METHODOLOGY SOURCE OF OUR NUMBERS

How we source
our numbers.

Every headline statistic on this site is documented here — source, sample size, caveats, and the date of the last underlying data refresh.

Statistic 1

62% of after-hours emergency calls go unanswered

Source: Proof Pulse intake telemetry — calls received by Proof customers between 18:00 and 06:00 local time, Q4 2025 through Q1 2026.
Sample size: n = 1,847 inbound emergency calls across 31 enrolled customers.
Caveats: Sample is biased toward small-to-mid restoration operators (1-10 trucks) who chose Pulse specifically because they had a known answering problem; pure-enterprise contractors with 24/7 staffed call centers were excluded. The 62% figure represents the share of calls that reached voicemail or were terminated before a human responded — it does not separate 'never returned' from 'returned >30 minutes later.'

Statistic 2

$4,200 average revenue lost per missed call

Source: Internal Proof OS billing-ledger analysis of water-damage emergency tickets, October 2025 - March 2026, where the work order was successfully dispatched after the original intake call had hit voicemail.
Sample size: n = 412 jobs. Distribution: median $3,860, mean $4,212, standard deviation $2,140. Rounded to $4,200 for marketing display.
Caveats: This measures revenue on jobs that we eventually won despite the missed call — it understates the loss for contractors who don't have a recovery system in place. The figure includes mitigation only (not reconstruction phase) and is U.S. restoration market specific. Carrier deductibles and customer OOP are included in the gross revenue figure.

Statistic 3

$48K/year captured by Launchpad / replaced by Engine

Source: Pulse missed-call data * Launchpad capture rate * average revenue per missed call. Calculation: 1.7 missed emergency calls per week * 52 weeks * 67% capture lift on Launchpad * $4,212 average ticket = $249K of captured revenue.
Sample size: We round the gross capture to a conservative $48K/year that is net of crew cost, COGS, and Launchpad subscription, because contractors will read the larger number with skepticism and a smaller defensible number is more honest.
Caveats: Assumes a contractor that today is missing roughly 7 emergency calls per month after hours. Operators with much lower missed-call rates won't see this lift; operators with much higher rates will see substantially more. The 'replaced by Engine' phrasing means that once a contractor scales past Launchpad capacity, the same dollar capture flows to Engine and grows from there.

Statistic 4

Claims paid 40% faster

Source: Time from first emergency dispatch to first carrier payment received, comparing the same contractor's pre-Proof submissions to their Proof-generated submissions.
Sample size: n = 2,180 closed claims across 47 contractors who migrated to Proof and had at least 25 closed claims pre- and post-migration. Pre-Proof median time-to-first-payment: 31 days. Post-Proof: 18.7 days. Median improvement: 39.7%, rounded to 40%.
Caveats: Selection bias: these are contractors who chose to adopt Proof and stayed on the platform — they were already motivated to fix their claims process. Carrier mix, TPA mix, and storm-season effects all contribute to variance. The 40% figure should be read as a typical-customer experience, not an SLA.

Statistic 5

97% Call Capture Rate

Source: Proof Pulse intake telemetry, January 2026 through March 2026. We define captured as 'answered by the AI, qualified, and converted to a ticket in the system without requiring a human dispatcher to intervene before the customer received an ETA.'
Sample size: n = 9,418 inbound emergency calls across all Pulse tenants. 9,158 captured / 9,418 total = 97.24%. Rounded to 97%.
Caveats: This is a capture rate, not a conversion rate. Captured tickets still require crew dispatch, scope acceptance, and claim execution to become revenue. The 3% of uncaptured calls fall into three buckets: spam/robocall (1.4%), non-restoration inquiry (0.9%), and edge-case AI handoff failures (0.7%).

Statistic 6

3 min average response

Source: Time from inbound emergency call termination to outbound dispatch SMS sent to the on-call technician, measured across all Pulse + Launchpad + Engine tenants.
Sample size: n = 6,341 dispatched emergencies, January 2026 through March 2026. Median: 2 minutes 47 seconds. Mean: 3 minutes 14 seconds. Rounded to 3 minutes.
Caveats: This measures system response (call-to-dispatch), not crew arrival on-site, which depends on contractor geography and traffic. Outliers above 10 minutes (1.2% of dispatches) typically involve commercial-routing scenarios that require explicit human approval before crew assignment.

Testimonial composite-disclosure policy

Some testimonials and quotes on this site are composite testimonials: they reflect feedback from real Proof customers in the same business segment (restoration operators of similar size and geography), but the specific customer-identifying details have been withheld at the customer's request, and quotes have been edited for clarity and length.

We follow the FTC's Endorsements Guides at 16 CFR § 255.2(b): composite testimonials are disclosed as such, and we do not present a composite as the voice of a single named individual. Where a quote is from a single named source with permission to attribute, we do attribute it; everywhere else, the quote is labeled as a composite.

If you are a reporter, analyst, or prospective customer and would like to verify a specific claim — including being introduced to a referenceable customer under NDA — contact trust@proofco.ai.

Last reviewed: May 13, 2026.

How we sourceour numbers.