Veritus growth playbook

Make the demo call the first sales step. Add work email, company and company type to the try page, using the same options as the contact form. At the end of each demo call, the agent offers, in its own voice, to book a 20-minute call with the team. It then sends a text with the booking link and the matching case study: AWA for third-party collections, Splash for post-offer loan sales, the AWA results for early-stage collections. Push every try-page lead into HubSpot with the agent chosen and the call length, and alert the AE when the company type is lender, issuer or agency. Add a "Talk to our team" link on the try page for visitors who would rather not take a call.

A prospect who has just heard the agent negotiate a payment plan is at peak belief, and asking for the meeting on that call shows the product doing its job. It also turns an anonymous phone number into a qualified lead at a company where "the constraint on growth is selling capacity".

[ ] Work email, company and company type fields (same options as /contact)
[ ] "Talk to our team" link on the try page
[ ] Closing turn in Alex, Jason and Maya offering a 20-minute call
[ ] Consent wording reviewed for the follow-up text
[ ] Follow-up text: booking link + matching case study
[ ] HubSpot record with agent chosen and call length
[ ] AE alert for lender, issuer and agency leads
[ ] Weekly booking-rate report by agent

• Extra fields lower try-page submissions

  Ask for email and company after the agent is chosen, and test with and without.

• The booking ask feels pushy

  Offer it once, after the scenario ends, and accept a no gracefully, which is itself a demo of good agent behaviour.

• Your websiteTry page (checked November 9, 2026): "Enter your details to receive a call from one of our AI agents"; fields are first name, last name and phone number, then a choice of three agents: Alex "Post-offer loan sales", Jason "Early-stage collections" and Maya "Third-party collections". No email, company or role field, no servicing or welcome-call agent, and no link to book a demo or reach sales; the only link is the privacy policy.Source
• Your websiteHomepage: H1 "The AI Communications Platform for Financial Services"; six use cases (customer acquisition, welcome calls, loan servicing, customer experience, early-stage and late-stage collections) for fintech lenders, card issuers, financial institutions and third-party agencies; CTAs "Speak to an AI agent" (to app.veritus.com/try), "Get in touch with a human" and a "Launch app" button for existing users; stats "71%" containment and "<5%" of borrowers asking whether they are speaking with AI; testimonials from AWA, Splash Financial, Second Alliance ("Veritus acts as my AI subservicer") and a former LendingClub consumer banking executive; "Backed by Y Combinator."Source
• Your websiteContact page: "See why financial institutions choose Veritus" and "Book a demo..."; a form asking name, work email, company, optional phone, "Number of call center agents" and company type (financial institution, fintech lender, card issuer, 3rd party agency, other). No calendar booking and no description of what happens after submitting.Source
• Your websiteWisetack case study: an AI voice agent (Morgan) called dormant merchants and followed up stalled applications, tested against a randomized control group: 20% higher reactivation, 13% more loan volume, reactivation up to 2.5x control for merchants inactive 5-10 months, 25% more converted loan volume and 41% more converted loans per merchant on stalled applications, "0 additional hires"; an 8-page PDF, ungated.Source
• HiringFounding Account Executive (San Francisco, posted July 29, 2026): "this is our first full-cycle sales hire" after "founder-led sales"; targets "Collections agencies and midsize lenders first", then enterprise lenders; works conferences RMAI, ACA, Money20/20 and LendIt; "the constraint on growth is selling capacity"; tools include HubSpot, Apollo and LinkedIn Sales Navigator. Calls consumer lending and collections "among the last corners of financial services that AI has not reached".Source

Add two scenarios to the try page: a servicing agent (due-date change, payoff question, payment with the secure keypad) based on the Finovate demo, and a welcome-call or card-activation agent for issuers and banks. Label each scenario by buyer (lender, issuer, agency) so visitors pick their own case.

Prospects believe what they hear in their own situation, and the industry pages for card issuers and banks finally get a demo that matches them.

• Too many choices

  Group scenarios by buyer type.

• #1 Turn "Speak to an AI agent" into a booked demo: the call form takes only a phone number and offers no next step

• Your websiteTry page (checked November 9, 2026): "Enter your details to receive a call from one of our AI agents"; fields are first name, last name and phone number, then a choice of three agents: Alex "Post-offer loan sales", Jason "Early-stage collections" and Maya "Third-party collections". No email, company or role field, no servicing or welcome-call agent, and no link to book a demo or reach sales; the only link is the privacy policy.Source
• Your websiteHomepage: H1 "The AI Communications Platform for Financial Services"; six use cases (customer acquisition, welcome calls, loan servicing, customer experience, early-stage and late-stage collections) for fintech lenders, card issuers, financial institutions and third-party agencies; CTAs "Speak to an AI agent" (to app.veritus.com/try), "Get in touch with a human" and a "Launch app" button for existing users; stats "71%" containment and "<5%" of borrowers asking whether they are speaking with AI; testimonials from AWA, Splash Financial, Second Alliance ("Veritus acts as my AI subservicer") and a former LendingClub consumer banking executive; "Backed by Y Combinator."Source
• Your websiteFinovate Fall demo page (October 1, 2026): a live inbound servicing call handled by agent Ivy, with observer agents and a secure keypad for card payments; stats "as of September 2026": "80%+" containment on inbound servicing calls and "95%+" on outbound calls; a full transcript; visitors could try a voice agent at Veritus's phone booth at the event.Source

Show a booking calendar right after the form is submitted. Agencies and midsize lenders go to the AE's calendar; large call centers get a longer slot with a forward-deployed engineer added. Tell visitors what happens next ("a 30-minute call; you'll hear an agent on your own use case") and send a confirmation that includes the try-page link.

Instant scheduling removes the email back-and-forth that loses B2B leads, and routing by size sends a one-person sales team to the biggest deals first.

• Unqualified bookings fill the calendar

  Show the calendar only for the listed company types; others get a reply by email.

• Your websiteContact page: "See why financial institutions choose Veritus" and "Book a demo..."; a form asking name, work email, company, optional phone, "Number of call center agents" and company type (financial institution, fintech lender, card issuer, 3rd party agency, other). No calendar booking and no description of what happens after submitting.Source
• HiringFounding Account Executive (San Francisco, posted July 29, 2026): "this is our first full-cycle sales hire" after "founder-led sales"; targets "Collections agencies and midsize lenders first", then enterprise lenders; works conferences RMAI, ACA, Money20/20 and LendIt; "the constraint on growth is selling capacity"; tools include HubSpot, Apollo and LinkedIn Sales Navigator. Calls consumer lending and collections "among the last corners of financial services that AI has not reached".Source

Publish one short metrics page: each figure with its definition (containment, resolution, payment rate), customer or program, time period and date. Every page that shows a number links to it. Replace "95%+ resolved" with the sourced outbound figure, or remove it until it has a source.

Precise, sourced claims build trust with risk-averse buyers, and they protect the company from marketing claims it cannot back up.

• Lower headline numbers

  Precision wins with compliance buyers; show ranges by program type.

• Your websiteIndustry pages: fintech lenders, financial institutions and card issuers each claim "95%+ resolved. End to end." with no source or period, next to sourced figures of "70%+" (AWA) and "74%" (Splash). The card-issuer and banks & credit unions pages show no customer story from their own segment. The agencies page promises "30-day" deployment; integrations are described only as "REST, webhooks, and native connectors across your existing stack", with no named systems.Source
• Your websiteFinovate Fall demo page (October 1, 2026): a live inbound servicing call handled by agent Ivy, with observer agents and a secure keypad for card payments; stats "as of September 2026": "80%+" containment on inbound servicing calls and "95%+" on outbound calls; a full transcript; visitors could try a voice agent at Veritus's phone booth at the event.Source
• Your websiteHomepage: H1 "The AI Communications Platform for Financial Services"; six use cases (customer acquisition, welcome calls, loan servicing, customer experience, early-stage and late-stage collections) for fintech lenders, card issuers, financial institutions and third-party agencies; CTAs "Speak to an AI agent" (to app.veritus.com/try), "Get in touch with a human" and a "Launch app" button for existing users; stats "71%" containment and "<5%" of borrowers asking whether they are speaking with AI; testimonials from AWA, Splash Financial, Second Alliance ("Veritus acts as my AI subservicer") and a former LendingClub consumer banking executive; "Backed by Y Combinator."Source
• Your websiteSplash Financial case study: outbound voice agent (Sofia), live since March 2026, calling applicants stalled at income documentation or identity verification: "20%+" relative conversion lift over the human baseline (April-June 2026, period-over-period), "74%" containment against a 70% target, "0" complaints in Q2; quote: "built for lending, not retrofitted for it."Source

Add "published" and "last reviewed" dates to every insight, plus a reviewer line on compliance pieces (for example, "reviewed with TCPA counsel" if that is true). Sort the hub by date and set a quarterly review for anything regulatory.

Dated, reviewed compliance content is more credible to buyers and to search and AI answer engines, which favour fresh sources.

• Old dates make content look stale

  Show the last-reviewed date prominently.

• Your websiteInsights hub: 9 items with no publication dates shown: 3 case studies (Wisetack, Splash Financial, AWA), 3 white papers (including "TCPA consent for AI voice agents" and "Turn application drop-off into revenue", whose CTA is "Download the playbook"), 1 blog post, 1 video and 1 compliance brief. No newsletter or subscribe option on the hub or the blog post.Source
• Your website"Q1 2026 Buyer's Guide: Virtual Agents in U.S. Consumer Collections", written by an outside collections-technology advisory firm, gated behind a form, with an Excel RFP requirements template. The TCPA compliance brief is ungated, carries no date, and ends with an optional "Building an AI voice program?" form.Source

Remove /design-library/ and /brand/ from the sitemap, mark them noindex (or move them behind the app login), and add the missing commercial pages to the sitemap as they ship (metrics, integrations, pricing approach).

A clean sitemap focuses crawling on the pages that sell, and it is a one-line change in the Astro sitemap config.

• Designers or partners rely on the public library

  Keep it reachable by URL, just not indexed.

• Your websiteSitemap (75 URLs, checked November 9, 2026): 49 are /design-library/ component pages (buttons, accordions, typography, brand tokens) plus /brand/; the rest are the homepage, about, careers, contact, 4 platform pages, 4 industry pages, 9 insights, privacy and terms. No pricing, integrations, comparison or customer-stories hub pages.Source

Make the header's main button "Speak to an AI agent" or "Book a demo", and move login to a plain "Log in" text link. Track header clicks before and after.

The always-visible button should serve the visitor most likely to be there; existing users know where to log in.

• Customers can't find login

  Keep "Log in" visible in the header.

• Your websiteHomepage: H1 "The AI Communications Platform for Financial Services"; six use cases (customer acquisition, welcome calls, loan servicing, customer experience, early-stage and late-stage collections) for fintech lenders, card issuers, financial institutions and third-party agencies; CTAs "Speak to an AI agent" (to app.veritus.com/try), "Get in touch with a human" and a "Launch app" button for existing users; stats "71%" containment and "<5%" of borrowers asking whether they are speaking with AI; testimonials from AWA, Splash Financial, Second Alliance ("Veritus acts as my AI subservicer") and a former LendingClub consumer banking executive; "Backed by Y Combinator."Source
• PressSeed round: $10.1 million announced February 5, 2026, led by Crosslink and Threshold, with Emergence Capital and others; Veritus went through Y Combinator's Summer 2025 batch; deployments include application-funnel outreach to pre-qualified borrowers and early-stage delinquency calls that take payments.Source

Package an "After-hours inbound" starter for collection agencies: the inbound agent answers evenings and weekends, takes card payments, sets up plans within the client's rules, and reports dollars collected that would otherwise have gone to voicemail. Fixed scope, a 30-day launch (as the agencies page promises), and a clear path to daytime overflow and 24/7.

Agency owners buy recovered dollars, and after-hours calls are revenue nobody is collecting today, so the first ROI report is easy to defend.

• Low after-hours call volume at small agencies

  Qualify on current after-hours voicemail volume first.

• Your websiteAWA case study (inbound collections agency): "25%" of dollars the agent (Linda) collects come from after-hours calls that previously went to voicemail; "67%" of customers pay on the call; "71%" containment; one in five fully handled calls ends in a payment; the agent started after hours, then went 24/7; the earlier vendor's SMS payment links were flagged as spam by carriers. PDF ungated.Source
• HiringFounding Account Executive (San Francisco, posted July 29, 2026): "this is our first full-cycle sales hire" after "founder-led sales"; targets "Collections agencies and midsize lenders first", then enterprise lenders; works conferences RMAI, ACA, Money20/20 and LendIt; "the constraint on growth is selling capacity"; tools include HubSpot, Apollo and LinkedIn Sales Navigator. Calls consumer lending and collections "among the last corners of financial services that AI has not reached".Source
• Your websiteIndustry pages: fintech lenders, financial institutions and card issuers each claim "95%+ resolved. End to end." with no source or period, next to sourced figures of "70%+" (AWA) and "74%" (Splash). The card-issuer and banks & credit unions pages show no customer story from their own segment. The agencies page promises "30-day" deployment; integrations are described only as "REST, webhooks, and native connectors across your existing stack", with no named systems.Source

Package "QA for your human team" as a paid 60-day entry product: upload or stream recordings, get every call scored on Mini Miranda, recording disclosure, right-party identification and cease-and-desist, and receive a report of compliance gaps and the call types most suited to automation. The report ends with a proposal for the first agent.

It sells to the compliance budget first, proves accuracy on the buyer's own calls, and turns QA findings into the business case for agents.

• Becoming a QA vendor instead of an agent platform

  Time-limit the package and build the agent proposal into it.

• Your websiteAutomated QA page: "Every call scored." AI scores 100% of calls, "human and AI", against custom compliance scorecards (Mini Miranda, recording disclosure, right-party identification, cease and desist); manual QA samples "1–3%" of calls; human agents can dispute flags. The page does not say whether QA can be bought on its own.Source
• HiringVoice AI QA Analyst (remote, posted October 4, 2026): "Every agent we deliver is tested against real conversations before it reaches a client"; grades live calls on accuracy, compliance and tone, analyses trends and prepares "quarterly client performance reports" with Customer Success; FDCPA, TCPA, ECOA and SCRA knowledge.Source

Make a standard pilot template: the success metric and control design agreed at kickoff (holdout group or matched cohort), weekly results from the QA analyst, and a day-60 readout that converts to annual automatically when the targets are hit. Publish an anonymized version as "How we run pilots" for prospects.

Pre-agreed success criteria shorten renewal decisions, and every pilot becomes a case study with numbers a risk team will accept.

• Holdouts reduce the client's short-term results

  Keep holdouts small (10-20%) and time-boxed.

• HiringCustomer Success Manager (posted September 29, 2026): runs mid-market "Growth" accounts (collection agencies, lenders, fintechs) from kickoff to renewal, owns "pilot-to-annual conversion" and expansion with Forward Deployed Engineers; success means "agents live, customers renewed, and results delivered". Veritus's agents "already handle thousands of conversations every day".Source
• Your websiteWisetack case study: an AI voice agent (Morgan) called dormant merchants and followed up stalled applications, tested against a randomized control group: 20% higher reactivation, 13% more loan volume, reactivation up to 2.5x control for merchants inactive 5-10 months, 25% more converted loan volume and 41% more converted loans per merchant on stalled applications, "0 additional hires"; an 8-page PDF, ungated.Source
• Your websiteSplash Financial case study: outbound voice agent (Sofia), live since March 2026, calling applicants stalled at income documentation or identity verification: "20%+" relative conversion lift over the human baseline (April-June 2026, period-over-period), "74%" containment against a 70% target, "0" complaints in Q2; quote: "built for lending, not retrofitted for it."Source

List every system Veritus has connected in production or tested (loan management and collection systems of record, dialers and CCaaS, payment processors, CRMs), each with its status (live, tested, on request) and what syncs. Add a "don't see yours?" form and link the page from the 30-day deployment promise.

Named integrations remove a top objection, rank for "<system> AI voice agent" searches, and give the AE an answer during the first call.

• Overstating depth of an integration

  Show status honestly: live, tested, on request.

• Your websiteIndustry pages: fintech lenders, financial institutions and card issuers each claim "95%+ resolved. End to end." with no source or period, next to sourced figures of "70%+" (AWA) and "74%" (Splash). The card-issuer and banks & credit unions pages show no customer story from their own segment. The agencies page promises "30-day" deployment; integrations are described only as "REST, webhooks, and native connectors across your existing stack", with no named systems.Source
• Your websiteVoice agents page: agents in "any language your customers speak" (no language is named and there is no Spanish example); fully managed telephony or "bring your own stack" over SIP to an existing dialer or CCaaS, with no vendors named; no pricing, pilot or launch timeline.Source
• HiringMember of Technical Staff (posted July 22, 2026): full-stack ownership in TypeScript, React and Next.js; builds "APIs and integrations with CRMs, payment platforms, and telephony" and helps design A/B test experiments.Source
• Competitor researchRivals' own homepages: Salient, "AI built for consumer lending", "$75M raised", logos such as Westlake Financial and Ally Financial and public release notes; Skit.ai names its integrations (Twilio, Genesys, TCN, PayNearMe, Finvi) and invites prospects to benchmark their portfolio; Prodigal names Five9, Genesys and NICE for onboarding "within weeks" and markets Spanish-language notes.See the research Source

Until a named issuer or bank story exists, publish the closest honest proof: anonymized results from any issuer or bank pilot, a servicing demo with full transcript (the Finovate call), and the examiner-ready audit trail shown in a sample report. Make the first issuer or bank customer's case study a contract term for a pilot discount.

Segment-specific proof lifts conversion on segment pages, and a negotiated case-study right speeds up the first reference.

• Banks refuse to be named

  Use descriptive anonymization ("top-50 card issuer") with approval.

• #4 Give every performance number one source and one date, starting with the unsourced "95%+ resolved"

• Your websiteIndustry pages: fintech lenders, financial institutions and card issuers each claim "95%+ resolved. End to end." with no source or period, next to sourced figures of "70%+" (AWA) and "74%" (Splash). The card-issuer and banks & credit unions pages show no customer story from their own segment. The agencies page promises "30-day" deployment; integrations are described only as "REST, webhooks, and native connectors across your existing stack", with no named systems.Source
• Your websiteFinovate Fall demo page (October 1, 2026): a live inbound servicing call handled by agent Ivy, with observer agents and a secure keypad for card payments; stats "as of September 2026": "80%+" containment on inbound servicing calls and "95%+" on outbound calls; a full transcript; visitors could try a voice agent at Veritus's phone booth at the event.Source

Write the Second Alliance case study around placements: how many more accounts the firm took on, settlement and payment-plan rates, staff time freed, and compliance results from QA. Publish it as a page and an ungated PDF like AWA's, and cut a one-minute audio clip of a redacted settlement call.

Agency owners buy on capacity ("more placements without headcount"), and a second agency story makes the agency page twice as persuasive.

• Client declines detailed numbers

  Use ranges and percentages.

• Your websiteHomepage: H1 "The AI Communications Platform for Financial Services"; six use cases (customer acquisition, welcome calls, loan servicing, customer experience, early-stage and late-stage collections) for fintech lenders, card issuers, financial institutions and third-party agencies; CTAs "Speak to an AI agent" (to app.veritus.com/try), "Get in touch with a human" and a "Launch app" button for existing users; stats "71%" containment and "<5%" of borrowers asking whether they are speaking with AI; testimonials from AWA, Splash Financial, Second Alliance ("Veritus acts as my AI subservicer") and a former LendingClub consumer banking executive; "Backed by Y Combinator."Source
• Your websiteInsights hub: 9 items with no publication dates shown: 3 case studies (Wisetack, Splash Financial, AWA), 3 white papers (including "TCPA consent for AI voice agents" and "Turn application drop-off into revenue", whose CTA is "Download the playbook"), 1 blog post, 1 video and 1 compliance brief. No newsletter or subscribe option on the hub or the blog post.Source
• HiringFounding Account Executive (San Francisco, posted July 29, 2026): "this is our first full-cycle sales hire" after "founder-led sales"; targets "Collections agencies and midsize lenders first", then enterprise lenders; works conferences RMAI, ACA, Money20/20 and LendIt; "the constraint on growth is selling capacity"; tools include HubSpot, Apollo and LinkedIn Sales Navigator. Calls consumer lending and collections "among the last corners of financial services that AI has not reached".Source

Run the phone booth at the collections shows (RMAI, ACA) with agency scenarios: inbound payment, settlement within limits, a hardship conversation. Badge-scan every caller, and have the agent call back the next day on the attendee's own phone to book a meeting. Pre-book meetings with target agencies from Apollo lists before the show.

Experiencing the agent is the best pitch Veritus has, and a structured follow-up turns booth traffic into pipeline for a one-AE team.

• Callback consent

  Collect explicit consent at the booth for the follow-up call.

• #1 Turn "Speak to an AI agent" into a booked demo: the call form takes only a phone number and offers no next step

• Your websiteFinovate Fall demo page (October 1, 2026): a live inbound servicing call handled by agent Ivy, with observer agents and a secure keypad for card payments; stats "as of September 2026": "80%+" containment on inbound servicing calls and "95%+" on outbound calls; a full transcript; visitors could try a voice agent at Veritus's phone booth at the event.Source
• HiringFounding Account Executive (San Francisco, posted July 29, 2026): "this is our first full-cycle sales hire" after "founder-led sales"; targets "Collections agencies and midsize lenders first", then enterprise lenders; works conferences RMAI, ACA, Money20/20 and LendIt; "the constraint on growth is selling capacity"; tools include HubSpot, Apollo and LinkedIn Sales Navigator. Calls consumer lending and collections "among the last corners of financial services that AI has not reached".Source
• Your websiteTry page (checked November 9, 2026): "Enter your details to receive a call from one of our AI agents"; fields are first name, last name and phone number, then a choice of three agents: Alex "Post-offer loan sales", Jason "Early-stage collections" and Maya "Third-party collections". No email, company or role field, no servicing or welcome-call agent, and no link to book a demo or reach sales; the only link is the privacy policy.Source

Publish one short data note a month from anonymized platform data (best call windows, payment-on-call rates, after-hours share, opt-out rates by channel), each with one chart and one practical rule. Add an email subscribe box to every insight, and send the note through HubSpot.

A regular stream of original data builds an owned audience of collections and servicing leaders, gives the AE reasons to reach out, and earns links and citations.

• Customer data sensitivity

  Aggregate across portfolios and state the sample.

• Your websiteBlog "What Happens to RPC Rates with Repeated Dials" (September 29, 2026, Head of Data Science): right-party-contact rate falls from 1.00 on the first attempt to 0.59, 0.42 and 0.26 on attempts two to four and 0.12 on the tenth; "The first three attempts produced 70% of all the right-party contacts"; the only blog post on the site, ending in "Get in touch".Source
• Your websiteInsights hub: 9 items with no publication dates shown: 3 case studies (Wisetack, Splash Financial, AWA), 3 white papers (including "TCPA consent for AI voice agents" and "Turn application drop-off into revenue", whose CTA is "Download the playbook"), 1 blog post, 1 video and 1 compliance brief. No newsletter or subscribe option on the hub or the blog post.Source

Within a day of each download, send Veritus's own answers to the RFP template (pre-filled, with links to the trust center, QA page and case studies) and offer a 30-minute RFP walkthrough. Co-host a webinar with the guide's author on running an AI-agent pilot, and publish a Q1 2027 refresh so the guide stays current.

Answering the buyer's template before they send it frames the evaluation around Veritus's strengths, and the independent author adds credibility.

• Looks like the guide is not neutral

  Keep the guide unchanged and send Veritus answers separately.

• Your website"Q1 2026 Buyer's Guide: Virtual Agents in U.S. Consumer Collections", written by an outside collections-technology advisory firm, gated behind a form, with an Excel RFP requirements template. The TCPA compliance brief is ungated, carries no date, and ends with an optional "Building an AI voice program?" form.Source
• HiringFounding Account Executive (San Francisco, posted July 29, 2026): "this is our first full-cycle sales hire" after "founder-led sales"; targets "Collections agencies and midsize lenders first", then enterprise lenders; works conferences RMAI, ACA, Money20/20 and LendIt; "the constraint on growth is selling capacity"; tools include HubSpot, Apollo and LinkedIn Sales Navigator. Calls consumer lending and collections "among the last corners of financial services that AI has not reached".Source

Add a Spanish option to the try page (start with the early-stage collections agent), with Spanish disclosures and a Spanish follow-up text. Publish a short page on multi-language agents: languages in production, how disclosures are validated per language, and language-switching mid-call.

Turning a vague claim into something a buyer can hear removes doubt, and it answers a specific staffing problem for agencies and lenders.

• Lower accuracy in Spanish

  Launch only after QA pass rates match English.

• #2 Add the servicing and welcome-call agents the homepage sells to the try page

• Your websiteVoice agents page: agents in "any language your customers speak" (no language is named and there is no Spanish example); fully managed telephony or "bring your own stack" over SIP to an existing dialer or CCaaS, with no vendors named; no pricing, pilot or launch timeline.Source
• Your websiteTry page (checked November 9, 2026): "Enter your details to receive a call from one of our AI agents"; fields are first name, last name and phone number, then a choice of three agents: Alex "Post-offer loan sales", Jason "Early-stage collections" and Maya "Third-party collections". No email, company or role field, no servicing or welcome-call agent, and no link to book a demo or reach sales; the only link is the privacy policy.Source
• Competitor researchRivals' own homepages: Salient, "AI built for consumer lending", "$75M raised", logos such as Westlake Financial and Ally Financial and public release notes; Skit.ai names its integrations (Twilio, Genesys, TCN, PayNearMe, Finvi) and invites prospects to benchmark their portfolio; Prodigal names Five9, Genesys and NICE for onboarding "within weeks" and markets Spanish-language notes.See the research Source

Build a simple calculator: agents, calls per month, after-hours share and average payment give estimated contained calls, staff hours freed and after-hours dollars recovered, with every assumption linked to its case study. Offer the result as a PDF that requires a work email, and pass it to the AE.

Buyers need a number for their own business case, and the calculator turns anonymous visitors into qualified leads with their size already known.

• Overpromising results

  Show ranges and cite each assumption.

• #4 Give every performance number one source and one date, starting with the unsourced "95%+ resolved"

• Your websiteContact page: "See why financial institutions choose Veritus" and "Book a demo..."; a form asking name, work email, company, optional phone, "Number of call center agents" and company type (financial institution, fintech lender, card issuer, 3rd party agency, other). No calendar booking and no description of what happens after submitting.Source
• Your websiteAWA case study (inbound collections agency): "25%" of dollars the agent (Linda) collects come from after-hours calls that previously went to voicemail; "67%" of customers pay on the call; "71%" containment; one in five fully handled calls ends in a payment; the agent started after hours, then went 24/7; the earlier vendor's SMS payment links were flagged as spam by carriers. PDF ungated.Source
• Your websiteSitemap (75 URLs, checked November 9, 2026): 49 are /design-library/ component pages (buttons, accordions, typography, brand tokens) plus /brand/; the rest are the homepage, about, careers, contact, 4 platform pages, 4 industry pages, 9 insights, privacy and terms. No pricing, integrations, comparison or customer-stories hub pages.Source

Each quarter, publish benchmark ranges for containment, payment-on-call, right-party contact by attempt, disclosure pass rates and complaint rates, split by use case (application funnel, servicing, early and late collections). Customers get their own position against the benchmark in their quarterly report; prospects get the public version and an offer to benchmark their own operation.

A recurring benchmark becomes the reference buyers, advisors and press cite, and placing each customer against it supports renewals.

• Small samples in some segments

  Publish only segments above the minimum sample.

• #4 Give every performance number one source and one date, starting with the unsourced "95%+ resolved"
• #15 Turn the RPC research into a monthly data note with a subscribe box

• HiringVoice AI QA Analyst (remote, posted October 4, 2026): "Every agent we deliver is tested against real conversations before it reaches a client"; grades live calls on accuracy, compliance and tone, analyses trends and prepares "quarterly client performance reports" with Customer Success; FDCPA, TCPA, ECOA and SCRA knowledge.Source
• Your websiteAutomated QA page: "Every call scored." AI scores 100% of calls, "human and AI", against custom compliance scorecards (Mini Miranda, recording disclosure, right-party identification, cease and desist); manual QA samples "1–3%" of calls; human agents can dispute flags. The page does not say whether QA can be bought on its own.Source
• HiringCustomer Success Manager (posted September 29, 2026): runs mid-market "Growth" accounts (collection agencies, lenders, fintechs) from kickoff to renewal, owns "pilot-to-annual conversion" and expansion with Forward Deployed Engineers; success means "agents live, customers renewed, and results delivered". Veritus's agents "already handle thousands of conversations every day".Source
• Competitor researchRivals' own homepages: Salient, "AI built for consumer lending", "$75M raised", logos such as Westlake Financial and Ally Financial and public release notes; Skit.ai names its integrations (Twilio, Genesys, TCN, PayNearMe, Finvi) and invites prospects to benchmark their portfolio; Prodigal names Five9, Genesys and NICE for onboarding "within weeks" and markets Spanish-language notes.See the research Source

Launch a "merchant and dealer activation" solution: an agent that calls inactive merchants, re-trains them on the financing flow and follows up stalled applications with them, sold to lenders that originate through partners. Add an industry page for merchant-network lenders built on the Wisetack data, and target a short list of POS and dealer-financing lenders.

It opens a buyer (partner and sales teams at lenders) with a revenue budget and lower regulatory exposure than consumer collections, backed by randomized proof.

• Distracts a small team from core segments

  Reuse the Wisetack flows and cap the effort at four pilots.

• #10 Run every pilot like Wisetack's: a control group, a scorecard agreed on day one, a renewal decision on day 60

• Your websiteWisetack case study: an AI voice agent (Morgan) called dormant merchants and followed up stalled applications, tested against a randomized control group: 20% higher reactivation, 13% more loan volume, reactivation up to 2.5x control for merchants inactive 5-10 months, 25% more converted loan volume and 41% more converted loans per merchant on stalled applications, "0 additional hires"; an 8-page PDF, ungated.Source
• Your websiteIndustry pages: fintech lenders, financial institutions and card issuers each claim "95%+ resolved. End to end." with no source or period, next to sourced figures of "70%+" (AWA) and "74%" (Splash). The card-issuer and banks & credit unions pages show no customer story from their own segment. The agencies page promises "30-day" deployment; integrations are described only as "REST, webhooks, and native connectors across your existing stack", with no named systems.Source
• Competitor researchRivals' own homepages: Salient, "AI built for consumer lending", "$75M raised", logos such as Westlake Financial and Ally Financial and public release notes; Skit.ai names its integrations (Twilio, Genesys, TCN, PayNearMe, Finvi) and invites prospects to benchmark their portfolio; Prodigal names Five9, Genesys and NICE for onboarding "within weeks" and markets Spanish-language notes.See the research Source

Invite agency customers into a small owners' circle: a quarterly call on what's working, early access to new agents, and a referral credit for each introduced agency that starts a pilot. Offer prospects a 20-minute call with a live customer before the AE demo.

Peer referrals close faster in a tight-knit industry, and the circle also improves retention by giving customers a voice in the roadmap.

• Customer fatigue from reference calls

  Cap at two calls per customer per quarter.

• #13 Turn Second Alliance's "AI subservicer" quote into the agency case study it promises

• Your websiteHomepage: H1 "The AI Communications Platform for Financial Services"; six use cases (customer acquisition, welcome calls, loan servicing, customer experience, early-stage and late-stage collections) for fintech lenders, card issuers, financial institutions and third-party agencies; CTAs "Speak to an AI agent" (to app.veritus.com/try), "Get in touch with a human" and a "Launch app" button for existing users; stats "71%" containment and "<5%" of borrowers asking whether they are speaking with AI; testimonials from AWA, Splash Financial, Second Alliance ("Veritus acts as my AI subservicer") and a former LendingClub consumer banking executive; "Backed by Y Combinator."Source
• HiringFounding Account Executive (San Francisco, posted July 29, 2026): "this is our first full-cycle sales hire" after "founder-led sales"; targets "Collections agencies and midsize lenders first", then enterprise lenders; works conferences RMAI, ACA, Money20/20 and LendIt; "the constraint on growth is selling capacity"; tools include HubSpot, Apollo and LinkedIn Sales Navigator. Calls consumer lending and collections "among the last corners of financial services that AI has not reached".Source
• Your websiteAWA case study (inbound collections agency): "25%" of dollars the agent (Linda) collects come from after-hours calls that previously went to voicemail; "67%" of customers pay on the call; "71%" containment; one in five fully handled calls ends in a payment; the agent started after hours, then went 24/7; the earlier vendor's SMS payment links were flagged as spam by carriers. PDF ungated.Source

Serve images in modern formats at the displayed size, set long cache lifetimes for hashed Astro assets and fonts, and drop the legacy JavaScript polyfill. Keep the case-study banner, the LCP element, as plain text with no blocking fonts.

It is a few hours of work in Astro and Cloudflare, it keeps the site fast as more media and case studies are added, and it reverses the slowing trend.

• Image quality loss

  Review key visuals after compression.

• Google PageSpeedGoogle PageSpeed (mobile, November 9, 2026): score 89; lab LCP 2.6s (the element is the case-study banner); 9.4 MB homepage with 3.5 MB of images (2.3 MB saveable) and about 5 MB of resources with short cache lifetimes. Real Chrome users pass Core Web Vitals: LCP 0.88s, INP 60ms, CLS 0.01 at the 75th percentile, though real-user LCP is trending 11% slower.Source

Publish a small set of plain, dated question-and-answer pages: AI voice agents and TCPA consent, Reg F call caps with AI, state AI-disclosure rules, PCI payments on AI calls, and how QA of 100% of calls works. Each answers in the first paragraph, cites sources and links the deeper brief. Add an llms.txt for the marketing site that matches the docs.

Search engines and AI assistants quote clear, dated, sourced answers, and these are the questions buyers ask before they ever fill a form.

• Answers read as legal advice

  Keep the not-legal-advice note and cite primary sources.

• #5 Date every insight, starting with the TCPA compliance brief

• Your websiterobots.txt: Content-Signal "search=yes, ai-train=no"; AI training crawlers are disallowed while search and assistant crawlers fall under the default allow; the docs site links an llms.txt index.Source
• Your website"Q1 2026 Buyer's Guide: Virtual Agents in U.S. Consumer Collections", written by an outside collections-technology advisory firm, gated behind a form, with an Excel RFP requirements template. The TCPA compliance brief is ungated, carries no date, and ends with an optional "Building an AI voice program?" form.Source
• Your websiteInsights hub: 9 items with no publication dates shown: 3 case studies (Wisetack, Splash Financial, AWA), 3 white papers (including "TCPA consent for AI voice agents" and "Turn application drop-off into revenue", whose CTA is "Download the playbook"), 1 blog post, 1 video and 1 compliance brief. No newsletter or subscribe option on the hub or the blog post.Source