Zealous Productions. Live theatre, Perth.

$39.2k in ticket sales from $6.8k of Meta ads.

A limited theatre season, a fixed number of nights, and one Meta sales campaign over 38 days. 197 ticket orders at $34.67 each, every one counted at the ticketing checkout.

5.74x
back on every dollar of the sales campaign
One purchase action, read at the ticketing checkout. · Meta Marketing API, ad account id withheld, available on request

Every dollar of that was written by the ticketing partner's own checkout, so it is what buyers actually paid.

A season with a hard deadline

Zealous Productions is a Perth independent theatre producer with a limited season and a fixed run of performances. Tickets sell through a national ticketing partner's checkout, not the producer's own website.

Sell the seats before the last night

Once the last night has played, an unsold seat is worth nothing, so every week of the campaign had to sell. The producer's own boosted posts were built for engagement, not sales. They needed ads that sold tickets, and a return they could budget the next season on.

What we did

Three calls, made before the first dollar was spent.

  1. One campaign built to sell tickets

    We ran every ticket through a single sales campaign optimised for purchases on one pixel, so Meta knew exactly which buyer to go and find, and every order taught it more.

  2. Tracking where the money changes hands

    The purchase event fires from the ticketing partner's own checkout, so every order carries the price the buyer paid. No guesses and no typed-in values.

  3. Every sale counted once

    Meta reports the same order under several names. We read one, so the return is built on real orders and never inflated by adding the aliases together.

Why it worked

A campaign optimised for purchases tells Meta who buys, not who taps like. Across the same season the boosted posts in the account sold no tickets at all. The sales campaign sold 197 orders.

The result

197 ticket orders and $39.2k of tracked ticket sales value from $6.8k of campaign spend. That is 5.74 times back, or $34.67 per order. Count every dollar in the account, boosted posts included, and it still returned 5.42 times.

VerifiedWhat it measures
5.74x
tracked ticket sales value on the sales campaign
Meta standard attribution: 7-day click, 1-day view, 1-day engaged view.
The query behind it
one 38-day campaign, queried 01/06/2026 to 31/08/2026
Meta Graph API. GET /{sales_campaign_id}/insights fields=spend,actions,action_values time_range={since:2026-06-01,until:2026-08-31}, ad set attribution_spec CLICK_THROUGH 7d, VIEW_THROUGH 1d, ENGAGED_VIDEO_VIEW 1d, action_type=offsite_conversion.fb_pixel_purchase: value 39203.25 on spend 6829.80 = 5.74
197
ticket orders from the sales campaign
One purchase action type, never summed with its aliases.
The query behind it
one 38-day campaign, queried 01/06/2026 to 31/08/2026
Meta Graph API. GET /{sales_campaign_id}/insights fields=actions time_range={since:2026-06-01,until:2026-08-31}, action_type=offsite_conversion.fb_pixel_purchase = 197. Re-derived from ad set, ad and daily rows
$39.2k
tracked ticket sales value from $6.8k of campaign spend
Order value at the ticketing checkout, booking fees included.
The query behind it
one 38-day campaign, queried 01/06/2026 to 31/08/2026
Meta Graph API. GET /{sales_campaign_id}/insights fields=spend,action_values time_range={since:2026-06-01,until:2026-08-31}, action_type=offsite_conversion.fb_pixel_purchase = 39203.25; spend = 6829.80
$34.67
per ticket order on the sales campaign
Campaign spend divided by the 197 orders.
The query behind it
one 38-day campaign, queried 01/06/2026 to 31/08/2026
Meta Graph API. GET /{sales_campaign_id}/insights fields=spend,actions time_range={since:2026-06-01,until:2026-08-31}: 6829.80 / 197 offsite_conversion.fb_pixel_purchase = 34.67
5.42x
over every dollar in the account, boosted posts included
All account spend in the denominator.
The query behind it
one 38-day campaign, queried 01/06/2026 to 31/08/2026
Meta Graph API. GET /{ad_account_id}/insights level=account fields=spend,action_values time_range={since:2026-06-01,until:2026-08-31}, default attribution, action_type=offsite_conversion.fb_pixel_purchase: 39203.25 / 7233.32 = 5.42

How we checked it

Every number on this page was pulled live from the ad platform and checked a second time by a reviewer whose only job was to prove it wrong.

Open the method, the caveats and the checks

Every figure above was read live from Meta Marketing API on 01/11/2026, then independently re-derived by an independent second agent instructed to refute every claim, re-derived by campaign, ad set, daily and account routes on 01/11/2026. Figures that failed that pass are not on this page.

What to read alongside it

  • Nobody ran a holdout test on this season, so nothing here is called incremental. Some buyers would have bought anyway.
  • A Google Search campaign also ran for about a week of the season, and Meta does not deduplicate its orders against Google.

What we left out, and why

  • No figure for what the producer kept from each ticket, because their share of the ticket price is a number we have never been given.

The checks it passed

Double countMeta lists each order under several purchase names. We read one. No custom conversion counts orders a second time.
Hardcoded valuePassed. The same tag passes a flat default on product views, and the purchase values look nothing like it. They move with the number of tickets bought.
Methodology breakNone inside the run. One sales campaign, one pixel and one attribution setting from launch to the last night.
Engagement attributionThe sales campaign was our build from its first day, on our own build records. Meta's log does not name who made each change, so the boosted posts are shown in the account figure rather than argued out.

Selling a season, an event or a launch?

Send us your ad account and we will show you what we would change, in plain numbers. It takes us about an hour, and you keep what we find either way. No obligation.

Impact Dents. Vehicle repair, Perth.

1,100 enquiries for one Perth workshop, at $9.53 each.

One Search campaign, twelve months, an average of $874 a month. 1,092 quote forms and 8 calls straight from the ads, at a 15.79 per cent click-through rate and $1.65 a click.

$9.53
per real enquiry, duplicates removed
$10,487.66 of media spend over 1,100 enquiries · Google Ads API, account id withheld, available on request

That is the clean number. The account's own column said 2,115 enquiries. We traced the double count, fixed it, and report only the 1,100 that were real.

One workshop, one campaign

Impact Dents is a Perth dent and vehicle repair specialist with one workshop. The job is simple to say and hard to do on a small budget: put a steady stream of quote requests in front of a technician, by form or by a call placed straight from the ad.

A small budget with no room for waste

On under $900 a month, every click has to be someone who wants a repair. Underneath, two conversion tags were recording the same form, so the bidding was chasing enquiries that did not exist.

What we did

Proof first, then the fix that saves money, then a number the owner can trust.

  1. One campaign, never split thin

    Every dollar runs through one Search campaign, so a small budget is never spread across half-funded experiments.

  2. Fix the count before the bidding

    We proved the double count from the daily rows, then took the duplicate out of the bidding on 30/07/2026 so Google stopped buying volume that was not there.

  3. Report the real number

    We report the 1,092 form enquiries plus 8 calls from the ads, 1,100 in all, and nothing padded on top.

Why it worked

A 15.79 per cent click-through rate at $1.65 a click is what tight relevance looks like: the right searches, bought cheaply. Clean tracking then let the bidding chase real enquiries instead of phantom ones.

The numbers

Cost per enquiry, before and after the correction

Fixing our own double count doubled the cost per enquiry we had been reporting.

Cost per enquiry, before and after the correctionFixing our own double count doubled the cost per enquiry we had been reporting.$4.96$9.53As the account reported itDeduplicated, 19/08/2026+92.3%
Show the numbers
PeriodValue
As the account reported it$4.96
Deduplicated, 19/08/2026$9.53

Deduplicated enquiries, month by month

Volume peaked at 124 in January 2026 and no month since has matched it.

Deduplicated enquiries, month by monthVolume peaked at 124 in January 2026 and no month since has matched it.037.575112.5150peak, 1242025-082025-102025-122026-022026-042026-062026-083042

The first and last months are part months, 13 days and 18 days, cut by the window and not by a spend change. The dashed segment is that final part month.

Show the numbers
monthDeduplicated enquiries, month by month
2025-0830
2025-0967
2025-1079
2025-1194
2025-12100
2026-01124
2026-02101
2026-03108
2026-0496
2026-0598
2026-0683
2026-0778
2026-0842

Average cost per click, month by month

The price of a click climbed from $1.41 in February to $1.92 in July, and August is running at $1.92 on 18 days counted so far.

Average cost per click, month by monthThe price of a click climbed from $1.41 in February to $1.92 in July, and August is running at $1.92 on 18 days counted so far.$0.00$0.62$1.25$1.88$2.502025-082025-102025-122026-022026-042026-062026-08$1.43$1.92

Spend held near $910 a month from October 2025, so this is the price moving and not the budget. The dashed segment is August 2026, which carries 18 days of the window.

Show the numbers
monthAverage cost per click, month by month
2025-08$1.43
2025-09$1.68
2025-10$2.16
2025-11$1.55
2025-12$1.49
2026-01$1.47
2026-02$1.41
2026-03$1.51
2026-04$1.64
2026-05$1.70
2026-06$1.77
2026-07$1.92
2026-08$1.92

The result

1,100 tracked enquiries in the twelve months to 18/08/2026, 1,092 form submissions and 8 calls from the ads, on $10,487.66 of spend. $9.53 each, at a 15.79 per cent click-through rate and $1.65 a click.

VerifiedWhat it measures
1,100
tracked enquiries, deduplicated
Counted per click and not per person.
The query behind it
19/08/2025 to 18/08/2026
SELECT conversion_action.name, conversion_action.id, metrics.conversions FROM conversion_action WHERE segments.date BETWEEN '2025-08-19' AND '2026-08-18' -- 1,092 native tag plus 8 calls from ads
$9.53
cost per tracked enquiry
Media cost only. No GST, no management fee.
The query behind it
19/08/2025 to 18/08/2026
SELECT metrics.cost_micros FROM customer WHERE segments.date BETWEEN '2025-08-19' AND '2026-08-18' -- $10,487.66 divided by 1,100
15.79%
click-through rate
The account's lifetime rate is 14.44 per cent.
The query behind it
19/08/2025 to 18/08/2026
SELECT metrics.ctr, metrics.clicks, metrics.impressions FROM customer WHERE segments.date BETWEEN '2025-08-19' AND '2026-08-18'
$1.65
average cost per click
$1.66 across the account's whole life.
The query behind it
19/08/2025 to 18/08/2026
SELECT metrics.average_cpc FROM customer WHERE segments.date BETWEEN '2025-08-19' AND '2026-08-18'

How we checked it

Every number on this page was pulled live from the ad platform and checked a second time by a reviewer whose only job was to prove it wrong.

Open the method, the caveats and the checks

Every figure above was read live from Google Ads API on 19/08/2026, then independently re-derived by an independent second agent instructed to refute every claim. Figures that failed that pass are not on this page.

What to read alongside it

  • 1,100 is a floor. The tags run 30-day and 90-day click lookbacks, and the last 18 days of the window were still filling on 19/08/2026, so the same query run in November returns more.
  • From 01/06/2026 to 18/08/2026 the account paid $11.65 an enquiry against $9.53 for the year. That is the rate you would be buying today.
  • We kept the native tag side of the duplicate, 1,092. Keeping the analytics side instead gives 1,068 enquiries and $9.82. We took the integer-clean series, and the choice is worth about three per cent.
  • Two more lead actions sit enabled here with no volume in the window, one of them a second phone action. Promote that one and the phone lane doubles the way the form lane did.

What we left out, and why

  • There is $2,115 of conversion value in this account and we will not show it to you as money, because every enquiry in it carries exactly $1.00 by default.
  • No comparison reaching back before 19/08/2025, because this one form has been measured by as many as four conversion actions over the account's life.

The checks it passed

Double countConfirmed and dated. Identical daily rows to 29/07/2026, then exact doubling on eight separate days from 30/07/2026. The deduplicated count is what we publish.
Hardcoded valueConfirmed. We probed by day, by month and by action, and every conversion came back at exactly $1.00 with no spread, so no dollar figure is published for this client.
Methodology breakPartial. The action set held steady across the window and the biddable column did not, so every trend here is read off the native tag alone.
Engagement attributionNot established from platform data. The campaign name carries an agency prefix and the tracking was rebuilt in mid 2025, and neither of those is evidence, so nothing here is stated as something we caused. No agency writes that sentence, which is exactly why it is on this page.

Want to know what your enquiries really cost?

Send us one number from your last ad report. We will show you the query that proves it, or tell you that nothing does. No obligation either way.

Fridgy Australia. E-commerce, Shopify.

3.66 times back, and the cost of a sale cut from $41.19 to $16.43.

$50,460.33 of attributed purchase value from $13,780.77 of Google Ads, across 551 tracked purchases. Every dollar of it written by the Shopify checkout, not a default someone typed in.

3.66x
return on attributed purchase value
Google Ads attribution, not Shopify's totals. · Google Ads API, account id withheld, available on request

Then the pace took off: 271 tracked purchases in the 69 days after the change, against 280 in the 192 days before it.

Two audiences, one budget

Fridgy Australia sells nationwide from a Shopify store, with a mean tracked order of $91.59. Inside the same account sits a second audience, corporate buyers, with their own Search campaigns and their own enquiry form. They do not behave the same.

It looked like the account was failing

From March to May 2026 the return slid to 2.31, then 2.26, then 2.30. The obvious read was a product or Shopping problem. It was neither.

What we did

We read the account by lane instead of by month, and the picture flipped.

  1. Split it into lanes

    Cost and value per campaign per month showed the corporate lane taking 53.9 per cent of March's spend, 66.6 per cent of April's and 73.9 per cent of May's, and returning 0.31 times.

  2. Back the lane that was winning

    Underneath it the consumer lane was having its best run of the window: 4.82 times in March, 5.99 in April, 7.72 in May. We cut the corporate lane and moved the budget across.

  3. Keep the value honest

    Purchase value flows straight from the Shopify checkout into the tag. No defaults and no rounded stand-ins.

Why it worked

Monthly totals were hiding two different businesses averaging each other out. Once the budget followed the lane that was actually selling, the account carried its highest volume months of the window, and cost per tracked purchase fell from $41.19 in March to $16.43 in August.

The numbers

Where the budget went, and what each lane was returning

In March, April and May 2026 the corporate lane took most of the monthly budget while the consumer lane was posting the best returns of the window.

Where the budget went, and what each lane was returningIn March, April and May 2026 the corporate lane took most of the monthly budget while the consumer lane was posting the best returns of the window.Corporate lane, share of monthly spend0.00%37.50%75.00%0.00%0.00%Consumer lane, return on attributed purchase value0.00x5.00x10.00x6.16x4.92x2025-122026-022026-042026-062026-08corporate lane starts

June 2026 is a mixed month. Shopping and Performance Max both served on 11/06/2026, so it belongs to neither lane.

Show the numbers
monthCorporate lane, share of monthly spendConsumer lane, return on attributed purchase value
2025-120.00%6.16x
2026-010.00%2.93x
2026-020.00%4.23x
2026-0353.91%4.82x
2026-0466.62%5.99x
2026-0573.88%7.72x
2026-060.00%2.77x
2026-070.00%3.92x
2026-080.00%4.92x

Cost per tracked purchase, by month

Cost per tracked purchase fell from $41.19 in March 2026 to $16.43 in August 2026.

Cost per tracked purchase, by monthCost per tracked purchase fell from $41.19 in March 2026 to $16.43 in August 2026.$0.00$12.50$25.00$37.50$50.00Shopping to Performance Max, 11/06/20262025-122026-022026-042026-062026-08$16.49$16.43

August 2026 covers 01/08 to 18/08, 18 of 31 days.

Show the numbers
monthCost per tracked purchase, by month
2025-12$16.49
2026-01$28.54
2026-02$25.54
2026-03$41.19
2026-04$34.68
2026-05$38.06
2026-06$33.73
2026-07$23.45
2026-08$16.43

Account return across the changeover

Account return rose across the changeover, and the 69 days after it carried nearly as many tracked purchases as the 192 days before.

Account return across the changeoverAccount return rose across the changeover, and the 69 days after it carried nearly as many tracked purchases as the 192 days before.3.45x3.93x01/12/2025 to 10/06/2026,192 days, 280 trackedpurchases11/06/2026 to 18/08/2026,69 days, 271 trackedpurchases+13.9%

Account level, not a campaign comparison. At campaign level Performance Max returned 3.94 times against the 4.55 the Shopping campaign it replaced had been returning. What rose is the account multiple, because the corporate lane was gone. What moved is volume.

Show the numbers
PeriodValue
01/12/2025 to 10/06/2026, 192 days, 280 tracked purchases3.45x
11/06/2026 to 18/08/2026, 69 days, 271 tracked purchases3.93x

The result

$50,460.33 of attributed purchase value on $13,780.77 of media: 3.66 times back across 551 tracked purchases at $25.01 each. 280 purchases in the 192 days to 10/06/2026, then 271 in the 69 days after it.

VerifiedWhat it measures
3.66x
return on attributed purchase value
Media cost only, before management fee. Read after the 30-day attribution window closed.
The query behind it
01/12/2025 to 18/08/2026
SELECT metrics.cost_micros, metrics.conversions_value FROM customer WHERE segments.date BETWEEN '2025-12-01' AND '2026-08-18' -- $50,460.33 / $13,780.77
551
tracked purchases at $25.01 each
$13,780.77 divided by 550.95 purchases.
The query behind it
01/12/2025 to 18/08/2026
SELECT segments.conversion_action, segments.conversion_action_name, metrics.conversions, metrics.conversions_value FROM customer WHERE segments.date BETWEEN '2025-12-01' AND '2026-08-18' -- $13,780.77 divided by 550.95, all on the native purchase tag
$91.59
mean tracked order value, $39.83 to $229.80 by day
Measured at the tag, from the checkout total.
The query behind it
01/12/2025 to 18/08/2026 for the mean, July 2026 for the daily spread
SELECT segments.date, segments.conversion_action, metrics.conversions, metrics.conversions_value FROM customer WHERE segments.date BETWEEN '2026-07-01' AND '2026-07-31' AND segments.conversion_action = 'customers/{customer_id}/conversionActions/{native_purchase_action}' -- mean: $50,460.33 / 550.95 over 2025-12-01 to 2026-08-18
15,907
clicks from 1,592,444 impressions at $1.01 a click
Traffic only, and the one twelve-month figure here.
The query behind it
19/08/2025 to 18/08/2026
SELECT metrics.clicks, metrics.impressions, metrics.average_cpc FROM customer WHERE segments.date BETWEEN '2025-08-19' AND '2026-08-18'

How we checked it

Every number on this page was pulled live from the ad platform and checked a second time by a reviewer whose only job was to prove it wrong.

Open the method, the caveats and the checks

Every figure above was read live from Google Ads API on 01/10/2026, then independently re-derived by an independent second agent instructed to refute every claim, re-derived through the Google Ads client library. Figures that failed that pass are not on this page.

What to read alongside it

  • 3.66 times on the native tag. 1.99 times on the analytics import. Same store, same purchases, a factor of two apart. It is on our board as a dated job, not a footnote at the bottom of a page.
  • Figures are gross of returns, and we have not confirmed at the tag whether the passed value includes shipping and GST.
  • 3.66 times is the whole account, so the money we spent finding out the corporate lane did not work is still inside it. Take that lane out and the consumer lane reads higher.

What we left out, and why

  • No hand-picked start date, because moving this window forward one month takes 3.66 times to 3.34 and you should have both numbers rather than the better one.
  • No return figure across the full twelve months, because the counted conversion set changed twice inside them and one multiple would be averaging two different definitions.
  • No channel or traffic-source claim for this brand, because its analytics property has an unresolved cross-domain session-join defect and that is a live job, not a finished one.

The checks it passed

Double countTwo actions could have counted the same order. Only the native tag is switched on for bidding and reporting, so nothing in the published count is counted twice.
Hardcoded valueThis is the test the study exists for. The action forces no default value, and the recorded order value moves day to day, so the dollars are the store's and not ours.
Methodology breakFound, and it is why the return window opens on 01/12/2025 rather than on the same day as the traffic window.
Engagement attributionNot established. The account was serving before this window opened and Google Ads cannot evidence who caused what, so nothing here is written as a claim about what we caused.

Running a Shopify store on Google Ads?

Send us one of your own reports and we will tell you which numbers in it trace back to real orders and which do not. No obligation either way.

West Coast Health Physio. North Beach, Perth.

229 physio appointments in twelve months, at $44.33 each.

One clinic, one Search campaign, a fifteen kilometre radius, up against multi-site groups with far bigger budgets. Not a spike. A run rate.

$44.33
per booked appointment
Media cost, excluding GST and fees. · Google Ads API, account id withheld, available on request

Click-through rate nearly doubled, from 6.79 to 12.62 per cent, and has held there for nine months.

One clinic, one catchment

West Coast Health Physio runs one practice in North Beach, Perth, with a fifteen kilometre catchment. The competition is not the clinic next door. It is multi-site groups bidding on the same Perth searches with budgets a single practice cannot match. A physio does not sell a click. They sell a slot in a diary, and an empty slot at 2pm is gone.

You cannot outspend a group

A single clinic can only out-target the groups. In November 2025 the account was buying reach it could not use: 2,483 impressions, 182 clicks, and $104.97 to book one appointment.

What we did

The fix was subtraction. We cut the wasted reach, not the budget.

  1. One campaign, fifteen kilometres

    Everything runs through one always-on Search campaign covering the clinic's catchment. A small budget split across campaigns starves all of them.

  2. Match control, not more money

    From December 2025 the account stopped matching searches it could not convert. Impressions fell from 2,483 in November to 1,521 in December. Clicks held at 181.

  3. Count only real bookings

    A completed booking is one conversion and a call from the ad is another. Nothing else counts, so the bidding works for appointments, not page views.

Why it worked

Click-through rate is a relevance meter you cannot buy with budget. It stepped from 6.79 to 12.62 per cent in December 2025 and has held for nine months inside the same campaign: the same people, less of everybody else. Cost per booking then fell six months in a row, from $104.97 in November 2025 to $29.36 in May 2026, on flat spend of $779 to $907 a month.

The numbers

Click-through rate, before and after December 2025

One unchanged campaign nearly doubled its click-through rate and then held it for nine months.

Click-through rate, before and after December 2025One unchanged campaign nearly doubled its click-through rate and then held it for nine months.6.79%12.62%Aug to Nov 2025Dec 2025 to Aug 2026+85.9%
Show the numbers
PeriodValue
Aug to Nov 20256.79%
Dec 2025 to Aug 202612.62%

Impressions and clicks, month by month

Impressions were cut by more than a third and clicks did not follow them down.

Impressions and clicks, month by monthImpressions were cut by more than a third and clicks did not follow them down.Impressions served per month02,5005,0001,4431,023Clicks per month0150300811382025-082025-102025-122026-022026-042026-062026-08match control tightened

November 2025: 2,483 impressions, 182 clicks. December 2025: 1,521 impressions, 181 clicks. Both August columns are part months, 19 to 31 August 2025 and 1 to 18 August 2026, so neither is comparable to a full month on impressions or clicks.

Show the numbers
monthImpressions served per monthClicks per month
2025-081,44381
2025-092,682218
2025-104,843297
2025-112,483182
2025-121,521181
2026-011,750227
2026-021,899230
2026-031,919238
2026-041,909217
2026-052,092261
2026-061,920258
2026-071,751242
2026-081,023138

Cost per booked appointment, by month

Six consecutive months of falling cost per booking, on flat monthly spend.

Cost per booked appointment, by monthSix consecutive months of falling cost per booking, on flat monthly spend.$0.00$37.50$75.00$112.50$150.00$104.97, the peak2025-082025-102025-122026-022026-042026-062026-08$29.35$49.19

Dashed months are provisional. The left endpoint is August 2025, which covers 19 to 31 August only, so its $29.35 is thirteen days and not a full month.

Show the numbers
monthCost per booked appointment, by month
2025-08$29.35
2025-09$41.82
2025-10$31.15
2025-11$104.97
2025-12$91.66
2026-01$59.52
2026-02$54.03
2026-03$44.91
2026-04$33.90
2026-05$29.36
2026-06$36.05
2026-07$51.96
2026-08$49.19

The result

229 booked appointments across twelve months at $44.33 each, on $10,152.24 of media, for one clinic in one catchment.

VerifiedWhat it measures
229
booked appointments, and that is the floor
Still filling on a 90 day lookback. Data-driven attribution splits credit across touchpoints, so the monthly totals underneath run fractional.
The query behind it
19/08/2025 to 18/08/2026
SELECT segments.conversion_action_name, metrics.conversions, metrics.all_conversions FROM campaign WHERE segments.date BETWEEN '2025-08-19' AND '2026-08-18'
$44.33
cost per booked appointment, on total account spend
All spend charged against bookings alone, the strict basis.
The query behind it
19/08/2025 to 18/08/2026
SELECT metrics.cost_micros FROM customer WHERE segments.date BETWEEN '2025-08-19' AND '2026-08-18' -- $10,152.24 divided by the 229 bookings
12.62%
click-through rate since December 2025, against 6.79 per cent before it
One unchanged Search campaign on both sides of the break.
The query behind it
01/12/2025 to 18/08/2026
SELECT metrics.ctr, metrics.clicks, metrics.impressions FROM customer WHERE segments.date BETWEEN '2025-12-01' AND '2026-08-18'
$3.67
average cost per click across $10,152.24 of media
Up from $3.41 before the break.
The query behind it
19/08/2025 to 18/08/2026
SELECT metrics.average_cpc, metrics.cost_micros, metrics.clicks FROM customer WHERE segments.date BETWEEN '2025-08-19' AND '2026-08-18'

How we checked it

Every number on this page was pulled live from the ad platform and checked a second time by a reviewer whose only job was to prove it wrong.

Open the method, the caveats and the checks

Every figure above was read live from Google Ads API on 19/08/2026, then independently re-derived by an independent second agent instructed to refute every claim. Figures that failed that pass are not on this page.

What to read alongside it

  • This account has been running since early 2022 under at least two earlier operators, and their fingerprints are still on the campaign names. We are not claiming we built it from nothing.
  • The account's own conversion column says 240, and 456 once Google's engagement proxies are counted. Only 229 are appointments. We publish the smallest of the three.

What we left out, and why

  • There is $23,400 of conversion value sitting in this account and we will not show it to you, because we typed those dollars in ourselves, $100 on a booking and $90 on a call, to steer the bidding.
  • We publish no before and after on cost per booking, because $43.79 before the break and $44.53 after it is the same number twice and calling that an improvement would be a lie.

The checks it passed

Double countWe went looking for a second action counting the same booking. There is not one. One conversion action per real event, with no analytics import on top.
Hardcoded valueFound on the first pass, on both conversion actions. That single finding decides what this study is allowed to publish, and the refusals above are its consequence.
Methodology breakFound one. Two Google engagement actions were created mid window, so no trend is claimed on total conversions. Bookings are read on their own action.
Engagement attributionNot established from platform data. This is the test we failed, and the wording throughout the page reflects it.

Running a clinic on Google Ads?

Send us your account and we will show you where the budget is going and what we would cut. It takes about an hour and you keep what we find either way. No obligation.

Enkindle Group logo
SEOAI SEOGoogle AdsConversion TrackingMeta AdsTikTok AdsLinkedIn AdsEmail MarketingSMS MarketingB2B Lead GenerationCustom CRMAI AutomationWeb Design & DevelopmentBrandingWeb HostingGraphic DesignVideo ProductionMeta Ads ConsultingGoogle Ads ConsultingSEO Consulting
InstagramTikTokFacebook
Klaviyo K:Partners Bronze partner badge
Enkindle Group
2026