The subject
Found from one Facebook reel. Everything below came from public sources: his own pages, his own published screenshots, the Meta Ad Library, DNS and WHOIS.
| Person | Jason Meng |
|---|---|
| Agency | theUpscale |
| Base | Phra Khanong, Bangkok |
| Niche | Health and beauty clinics: aesthetic, chiropractic, wellness / IV drip |
| Sells | Booked, pre-qualified patient appointments on a performance promise |
| Landing pages | theupscale.io/automated-patient · the-upscale.com |
theupscale.clinicagency · 61587195276244TheUpscale Agency · 61553696545402 | |
| Elsewhere | LinkedIn · YouTube · Klixy.io (his agency incubator) |
| Stack | GoHighLevel, white-labelled via sites.ludicrous.cloud |
| GHL location | m5NREswZegfp4oWExNAt |
| Meta pixel | 547105601186501 |
| Entry point | FB reel 1708729853583167: "WARNING: Our AI Booked 10-40 Patients a Month To Your Clinic" |
His claim About $2M a year. Price went from $200 per month per client in 2023 to $26,000 per client in 2026. Unverified, from his own LinkedIn.
He does not run Meta ads
The least expected finding, and the most useful one.
How it was checked
Ad Library API on both page IDs, every country filter. Then the Ad Library web interface at country=ALL, Thailand and US. Every query came back empty.
Why you can trust the zero Verified
The identical query was run against MAG Security System in the same session. It correctly returned 44 ads. So the tool works. The zero is real, not a lookup failure.
What he does instead: organic short-form video on Facebook and YouTube, plus outbound. The reel that started this teardown is an organic post, not a paid ad.
Lesson for us: an agency selling paid ads can reach $2M a year without running paid ads on itself. A competitor missing from the Ad Library is not necessarily small.
Method note Ad Library keyword search is worthless. Searching "theUpscale clinic" returned mobile game and web novel ads. Page-ID search is the only reliable route. Second time we have proved this.
His funnel, end to end
Reconstructed from his landing pages and his own published delivery screenshots.
Organic reel or outbound DM
No paid traffic to himself. Short-form video with a warning-style hook, or direct outreach.
Landing page with a size gate at the top
"For Aesthetic Clinics Above $10K/month" sits above the headline. Filters before anyone reads the offer.
Form-gated calendar
Form and calendar sit side by side above the fold. The calendar is visible but locked until the form is filled. Phone number is asked before the name.
Free clinic audit call
30 minutes. This is his sales call, dressed as a diagnostic.
Client onboard in 7 days
He builds the ads, the copy, the video edits, a custom sales page and the automations. All inside a week.
Client ads sell a named treatment at a named price
Not "book a consult". A specific product with a number on it. See section 05.
Lead lands in a tracking sheet, then a human chases it
Called, Email, SMS, WhatsApp. Four manual channels, worked by a named account manager. See section 06.
Weekly scoreboard posted into a shared group with the client
Same Monday scoreboard ritual we designed for MAG. See section 07.
His landing page, taken apart
theupscale.io/automated-patient/. One screen of persuasion, then the booking widget. That is the whole page.
The one detail worth copying today: he asks for the phone number before the name. The most valuable field is captured first, so a drop-off still leaves a contactable lead.
His second site, the-upscale.com, is the fuller one: nav is Results, Free Masterclass, Pricing Plans, Book a Call. It carries a scarcity bar ("Only 4 spots for new clients this month"), a published pricing page and a masterclass as a softer second entry point. Best line on either site:
"Your business lives or dies by the number of show-ups you get to your doors and that's what we GET YOU. Not 'leads', but show-ups."
The ad format: a price, not an invitation
This is the single biggest difference between his client ads and ours.
What his client ads actually offer Observed
Read straight off the Offer Claimed column in a live client tracking sheet:
NAD+ Drip THB 7,500NAD+ Drip THB 5,460same offer later. He cut the price and kept testing.Metabolic Fire IV 30% OFFWeight Loss Programme
A named product with a number attached. The lead arrives already knowing roughly what they are buying, which is what his "prepaid patients" headline is built on.
This is the same principle as Verdant Solar in our own SOP, by a different mechanism. Verdant publishes a floor (RM17,000 minimum). Jason publishes the actual price of one entry treatment. Either way: a number filters, an invitation does not.
Our MAG offer is currently a free audit, which is an invitation with no number in it. Worth testing a priced entry offer beside it.
His delivery sheet
He published this as proof. It is the most valuable thing in the whole teardown.
Offer Claimed and Lead Condition Details. Currency is Thai baht, so this is a Bangkok wellness clinic.
The column map
Time | Date | First Name | Last Name | Whatsapp | Email
Offer Claimed | Lead Condition Details | Additional Notes
Contact Time | Called? | Email? | SMS? | WhatsApp?
Booked | Showed | Package | Booking Date | Booking Time | Sales | NOTES
Lead Condition Details is the field that matters. It holds the lead's own stated reason, captured at opt-in. Observed values: longevity and anti-aging, better sleep and mood, overall wellness boost, energy and mental clarity, boost energy and metabolism, stress and burnout recovery, weight regain, kickstart weight loss.
The caller opens the conversation already knowing why the person raised their hand. It is one extra form field and it is the cheapest quality lever in the whole funnel.
Caution The "AI" is a human call-down. His ad says "Our AI Booked 10-40 Patients a Month". His own sheet shows four manual contact checkboxes worked by a named human account manager, and the NOTES column is dominated by Unreachable and Left on read. On the page sampled, roughly 11 of 26 leads were never reached at all.
Useful two ways: it shows what this market will tolerate as a claim, and it shows the real cost base is labour, not software.
His real numbers
Three client scoreboards he published himself. All three reconcile arithmetically, so they are very likely real.
Verified Why we believe these. On all three: leads × cost per lead reconciles to ad spend, and cash collected minus ad spend reconciles exactly to his stated profit. Invented numbers do not usually survive that check.
| Stage | 🇲🇾 Malaysia | 🇹🇭 Thailand | Chiro (USD) |
|---|---|---|---|
| Leads | 2,793 | 1,590 | 1,079 |
| Cost per lead | RM29 | ฿405 | $16 |
| Ad spend (derived) | ~RM80,860 | ~฿643,846 | $16,936 |
| Booked appointments | 354 | 300 | 309 |
| Booking rate | 13% | 19% | 29% |
| Cost per booked appointment | RM228 | ~฿2,146 | ~$55 |
| Show-ups | 180 | 151 | 289 |
| Show-up rate | 51% | 50% | 94% |
| Cost per show-up | RM449 | ~฿4,264 | ~$59 |
| Packages sold | 84 | 47 | 9 |
| Show to sale | 47% | 31% | 3% |
| Cost per sale | RM963 | ~฿13,699 | ~$1,882 |
| Cash collected | RM555,631 | ฿1,068,615 | $93,097 |
| His stated "ROI" | 687% | n/a | 550% |
The number we did not have
Our SOP and HANDOFF both record that cost per booked appointment for Malaysia does not exist publicly. It does now, for aesthetic and wellness clinics.
Use it as a discovery anchor, not a promise. One agency, one vertical, consumer health. Our archetypes are B2B project work with a bigger ticket and a longer cycle, so expect a higher cost per appointment and a much longer path to a sale. The useful part is the shape: roughly 8x from lead to booked appointment, and roughly 33x from lead to sale.
His "ROI" is ROAS, and it hides his fee. 687% is only cash collected divided by ad spend. It does not subtract his retainer. If his stated $26k per client is close to right, adding it drops that account from about 6.9x toward roughly 2-3x. Still good. Not the number on the screenshot.
So we report ad spend plus our fee, and the client's true cost per sale. He does not. Cheap credibility moat.
The chiro account is the warning. 94% show-up looks spectacular until you read the next row: 3% show to sale, 9 sales from 289 attendances. High show-up with a collapsed close means the offer pulled the wrong people cheaply.
Show-up rate alone is a vanity metric. Our scoreboard must always carry show-to-sale next to it.
His claim He misses his own targets. His published internal goals are 55%+ booking, 65%+ show-up, 20%+ closing. His actual accounts run 13%, 19% and 29% booking. A $2M a year agency is nowhere near its own targets. Read that as permission.
His clients
The hosting route failed. The logo wall worked. Both results are worth keeping.
Verified dead end Shared hosting cannot identify his clients. Both his domains point to sites.ludicrous.cloud, but that is not his server. It is a third-party white-label GoHighLevel platform sold to many agencies. Certificate transparency shows whitelabel., clientportal., brand. and vibe. subdomains with matching staging copies. WHOIS: Cloudflare registrar, Texas, created March 2024. One of its edge IPs alone reverse-resolves to 499+ unrelated domains: churches, accounting courses, browser games.
Rule for next time: a CNAME into a white-label SaaS identifies the vendor, never the customer list. Reverse-IP on any Cloudflare address is worthless for attribution.
Caution These named clients are not running his funnel. The Clifford Clinic runs 5 active ads in Singapore: EMFACE, Pico Laser, acne treatment, nose thread lift. Every call to action is "Send message", a phone number, or their own website page. Not one goes to a landing page or a booking calendar.
A sweep of go. book. offer. promo. lp. apply. start. get. funnel. landing. on all six client domains found nothing pointing at his host. So the logo wall is either historical or aspirational. His current live accounts are the ones in the screenshots, and those are deliberately unnamed.
Our version of this: the logo wall only carries accounts we can show numbers for.
Where his proof is soft
Not to sneer. To calibrate. This is the credibility bar we have to clear, and it is lower than it looks.
Client count moves
"200+ practices", "250+ clinics" and "300+ thriving clinics" all appear across his two sites.
Same money, two years
The same $5,240,000 figure is credited to both 2024 and 2025, in two blocks on the same page.
The qualifier moves
His Facebook page says clinics above $20k/month. His landing page says above $10k/month.
The AI claim
Does not match his own delivery sheet, which shows a human working four manual channels.
"ROI 687%"
Is ROAS on ad spend only. Excludes his own fee entirely.
The logo wall
Named clinics whose live ads do not run his funnel at all.
None of this means he is not good. The numbers in section 07 look real and the operation is clearly working. It means the proof bar in this market is low, and an agency with genuinely reconciled, fee-inclusive numbers can beat him on credibility alone.
What we steal, and what we do differently
Three moves worth taking, in order of value. Then three places we deliberately go the other way.
Put a price on the offer in the ad
Ours offers a free audit, which is an invitation. His offers a named treatment at a named number. A price qualifies before the click, exactly like Verdant Solar's published RM17,000 floor. Different mechanism, same principle: a number filters, an invitation does not.
Capture the "why" at opt-in, then route the first call on it
His Lead Condition Details field holds the lead's own stated reason, so the caller opens knowing it. One extra form field, and the cheapest quality lever available. Pairs with Verdant's language-routing field. Both go straight into the AI qualifier build with Steven.
Sell show-ups, not leads
His line is sharper than ours: "Not 'leads', but show-ups." Our unit is already the booked appointment. His framing says it better. Take the phrasing.
Where we go the other way, on purpose
Fee-inclusive economics
He shows ROAS on ad spend only. We show cost per sale including our retainer. Harder number, much stronger trust.
Show-up never travels alone
Always paired with show-to-sale. His chiro account proves why: 94% show-up, 3% close.
No fake AI claim
We do not say AI does the calling unless it actually does. His sheet shows a human doing it.
Still open: his pricing page was not opened in this pass, and it is worth a look before we set our own ladder. Klixy.io, his equity-share agency incubator, is a separate business model worth studying if we ever productise the SOP.
Method and sources
Everything here came from public sources. Nothing was submitted, logged into or scraped behind a login.
What was used
- His two landing pages and their page assets
- His own published proof screenshots
- Meta Ad Library, page-ID method
- DNS, WHOIS, certificate transparency, reverse IP
- His public LinkedIn and YouTube
Two method rules earned here
- Ad Library: search by page ID, never by keyword. Keyword search returned unrelated mobile game ads.
- Hosting: a CNAME into a white-label SaaS identifies the vendor, not the customer. Reverse-IP on Cloudflare is worthless.
Where this lives
- Full write-up:
projects/lead-gen-service/competitor-teardown-theupscale.md - Google Doc: Drive › 1-Project › Lead Gen
- Lessons folded into the SOP:
methodology-sop.mdPart 1C and Part 4
Sibling studies: the visual playbook is our own method. The delivery blueprint is technique distilled from 12 YouTube educators. This page is one live operator, observed. Kept separate on purpose.