Insurance Vendor Diligence Tracker

Definition

A single running comparison across every third-party insurtech vendor Directions has researched as a potential rail for insurance-tech-integration-stack — one row per vendor, evaluated against the same fixed set of questions every time, so vendors stay comparable instead of each getting a differently-shaped writeup. This page holds only the condensed comparison; full detail, sourcing, and notes live on each vendor’s own company page under wiki/companies/ — this page links out rather than duplicating, per the KB’s one-fact-one-file rule. Workflow: the next time we dig into a new product, add one row to the table below, one subsection under Key points, and update its company page as usual — that’s the whole maintenance loop.

Two categories live here, and they are not comparable to each other: ingestion rails (get the prospect’s existing policies) and quoting/placement rails (get a price from carriers). A third, contributory-database, was added 2026-08-17 when lexisnexis-risk-solutions surfaced — it answers the ingestion question by a mechanism no startup in this tracker can use. Compare within a category, never across.

The standing framework

Expanded 2026-08-17 from 7 to 9 questions, after Reliability and Ownership kept surfacing as findings that didn’t fit cleanly under any existing question across the first four vendors. Every vendor gets evaluated on the same nine:

1. Users & journey — who uses it, and step by step how the flow works.
2. Data persistence — does the vendor store data after a pull/interaction, or is it pass-through? Retention terms?
3. Carrier relationship — how (if at all) do carriers use the vendor, and is user data shared back to/with carriers?
4. Carrier count — how many carriers does it actually connect to (distinguishing live connections from index/marketing claims)?
5. Reliability / completion rate — what % of attempts actually succeed end-to-end, and is that vendor self-reported or third-party-corroborated (G2, etc.)? (added 2026-08-17)
6. API maturity — auth model, docs quality, sandbox, webhooks, schema depth.
7. Ownership & competitive stability — who owns them, funding trajectory, and is their neutrality structurally at risk (acquisition by a conflicted carrier/MGA/AMS, or strategic drift into becoming a competitor)? (added 2026-08-17 — motivated directly by insurance-distribution-aggregator-layer’s finding that every neutral rail reaching distribution relevance since 2021 has been acquired by a conflicted owner)
8. Pricing — tiers, metering, what commercial/premium features cost.
9. Our use case — if Directions opened a brokerage tomorrow, what would this vendor concretely do for us, and where does it not help?

Vendor comparison table

Vendor	Category	Mechanism	Carrier count (claimed)	Commercial lines	Reliability	Pricing	Ownership & funding	Our use
canopy-connect	Policy ingestion (personal + commercial)	Consumer-permissioned credential login (Plaid-style)	300–400 (site inconsistent), ~95% market claimed	Yes, credential-pull attempted — G2-flagged as unreliable	Self-reported ~70% credentials-to-full-pull; G2 independently confirms commercial pulls fail more often	$100–600/mo,metered"quotepoints";commercial=2pts/pull($4–6 marginal)	$6.5M Series A (Oct 2023), no round since — independent, founder-led, capital-disadvantaged vs. Axle	Primary ingestion rail for the check-up funnel; strongest on personal lines
insurgrid	Policy ingestion (personal credential-pull + commercial via upload only)	Same credential-login mechanism for personal; document upload + AI extraction for commercial	150 (live) / 350 (press) / 450 (formats parsed) — internally inconsistent	Extraction/upload only — no commercial credential pull	No completion rate published; the carrier-count self-contradiction is itself a marketing-reliability red flag	$99/mo flat, unlimited pulls, 20% off annual	Acquired by Helium Ventures (generalist $500K–$3M-revenue holdco, other 2026 buys are wedding platforms) Jan 2026 — roadmap-durability risk, not a competitive threat	Cost-floor backup for the document door; cannot replace Canopy on commercial credential-pull
herald	Quoting/binding (not ingestion)	Broker submits application data to carrier via API; requires the broker’s own pre-existing carrier appointment	Publishes an 86-carrier/228-product market index (not its own live-connection count)	Yes — this is its core product (BOP, GL, WC, cyber, E&O, D&O)	Not directly comparable (quoting, not pulling); its Connections mostly activate immediately, a minority go “referred” for ~24h carrier/Herald review	Unpublished; enterprise sales cycle	~$20M total, independent, no round since Oct 2024 — but post-Babbix drifting into broker-facing AI products, a strategic-drift risk rather than acquisition risk	Quoting-side rail for P1+, once Directions or its partner brokerage holds real appointments — doesn’t grant appointments itself
axle	Policy verification (personal lines only)	Same consumer-permissioned credential/upload mechanism as Canopy/InsurGrid, via the “Ignition” embed; live monitoring while connection is open	Unpublished — vaguest of any vendor here (“coverage across all major carriers,” no number)	No — funded roadmap item. Series A explicitly earmarks commercial as a 2026 build target	No completion rate published; effectively zero independent review signal (the one G2 listing found appears to be an unrelated “Axle” product)	Fully unpublished — no tiers or per-verification cost found anywhere	~$21.5Mtotal($4M seed 2023 + $17.5M Series A Aug 2026, Base10) — independent, most heavily capitalized personal-lines vendor here	Track, don’t build on: serves lenders/dealers verifying a third party’s coverage, not agencies pulling a prospect’s portfolio; revisit once/if commercial ships
fize	Policy ingestion (personal + commercial auto)	Same consumer-permissioned credential link, plus document upload; “like linking a checking account to Venmo"	"~95% of policies” claimed; no carrier number published	Commercial auto only, and tier explicitly labelled beta	Nothing published; no independent signal	Most transparent in the cohort: $89/$200/$325permobytier,50–150syncs,\ast \ast$2/extra sync**; 7-yr doc history, 14-day trial	Founding, funding, HQ and team all undisclosed — least knowable vendor here	Middle option on cost: cheaper per pull than Canopy, dearer than InsurGrid’s flat rate; commercial too immature to lead with
quotesweep	Quoting (portal automation, not API)	AI agents log into carrier portals in parallel with the agency’s own credentials, handling MFA and CAPTCHAs	~70 actively automated vs “550+ tracked” (internally inconsistent); reaches the ~2,600 portal-only carriers APIs can’t	Yes — BOP, GL, WC, auto, umbrella, professional, cyber, EPLI	No completion rate published; per-portal reliability unassessed	Public and submission-metered: $49/$99 / $299 per mo for 10 / 30 / 100 submissions	Founded 2025 (Ankur, solo-founder-led), funding undisclosed — independent, and the only unowned rater of scale in its lane	Automates Job 4 (portal-raking) directly; still requires our own appointments, so P1+ not P0
lexisnexis-risk-solutions	Contributory database (ingestion, different mechanism)	Carriers pool their own policy data; no consumer consent, no login, no upload	Top-25 commercial auto carriers participating; 37 yrs of contributory operation	Commercial auto only	Claims up to +33% drivers / +22% vehicles found (extrapolated from personal auto)	Unpublished, enterprise	RELX-owned incumbent — zero neutrality risk, zero acquisition risk, and zero access for us	Structurally closed to brokers — carrier-participants only, and contributory access requires a book we don’t have. Track for scope expansion

Key points

Canopy Connect (usecanopy.com) — full detail: canopy-connect

• Three distinct user segments: agencies (core intake product), automotive/lending verification (dealers, rental, lenders — direct Axle competitor), and carriers themselves (buy access for competitive comparisons at their own point of sale).
• Data is stored, not pass-through — retained “as long as necessary,” shared with the “Innovator” whose flow the consumer used; the exact mechanism for carrier-side data sharing is undocumented publicly and is a named open item.
• API is genuinely developer-grade: OAuth-style keys, native SDKs, public sandbox, signed/retried webhooks, 50+ schema objects including commercial-specific ones. Best-documented vendor in this tracker so far.
• Reliability: self-reported ~70% completion, and G2 reviewers independently confirm commercial-lines pulls are the specific weak point — corroborating the design decision to rank document upload ahead of carrier login in the check-up funnel’s intake doors. This is the only vendor here with any independent (not just vendor-stated) reliability signal at all.
• Ownership: no acquisition signal, still founder-led (Tolga Tezel) three years after its Series A — but three years with no new round, against a freshly-capitalized Axle circling the same automotive/lending lane, is a competitive-financing gap worth watching.

InsurGrid (insurgrid.com) — full detail: insurgrid

• Resolves what was an open question as of 2026-08-15: commercial support is extraction-only, never credential-pull — the FAQ scopes carrier-connection strictly to personal lines.
• Cheapest option in this tracker by a wide margin at real volume ($99/mo flat vs. Canopy’s per-pull metering), but the cheapness only covers what it actually does — personal-lines connection plus commercial-by-upload.
• Ownership: acquired by a generalist holdco (Helium Ventures — wedding/event platforms are its other 2026 buys) rather than an insurance strategic. Reads as less of a competitive threat (a holdco will not build a brokerage) but more of a roadmap/support-quality risk.
• Reliability: no completion rate published anywhere; the site’s own carrier-count claims are internally contradictory (150 / 350 / 450 across different pages) — not a pull-failure statistic, but the same category of red flag: it signals under-maintained marketing post-acquisition, which is itself a reliability concern about the vendor, not just the product.

Herald (heraldai.com) — full detail: herald

• Structurally different from the other two: it’s not pulling existing policy data, it’s the pipe for submitting new business to a carrier for a quote. “Does it persist / share with carriers” doesn’t apply the same way — carrier-sharing is the product itself.
• Data model made the appointment mechanic concrete: Distributor (brokerage) → Producer (agent) → Product (one carrier’s offering) → Connection (Producer+Product, requires the producer’s actual carrier credentials to create) → Application → quote/bind/issue. A Connection literally cannot be created without a pre-existing carrier appointment — confirms Herald grants zero appointments itself.
• Ownership/stability: post-Babbix pivot toward broker-facing AI products (Workbench, Inbox) makes it a rail that could converge into a direct competitor, not pure infrastructure — a strategic-drift risk distinct from the acquisition risk seen elsewhere in this tracker, but the same underlying caution: keep it swappable, per insurance-distribution-aggregator-layer.
• Reliability isn’t a pull-completion number here since there’s no pull — the closest analog is Connection activation (mostly immediate, a minority “referred” for ~24h review), which is a different kind of friction (carrier underwriting review, not technical failure).

Axle (axle.insure) — full detail: axle

• Same consumer-permissioned mechanism as Canopy/InsurGrid, but with a cleaner privacy boundary: Axle’s End User Terms state clients never receive login credentials, and Axle explicitly disclaims any relationship with carriers — a tighter three-party model than Canopy’s (which adds carriers as a fourth, data-buying party) or InsurGrid’s (whose policy disclaims control over what “Providers” do with shared data).
• Serves a different customer base than the other two entirely: lenders, dealers, rental companies, and employers verifying someone else’s coverage at a transaction moment — not agencies. Overlaps with Canopy specifically in the automotive/lending verification lane, not the agency-intake lane.
• Weakest transparency in the tracker on two fronts: no published carrier count (vaguer than either Canopy or InsurGrid) and fully unpublished pricing. Best developer-experience docs of the three on auth/webhook design, though a marketing claim of a “full-featured sandbox” wasn’t corroborated in the docs pulled.
• Reliability: essentially no independent signal exists — the only third-party review found under “Axle” on G2 appears to be a mismatched, unrelated product (context clues: compared against dental-software competitors). Treat as unverified in either direction, not as evidence of quality.
• Ownership: the most heavily and most recently capitalized of the personal-lines vendors (~$21.5M total, Aug 2026 Series A), independent, YC/Base10-backed. Its Series A explicitly funds a commercial-lines expansion — the most concrete, capitalized move yet toward the disconfirmer named in policy-ingestion-requires-a-data-rail. Revisit on a fixed cadence rather than treating personal-lines-only as a permanent constraint.

InsurGrid — vendor call addendum (2026-08-17, first live call)

This is the tracker’s first entry sourced from a conversation with the vendor rather than desk research, and it disagrees with the desk research on the single most important question.

• Contested: commercial credential-pull. Customer Success states ~300 personal / 120+ commercial carriers, with upload framed as a fallback added because most commercial agencies lack online accounts. The public FAQ reading above says extraction/upload only. Both readings are now recorded; neither is confirmed. Do not treat the “no commercial credential pull” line as settled — resolve against the carrier list and API docs the vendor committed to send. Same-lane consequence: the Canopy-vs-InsurGrid commercial comparison in the table above is currently resting on the weaker of the two sources.
• Pricing (question 8), now concrete: $99/moincluding5users,+$19.99/user beyond 5, no pull cap on standard; separate API agreement for enterprise, tiers from ~500 pulls/month, custom above ~100K/month. No free trial — minimum API contract, or 50% off month one.
• Data persistence (question 2), answered: pulled data is stored on the dashboard indefinitely until account deactivation, even though live carrier access expires with the session.
• Reliability (question 5), still unanswered: “about 100%” self-reported; the ~1% is glitches/partial pulls fixed manually on user report. No completion rate, no third-party signal. The internally inconsistent carrier counts were not raised on the call — worth doing next time.
• The refresh model is the real differentiator, and it cuts both ways. No auto-refresh; a login opens a 15–20 minute pull window; credentials are never stored. The vendor’s counter-story — an enterprise customer left for Canopy’s auto-refresh and came back on reliability and price — is from an interested party and needs independent checking. Either way, customer-initiated freshness is a hard constraint on any “continuous portfolio ownership” product, which is a design input for AX-INS-7, not just a vendor detail.
• Acquisition-phase usage is explicitly allowed — pre-client prospect pulls raise no T&C issue, with staging-period pricing available. This closes a real unknown for insurance-checkup-funnel-design.
• Under-weighted in the desk research: the custom form builder. Unlimited questions, producer branding, and — the part that matters — a single customer-facing step that combines custom questions with the carrier login flow, returning dec pages and structured answers together. That is closer to the check-up funnel’s actual intake shape than the raw pull API is.
• Setup: 2–3 working days from contract and payment to a staging account with API/webhook testing and developer support.

Fize (getfize.com) — full detail: fize

• The fifth member of the consumer-permissioned cohort, mechanically identical to Canopy/InsurGrid/Axle (credential link + document-upload fallback), 250+ fields, EZLynx/HawkSoft/Zapier integrations.
• Pricing is the most transparent of any vendor in this tracker — published tiers, published per-sync overage ($2),7-yeardocumenthistoryateverytier.SitseconomicallybetweenCanopy($4–6/commercial pull) and InsurGrid (flat unlimited).
• Its commercial tier is explicitly labelled “Beta” — the third independent signal that commercial credential pull is unsolved industry-wide, alongside Canopy’s G2-reported commercial flakiness and InsurGrid’s contested claim. Three vendors, three different ways of admitting the same weakness, supports policy-ingestion-requires-a-data-rail.
• Least knowable vendor here: no founding date, funding, HQ, team, carrier list, or reliability figure disclosed anywhere accessible.

QuoteSweep (quotesweep.com) — full detail: quotesweep

• Different category from everything else in this tracker except Herald — and a different mechanism from Herald. No carrier APIs at all: AI agents log into portals in parallel with the agency’s own credentials, handling MFA and CAPTCHAs.
• ⭐ The corroborating number. Founder’s premise: “2,700 carriers, and only 86 expose any commercial API” — 86 is exactly herald’s own published Insurance API Index count. So the entire API-first commercial quoting category addresses roughly 3% of US commercial carriers by count. That is the supply-side twin of H-INS-10: no API rail exists in either direction, only credential-driven brute force.
• It is the “portal-raking” job, productised — Job 4 in broker-jobs-to-be-done-map, automated, at $49–299/mo. Direct market evidence for that job’s cost and automatability.
• Strong documented security for a 2025 startup (credentials in AWS Secrets Manager, never passed into AI prompts, 30-day destruction on offboarding) — but the cited SOC 2 Type II belongs to its infra providers, not to QuoteSweep, unlike Canopy’s and Axle’s own certifications.
• Carrier counts internally inconsistent (70+ automated vs 550+ tracked) — same red-flag category as InsurGrid’s 150/350/450.

LexisNexis Risk Solutions — full detail: lexisnexis-risk-solutions

• A third mechanism, and the most thesis-relevant find of the sweep. Current Carrier Commercial (launched 2026-06-16) is a carrier-contributory policy database: participating carriers pool their books and get the aggregate back. No consumer permission, no portal login, no documents.
• It is the closest thing to a real US commercial policy rail that exists — and it is closed to us by construction. Commercial auto only (not the BOP/GL/WC/property spread an SMB actually holds); sold to carriers with access “limited to participating companies and authorized affiliates”; and contributory, so entry requires a book a new brokerage doesn’t have. It does not disconfirm policy-ingestion-requires-a-data-rail; it explains why the credential-pull cohort exists at all.
• Reframe worth carrying into the deck: consumer-permissioned scraping is not a technology choice, it is the workaround for not being a carrier. The data pool exists; membership is the gate.
• Revisit trigger: scope expansion beyond commercial auto, or any broker-accessible tier. Track on a cadence, like axle’s funded commercial expansion.

Evidence

• Web research sessions, 2026-08-15 through 2026-08-17 — vendor sites, public API docs, G2 reviews, funding press releases; full citations on each company page.
• insurance-tech-integration-stack — the architecture decision this tracker’s findings feed.
• insurance-checkup-funnel-design — the funnel whose intake-door ranking this tracker’s Canopy/InsurGrid findings directly informed.
• insurance-distribution-aggregator-layer — the acquisition-pattern hypothesis that motivated adding the Ownership & competitive stability question.
• 2026-08-17-directions-insurgrid-vendor-call — first live vendor call (InsurGrid Customer Success); source for the pricing, persistence, refresh-window, forms, and acquisition-phase findings, and for the commercial-credential-pull conflict.
• Adjacent-vendor sweep, 2026-08-17 — web research across G2/CB Insights alternative lists, vendor sites and pricing pages, the Nami VP infrastructure landscape, and QuoteSweep’s (vendor-authored) platform comparisons; added fize, quotesweep, lexisnexis-risk-solutions and surfaced the unprofiled remainder listed below.

Open questions

• Canopy’s carrier-side data-sharing mechanism (segment 3) is undocumented publicly — close via a sales call before committing consent-language copy.
• Herald’s actual live carrier-connection count (vs. its 86/228 market index) is unpublished — matters for realistic P1 quoting coverage estimates.
• Axle’s real carrier count and pricing remain undisclosed anywhere public — close in a direct vendor conversation if it moves from “track” to “candidate.”
• No vendor here has a third-party-verified completion/reliability number except Canopy’s G2-corroborated commercial weakness — worth deliberately seeking out reviews/references for InsurGrid, Herald, and Axle rather than relying on vendor self-report, next time capacity allows.
• Next vendors worth adding: coverforce (assigned as a research follow-up in 2026-08-16-directions-broker-jobs-to-be-done-map), momentum-amp (AMS, different category — may need a variant framework).
• Surfaced but not yet profiled (2026-08-17 sweep): ingestion — Trellis (embedded/auto comparison), Insurely (European open-finance; relevant only if the UK second market under AX-INS-9 is ever activated), Verisk (the other incumbent data giant). Quoting — Tarmika (applied-systems-owned, ~31 carriers), Semsee (iBynd-owned, 48), vertafore Commercial Submissions (~27), Pathpoint (E&S digital wholesale), 1Fort ($7.5M, ~12–15 carriers, cyber-leaning), Appulate, Highwing, Agentero.
• Legal durability of the whole credential-automation mechanism is unexamined. Both cohorts — ingestion (Canopy/InsurGrid/Axle/Fize) and portal quoting (QuoteSweep) — depend on automated logins to carrier portals using someone else’s credentials. Nobody has been asked what carrier terms of service say about that, or what happens if a large carrier decides to enforce. This is a single shared point of failure across most of this tracker and deserves a real answer before either rail becomes load-bearing.
• New, and the highest-priority item on this page: InsurGrid commercial support — credential-pull (vendor CS) or upload-only (public FAQ)? Resolve against the promised carrier list and API docs before the Canopy-vs-InsurGrid decision.
• Canopy’s auto-refresh reliability is now disputed by a direct competitor. Neither account is independent; verify before treating either as fact.
• The 15–20 minute pull window means portfolio freshness is customer-initiated for any vendor on this mechanism. Worth asking Canopy the same question directly — if its auto-refresh has the same underlying constraint, the differentiator dissolves.
• Method note: one call answered more standing questions than three days of desk research. Prefer vendor calls over site-reading for the remaining entries (coverforce, momentum-amp).

Related

• insurance-tech-integration-stack · insurance-checkup-funnel-design · insurance-distribution-aggregator-layer
• canopy-connect · insurgrid · herald · axle · coverforce