Bindable vs Indication: The Gap API Raters Don't Tell You About
An agency owner put it this way to us, unprompted: "The prior products, we tried Tarmika and Bold Penguin, and it would just give you a list of 10 carriers, you know, rated by price. And then we would essentially link over to that quote, and maybe update the coverages, and then it finalizes. So a lot of times, those numbers are way off."
The numbers are way off. That's the complaint, and it's specific. Not "the tool was slow" or "the interface was clunky." The number the tool showed him, the premium on the comparison screen, wasn't the premium he ended up with once he bridged to the carrier and ran the quote for real.
This is the thing the comparative-rater category has a hard time talking about, because their marketing is built around the premise that the rater number is the quote. In practice, for a large portion of the carrier list, it isn't.
The Two Kinds of Rates
Tarmika's public documentation is explicit about this. Their carrier grid labels every carrier with a rate type:
- B (Bindable) — a real quote the agent can bind
- I (Indication) — an estimate only; the agent has to log into the carrier portal to get the bindable quote
- P (Bind/Issue Policy) — the rare case where policy issuance happens on the platform
Bold Penguin, Commercial Submissions, and Semsee have similar distinctions, just named differently. Indication rates dominate the grids. Bindable-on-platform is the exception, not the rule. (See how quote, indication, and binder are defined in commercial insurance.)
The marketing doesn't always surface this. A typical demo of an API rater shows ten carriers in a side-by-side comparison with premiums and coverage summaries. The demo doesn't pause to explain that for six or seven of those ten, the agent will have to leave the comparison screen, log into the carrier portal, answer additional questions, wait for MVRs and credit pulls, and then get the actual bindable number — which will often be different from the number on the comparison screen.
Why the Number Changes
The agency owner quoted above walked through the exact mechanic: "The API brings back kind of generic rates, and then we have to bridge over and finalize it. Typically, we have to bridge over to the quote, say Progressive, we'll bridge over, and then it still has to run reports... run the MVRs for the auto, run the reports, and then that has an impact on the premium."
Three things happen when the agent bridges from the API rater to the carrier portal:
-
The carrier runs external data reports. MVRs on drivers, credit pulls on the insured, loss runs from prior carriers. The rater doesn't have access to these reports — or the API call returned a rate without them. The carrier portal pulls them, and the pricing model adjusts.
-
The carrier applies its own underwriting rules. Some of these rules are encoded in the API call; many are not. State-specific surcharges, agency-specific underwriting tiers, carrier-specific moratoriums on certain class codes. All of these can adjust the rate between the API response and the portal-facing quote.
-
The agent has to answer supplemental questions. These are the carrier-specific questions that didn't fit into the standard ACORD 125 + 126 + 140 data model. Fire sprinkler systems, security cameras, signed contractor agreements, named drivers added after the quote started. Each answer adjusts risk, which adjusts premium.
By the time all three have happened, the rater number is historical. The portal number is real. And the portal number is different.
Where This Leaves the Agent
Agents using API raters for commercial lines end up doing a kind of two-pass quoting: first pass to shortlist carriers based on the API indication, second pass in the carrier portal to get the actual bindable number. The agent hasn't saved time compared to going straight to the portal. They've just moved the work.
In some cases, they've added work. If the API rater's shortlist is based on inaccurate indications, the agent might bridge to a carrier, run reports, and find out the real number is 30% higher than the API said — at which point the carrier is no longer competitive and the whole exercise was wasted.
For lines where the rater has bindable-on-platform carriers (the P and B carriers), the story is better. The agent can stay in the rater for those carriers. But the bindable list is smaller than the total carrier list, and in commercial lines with any complexity, it's usually the smaller half.
What Bindable-From-the-Portal Actually Looks Like
AI agents that operate carrier portals don't have this problem. When the agent submits the ACORD data through an AI-driven portal session, the reports run in real time, the supplementals get answered, and the number that comes back is the carrier's actual bindable premium. There's no bridge-over step, because the AI was in the portal the entire time.
This is a structural difference, not a tooling difference. API raters return rater-layer estimates; portal automation returns portal-layer bindable quotes. Both have costs — portal automation takes longer per carrier because the portal runs reports, and rater APIs are faster because they skip reports. But the speed advantage of the rater is only real if the indication is close to the final number. When it isn't, the agent pays the time cost twice.
Frequently Asked Questions
What is a bindable quote in commercial insurance?
A bindable quote is a premium figure the agent can accept on the customer's behalf to initiate policy issuance. The carrier has run all necessary underwriting checks (MVRs, credit, loss runs, appetite validation) and is committing to the number. An indication, by contrast, is an estimate that can change before binding.
What is an indication rate?
An indication rate is an estimated premium returned by a carrier or rating engine before full underwriting is complete. It's useful for shortlisting carriers and giving the insured a ballpark, but it's not a firm offer. The agent must still complete the full submission in the carrier portal to receive a bindable quote, and the final number can be higher or lower than the indication.
Why are some Tarmika quotes indication-only?
Carriers choose which rate types they expose through Tarmika's API. For many carriers, the full underwriting model (including MVRs, credit data, and supplemental question flows) isn't exposed via API. Those carriers return indication rates to the Tarmika interface and require the agent to bridge to the carrier portal for the bindable quote.
Do all API comparative raters have this indication-only issue?
Yes. The pattern applies across Tarmika, Bold Penguin, Semsee, and Vertafore Commercial Submissions. Any tool that relies on carrier APIs for rating inherits the carrier's decision about what to expose via API, and most carriers don't expose full bindable underwriting to third-party APIs. Portal-based automation is the alternative that doesn't face this constraint.
How much do premiums typically change between indication and bindable quote?
Premium variance between indication and bindable is hard to measure publicly because the rater companies don't publish this data. Practitioner reports suggest variance of 10-30% is common for personal auto (driven by MVR and credit pulls), with narrower variance in BOP for simple risks and wider variance in complex commercial lines where supplemental questions drive a lot of the final pricing.
What We're Building
QuoteSweep's AI agents operate carrier portals directly, so the number that comes back is the carrier's actual bindable premium — not an API indication that still needs portal work to finalize. If you've been burned by indication-based quotes that moved on bridge-over, this is the approach that doesn't have that problem.
