The new homepage sells to financial services, healthcare and real estate operators. Those industry pages carry the strongest numbers on the site: $120K a year saved for an 8-partner CPA firm, a 34% drop in no-shows, and $85K saved plus a 90% faster intake for a dental group. Yet none of the six "Read Case Study" buttons opens the story it advertises. The voice AI button goes to /case-studies/voice-ai-patient-appointment/, which served the homepage both times it was fetched. The other five send the visitor to the general listing, and none of the 55 case studies in the sitemap is a CPA, tax, bookkeeping, dental or clinic project. A buyer who clicks for proof lands back where they started.
First action
Repoint the voice AI card to /case-studies/AI-Voice-Receptionist-Booking-System today.
How to measure it
Industry-page consultation bookings
- Now
- –
- Target
- +30%
- By
- 60 days
Steps
Fix the dead link
Repoint the voice AI card to the existing receptionist study, and hide the buttons on cards with no page.
Frontend Developer 1 hour
Gather the facts
Scope, systems, timeline and before and after numbers for the six projects on the cards.
Head of Delivery 3 days
Get sign-off
Ask each client to approve an anonymised write-up, or a named one if they agree.
Sales & Client Success Lead 1 week
Write the studies
One page each: problem, build, architecture, results, the client's acceptance and a quote.
Digital Marketing Manager 1 week
Publish and link
Lowercase URLs, with each card linking to its own study.
Frontend Developer 1 day
Feature on the homepage
Put the two strongest in "See client results".
Digital Marketing Manager 1 day
Track
Measure case-study clicks through to consultation bookings, by industry.
Digital Marketing Manager monthly
Why it works
Write up the six projects the cards already summarise. Anonymised is fine: firm size, region, the systems connected, before and after, the timeline and what the client signed off. Publish each at its own URL and link every card to it directly. Today, point the voice AI card at the existing AI Voice Receptionist Booking System study, and remove the button from any card that has no page yet. Then add the two strongest studies to the homepage "See client results" section, and use one accuracy figure on the finance page instead of both 99% and 97%.
Industry-page visitors are the buyers the new homepage wants, and clicking for proof is the highest-intent action on those pages. Buyers in regulated industries want to see a firm like theirs before they book an architecture call. The numbers already exist, so this is writing work, not new delivery.
Ready to use
ChecklistIndustry proof checklist
[ ] Voice AI card repointed to an existing study [ ] CPA firm study ($120K a year, 38 hours a week) [ ] Tax-season document pipeline study [ ] Bookkeeping onboarding study (12 hours to 45 minutes) [ ] Clinic insurance verification study [ ] Dental intake and billing study ($85K, 90%) [ ] Each card links to its own URL [ ] One accuracy figure on the finance page [ ] Two studies on the homepage [ ] Client approval on file for every number
Risks and how to handle them
Clients will not be named
Publish anonymised studies with firm size and region, and keep named references for calls.
Figures can't be backed up
Only publish numbers the client has seen and approved.
Evidence
- Your websiteHealthcare page: headline "Your Staff Spends 60% of Their Day on Admin Instead of Patients"; three case cards with results ("34% reduction in no-shows", "85% of insurance verifications automated", a dental group with "$85K" saved and a 90% cut in intake time). The voice AI card's "Read Case Study" button goes to /case-studies/voice-ai-patient-appointment/, which served the homepage when fetched twice on November 19, 2026; the other two cards link to the general /case-studies/ listing.Source
- Your websiteFinancial services page: headline "Stop Losing $150K+ Every Tax Season to Manual Work"; three case cards (an 8-partner CPA firm with "$120K annual savings" and 38 hours a week saved, a tax-season document pipeline, and a bookkeeping practice whose onboarding fell from 12 hours to 45 minutes), and all three "Read Case Study" buttons link to the general /case-studies/ listing. Extraction accuracy is given as 99% in one place and 97% in another.Source
- Your websiteCase study sitemap (read November 19, 2026): 55 case study pages, none for a CPA, tax, bookkeeping, dental or clinic client; clients are unnamed (for example a Mumbai mattress retail chain with "2+ hours" saved daily on /case-studies/google-review-automation); slugs mix cases ("LinkedIn-" and "Linkedin-"), one has a typo ("Advanced-Googl-Apps-Script") and three contain "&". The /case-studies/ listing showed "Showing 0 projects" to a reader that does not run JavaScript, and the services page refers to "52 real-world case studies".Source
- Your websiteHomepage (read November 19, 2026): H1 "Enterprise AI, actually shipped.", aimed at financial services, healthcare and real estate operators in the US, UK and Australia; "2,000+ Projects Delivered"; CTAs "Book an Architecture Call" and "See client results"; a "live architecture" diagram of an after-hours patient call booked by a voice agent ("avg. handle 38s", "0 missed calls") that is not tied to a client; no testimonials, client names or logos. The meta description still reads "Make.com Platinum Partner with 100+ experts and 2000+ projects delivered. Free automation audit."Source