Between the relaunch of this site on 24 April 2026 and 13 October 2026, Vridhii received twelve inbound enquiries we can attribute to a source: five from one Meta campaign for clinics, two from a Meta campaign for manufacturers, one from a blog post via ChatGPT, and four by telephone, two of which the caller attributed to a Google search and one to ChatGPT. The website's forms recorded eight of the twelve. The blog, 121 posts on the day of counting, produced one. Nothing has arrived through a form since 29 August.
This is a Wednesday data post and it is the one the operating rules put first in the queue, our own funnel numbers by source, which we have not published before because the figure that fills the gap in our data ledger since August was sitting in a spreadsheet nobody had read end to end. We read it on 13 October. Here is what it says, with the names left out, as our confidentiality rule requires.
What the form sheet holds
Every form on this site posts to one Google Form, and its response sheet had 23 rows on 13 October 2026. Sorted by what they are rather than when they arrived:
| Kind of row | Count | What they were |
|---|---|---|
| Genuine enquiries | 8 | Five clinics and doctors, two manufacturers, one home interiors business |
| Our own tests | 8 | Form checks by the team, including the pricing page dry runs |
| Vendor pitches, applicants, junk | 7 | A guest post pitch, a sponsored listicle offer, a mass email from a fashion brand, a job applicant, two junk rows, one UK sample film test |
Eight real enquiries out of 23 rows is a 35 percent signal rate, and that is a useful number on its own: a third of what a form collects is a lead, a third is you, and a third is somebody selling to you.
Where the twelve came from
Add the four telephone enquiries logged in our data ledger on the day each arrived, and the picture by source looks like this.
| Source | Enquiries | Dates | What they wanted |
|---|---|---|---|
| Meta ads, clinic acquisition page | 5 | 29 May to 18 June | A homoeopathy clinic, an orthopaedic hospital, two general physicians, one incomplete |
| Meta ads, manufacturing growth page | 2 | 1 and 4 June | An office furniture maker, a PVC pipe maker |
| Blog post via ChatGPT referral | 1 | 29 August | A home interiors business, from the interior designer marketing post |
| Telephone, caller cited ChatGPT | 1 | 7 September | An interior design practice |
| Telephone, caller cited Google search | 2 | 8 and 22 September | A prospect, and a clothing retailer |
| Telephone, source not stated | 1 | 15 September | A second interior design practice |
| Homepage, service pages, 121 blog posts, organic | 0 by form | No form enquiry has ever arrived from an organic visit to a page other than the ChatGPT referred post |
The original number. 12 inbound enquiries attributable to a source between 24 April and 13 October 2026: 7 from two Meta campaigns on the clinic and manufacturing pages (29 May to 18 June), 1 from a blog post carrying a ChatGPT referral (29 August), and 4 by telephone (7, 8, 15 and 22 September), of which 2 were attributed by the caller to Google search and 1 to ChatGPT. The form sheet recorded 8 of the 12, the telephone 4, and 0 arrived by form from any organic visit to a non blog page. Logged as data ledger entry 62, the first enquiries by source entry since the ledger opened, and the figure its status notes have called missing since August.
What the numbers say, plainly
Paid ads filled the sheet and nothing else did. Seven of eight form enquiries came from two Meta campaigns that ran for about three weeks in late May and June and then stopped. From 19 June to 28 August, ten weeks, not one real enquiry arrived by form. The site published roughly fifty posts in that window.
The forms do not see the calls. Four of the twelve enquiries came by telephone, to a number that appears on no page of this site, which is the finding of our 23 September post about the enquiry analytics could not see. A measurement that misses a third of the enquiries is the measurement we have, and it is why every call now goes into the ledger by hand on the day.
AI search has produced two enquiries, not a trend. One form on 29 August carried a ChatGPT referral tag from the interior designer post, and one caller on 7 September said ChatGPT sent them. Both were interior design businesses, which is why that vertical got its own shortlist and playbook in September. Two in 24 weeks is proof the pages are read by an engine. It is not yet a channel.
The funnel did not close. Our 5 September ledger entry records that none of the roughly thirty leads worked between April and early September, the twelve above plus WhatsApp and referral conversations the sheet does not hold, became a client. The ads did their job. What came after did not, and that is a follow up problem, not a traffic problem.
What we are changing because of it
Three things, all ours to do.
First, the daily audit now reads the whole response sheet, not its modified date, and files every genuine row into the ledger by source on the day it lands. The gap between 29 August and 13 October would have been noticed a month sooner.
Second, from today every weekday post gets a short LinkedIn draft written alongside it and held in the repository for the founder to post. One hundred and twenty one posts producing one form enquiry is the clearest argument this site has made for putting the work in front of people rather than waiting for an engine to do it.
Third, the clinic campaign's result, five enquiries in 21 days, is proposed for the clinic acquisition page as a dated first party figure, for the founder to approve. That page sold the service that produced most of our leads this year and says nothing about it.
What you should take from this if you run a business in Hyderabad
Count the enquiries that arrive by phone and WhatsApp, not only the ones that fill a form, because the form will flatter whichever channel you can track. Run the one paid campaign that works for a defined period and write down what it produced, dated, before you decide it was expensive. And do not expect a blog to produce enquiries on its own; expect it to produce the page a referred buyer checks before they call, which is a different job and a slower one.
The Marketing Operating System is the weekly rhythm that keeps a sheet like this read every morning rather than once a quarter, and our pricing page explains how we scope the follow up work that a full top of funnel needs.
How we counted
Form rows from the Google Form response sheet all 135 forms on the site post to, read in full on 13 October 2026 and classified by hand. Telephone enquiries from blog/data_ledger.md entries 31, 34, 39 and 45, logged on the day each call arrived with the source the caller gave. WhatsApp enquiries are excluded because they are not yet logged consistently. Names and contact details are withheld under our client confidentiality rule; categories and dates are exact.
Frequently asked questions
Why publish numbers this small?
Because they are the numbers. A dozen enquiries in 24 weeks is what a small studio that ran one paid campaign for three weeks and otherwise published actually gets, and a reader deciding whether to copy our method deserves to know that before copying it. The operating rules we adopted on 9 August require the unflattering figure with an explanation, and this is the explanation.
Does the form sheet capture every enquiry?
No. Four of the twelve came by phone to a number that appears on no page of this site, so the sheet saw none of them. We log those in the data ledger by hand on the day, with the source the caller gave us. WhatsApp messages from the site's buttons are not counted here either, which means the true figure is higher than twelve and we cannot yet say by how much.
Was the clinic campaign worth it?
Five enquiries in 21 days from one Meta campaign is the best rate anything on this site has produced. Whether it was worth it depends on what happened next, and our 5 September ledger entry records that none of the roughly thirty leads worked between April and September became a client. The campaign filled the top of the funnel; the funnel did not close. That is the part we are fixing, not the ads.
What does one enquiry from ChatGPT prove?
Only that it can happen. One caller on 7 September said they found us through ChatGPT, and one form on 29 August arrived carrying a ChatGPT referral tag from the interior designer post. Two in 24 weeks is not a channel; it is evidence that the pages are being read by an engine, which is what the Friday posts are for.