingShowing ads to users who have previously visited your site or interacted with your brand, to bring them back and drive conversion.View full definition → that person harder buys loss ratio, not growth. A fall in quote starts after you removed a "quote in 60 seconds" claim is cosmetic if binds held flat, and it may have improved lead quality.

The drop-offs that are diagnostic are the ones that moved without a corresponding change in mix: completion rate falling in a single browser, viewed-to-bind falling in one state the week after a rate filing took effect, referral volume spiking on a vehicle group after a rules update. Those have a cause you can name.

A worked example

A US auto insurer runs paid search and generates 10,000 quote starts in a month.

• Completion: 82% → 8,200 completed forms
• Of those, 7,400 priced instantly, 800 referred; 300 referrals return a price, 500 lapse or decline → 7,700 priced quotes
• Quote-viewed: 70% → 5,390 viewed
• Viewed-to-bind: 12% → 647 binds

Overall quote-to-bind rate = 647 ÷ 10,000 = 6.5%

At $150,000 of campaign spend, that is roughly $232 per bind. Now apply a 5% day-30 fallout (non-payment, cooling-off cancellation, repriced and refused): 615 policies in force, and the same spend becomes about $244. A 5% difference in denominator, and it is the smaller number that everything downstream depends on. That per-policy figure is the acquisition costacquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition → the channel lesson decomposes; whether $244 is affordable is settled against the multi-year value the CLVCLVLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition → lesson models, not against the premium on the first invoice.

Why funnel mapping matters more than single-metric tracking

A falling blended rate could mean worse traffic quality, a technical delivery failure (a new required field broke completion), a rules change pushing more quotes into referral, weaker follow-up on viewed-but-unbound quotes, or genuine price uncompetitiveness. Five diagnoses, five different owners: media, engineering, underwriting, CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition →, pricing. Stage-level segmentationsegmentationDividing a market into distinct groups of customers who share similar needs, characteristics or behaviours, so each group can be served with a tailored approach.View full definition → turns "conversion is down" into a brief someone can act on this week.

Channel and product differences worth remembering

• Price comparison sites generate high quote-start volume at low per-carrier bind rates, because the prospect is quoting with five to ten competitors at once. Carriers bidding there must price acquisition assuming that lower close rate from the start.
• Owned direct channels convert better after the quote (the prospect already chose the brand) but produce smaller, more expensive top-of-funnel volume.
• Agent-assisted channels show the highest viewed-to-bind rates and the highest cost per lead.
• Embedded distribution breaks the funnel model entirely. ZhongAn, the Chinese digital-only insurer, built its early volume on shipping return insurance attached at Taobao checkout: hundreds of millions of policies a year at a few jiao each, with no quote form, no viewed stage, and effectively one click between offer and bind. The metric that matters there is attach rate on the host platform's transactions, and the whole economics rest on someone else's checkout flow. Lose the placement and there is no funnel to optimise.

In Europe, comparison sites carry far more of the motor funnel than in the US, and the UK's Financial Conduct Authority has scrutinised their pricing practices and add-on sales, which constrains how quotes may be presented and bundled.

🎬 [VIDEO: "How Insurance Companies Make Money (and Lose It)" - youtube.com/@PatrickBoyleOnFinance - a plain-language breakdown of insurance economics that helps non-technical learners see why acquisition efficiency matters to the business model]

A simple way to instrument this yourself

Tag each stage as an event and compute stage-over-stage rates:

SELECT
  COUNT(DISTINCT CASE WHEN stage = 'quote_started' THEN user_id END) AS started,
  COUNT(DISTINCT CASE WHEN stage = 'quote_completed' THEN user_id END) AS completed,
  COUNT(DISTINCT CASE WHEN stage = 'referred' THEN user_id END) AS referred,
  COUNT(DISTINCT CASE WHEN stage = 'quote_viewed' THEN user_id END) AS viewed,
  COUNT(DISTINCT CASE WHEN stage = 'bind' THEN user_id END) AS bound
FROM funnel_events
WHERE event_date >= '2026-01-01'

Add a sixth column for policies still in force 30 days after bind, and the dashboard stops flattering you.

Key Takeaways

• Track five stages, not one blended rate: quote start, completion, referral, viewed, bind. Then check day-30 in force, because binds unwind.
• Completion failures are the underrated leak, estimated at 15-25% of US auto starts. Shorter forms fix it and push data qualitydata qualityThe degree to which data is fit for purpose: accurate, complete, consistent, timely, valid and unique. Poor quality data undermines analytics, reporting and AI.View full definition → problems downstream.
• Referral rate and referral turnaround belong on the marketing dashboard; a 48-hour manual review loses an applicant who could get an instant price elsewhere.
• Life insurance leaks hardest at quote-viewed, because the quote is a range pending underwriting rather than a price.
• Rough benchmarks: 8-15% quote-to-bind for US direct auto, 3-8% for online term life, 20%+ agent-assisted. Embedded models like ZhongAn's checkout attach make these ratios meaningless and substitute a different one.
• Ask whether a drop-off has a nameable cause before funding a fix. Declines at referral are the system protecting margin.