Round one creative, the funnel architecture, the written copy and the weekly operating loop. Every ad has its video attached, with the full transcript underneath it.
Play each video, then read the copy written for it. The transcript sits behind the toggle under every ad so you can check the voiceover against the primary text without leaving the page. Videos are compressed previews for review; the campaign uses the full-resolution masters in the drive.
A renovation is a considered purchase made over months, not a decision taken in a feed at nine in the evening. Asking a cold stranger to book an hour at a showroom is asking too much too early, and it is why cold-only lead campaigns in this category produce expensive, unqualified enquiries.
So three jobs run at once. Cold ads make people problem aware and build the warm audience out of whoever watches them. Warm ads build trust and buying criteria, so that when the offer finally arrives the person already believes in the criteria only Arch KBB meets. Hot ads convert, and only those are judged on cost per booking.
In week one there is no warm audience, so there is nothing to retarget. Cold and conversion therefore run in parallel from day one while the pools fill. Stages two and three switch on when the audience gates are met, usually two to four weeks in. Do not judge the retargeting stages before they have an audience to retarget.
| Job | Make people recognise the problem, and build the video-viewer audience that stages 2 and 3 depend on |
| Objective | Video views, optimised for ThruPlay. Not leads |
| Audience | 5 mile radius of the W6 showroom, plus Chiswick, Ealing, Putney, Richmond, Twickenham and Wimbledon. Age 35 to 60. Broad, no interest stacking. The creative does the qualifying |
| Creative | Ad 1 and Ad 2 below |
| Judged on | Cost per ThruPlay, hook rate, hold rate, pool growth. Never on leads |
| Job | Set the buying criteria, and make the in-house structure feel like the only sane way to do this |
| Objective | Traffic to the landing page, or ThruPlay while pools are small |
| Audience | Watched 50%+ of any stage 1 video in 90 days, plus Instagram and Facebook engagers in 180 days |
| Creative | Ad 3 and Ad 4 below, plus Explainer 01 recut once the years claim is settled |
| Judged on | Link CTR, cost per landing page view, and growth of the stage 3 pool |
| Job | Book The Arch Blueprint |
| Objective | Leads for three weeks, then Consultation Attended as an offline conversion from HubSpot |
| Audience | Landing page visitors who did not book (30 days), form abandoners, 75%+ viewers of stage 2, plus the customer list for lookalike seeding once volume allows |
| Creative | Ad 5 below, plus the six existing project carousels as static proof |
| Judged on | Cost per booked Blueprint under £150, then cost per attended consultation under £400 |
The percentages above were written against an illustrative £150 a day. The agreed opening budget is £500 a month, which is about £16 a day, and that changes how this has to be built.
Split three ways, that is roughly £6.60 cold, £4.10 warm and £5.75 hot per day. Meta will not exit the learning phase on any of them at that level, and three underfed ad sets compete with each other for the same audience while none of them gathers enough data to optimise. So month one runs as two ad sets, not three: cold ThruPlay at roughly £6 a day to build the viewer pool, and conversion at roughly £10 a day pointed at the landing page. Stage 2 opens only when the pool justifies it and the budget rises.
What £500 buys, honestly: enough data to prove the funnel and produce a handful of consultations, not a full three-stage system running at once. It is the right way to start, and the numbers in Part Four are how we earn the case for increasing it.
Meta truncates primary text at roughly 125 characters on mobile, so every one of these front-loads the hook into the first line. Headlines are written to sit under the video without repeating what the voiceover just said. No prices, no discounts, no deadlines and no budget figures of any kind, per the messaging platform.
The strongest single piece of direction from the confirmation call, and it applies to all five ads below. An ad has no cover image. Whatever the first frame of the video is, is the ad, and every one of these opens on a presenter standing against a background that does not say kitchen, bathroom or renovation to anyone scrolling. In his words: "if I look at it without playing it, there isn't a picture behind her that draws us to a kitchen or to a bathroom."
He is right, and it is the cheapest available improvement to this whole system. The rule for every future shoot: open on the room, not on the person. Establish the space wide enough that it reads instantly as a finished kitchen or bathroom, then push in to the presenter. He named two competitors doing exactly this, Smallbone and The Woodworks, both of which open on the work rather than on a face.
Round one still runs the footage that exists, because Meta decides sequencing and a talking head can earn its place if the copy is right. But the hook rate threshold in Part Four is the test of it, and the brief goes to the content team before anything else is filmed.
The source already covers listening, ideas, materials, then budget and timeline, so the structure is right. Trim to 45 seconds, crop to 1080×1920, and rewrite the voiceover to name The Arch Blueprint and describe the two meetings honestly: an hour of discovery, then a presentation of the design and the costing. Keep the "free, relaxed, no obligation" line, it is doing real work, and close on seeing the whole thing before committing to anything. No £750 valuation and no take-away claim, both withdrawn on 1 September. Per Munty's hook note above, the recut must open on a finished room rather than on the presenter.
This is the "meet your designer" video the landing page already calls for. It thanks the client for booking and tells them exactly what to bring: inspiration images from Pinterest or Instagram, their must-haves and wish list, floor plans and rough measurements.
Drop it straight into the booking confirmation email and reference it in the SMS reminder. It costs nothing, it is already made, and it protects the show-up rate, which is the metric the entire system is ultimately judged on. Nothing else in this document is as cheap to action.
A strong trust film: one-stop shop, collaboration with your architect, one point of contact, design-led rather than cookie-cutter, and a one-year workmanship guarantee. Usable in stage 2 once recut vertically and trimmed to about 45 seconds.
"For over 18 years." Companies House shows The Arch Bathrooms Limited incorporated in 2014, which is twelve years, and the landing page says "trading since 2014". These cannot both run. If the eighteen years refers to the founders' experience rather than the company, the line has to say so.
"One-year workmanship guarantee." This appears nowhere in the messaging platform or on the landing page. If it is current it belongs in both, and in the risk reversal on the booking page.
All three February videos run the same discount campaign: free dry fitting "worth up to £3,900" on kitchens and bedrooms, free underfloor heating or a heated towel rail on bathrooms.
They contradict the messaging platform on its most deliberate point, that no prices and no discounts appear in advertising, because both recruit the comparison shopper who does not close at 20%. They are also, on the record, what was running when the account went dark in late March. Re-running them rebuilds exactly the audience this project has been built to filter out.
If a promotional angle is wanted later it has to come from somewhere other than price. The £750 design value that used to sit inside The Arch Blueprint was withdrawn on 1 September, so there is currently no sanctioned number of any kind in Arch KBB advertising. See section 12b of the messaging platform for how the budget filter is done instead.
An ad system that is set live and reviewed monthly is a spending plan, not a system. The loop below is what actually moves people along the journey, because it does three jobs every week: it kills what is not working, it promotes what is, and it feeds the next stage with audience.
| Day | Action |
|---|---|
| Monday | Pull the numbers for all three stages. Score every ad against the thresholds below. Nothing is judged before 8,000 impressions |
| Monday | Apply the decision rules. Pause, recut or promote. One change per ad per week, so cause and effect stay readable |
| Wednesday | Check audience pool sizes and open the next stage if a gate has been passed |
| Friday | Reconcile HubSpot. Booked, attended, won. This is the only number that finally matters |
| Monthly | Refresh creative that has hit frequency, and brief the next shoot from the gap list |
Deliberately mechanical, so decisions are made on data rather than on how anyone feels about a piece of creative on a Monday morning.
| Stage 1, cold | |
|---|---|
| Hook rate 3s views ÷ impressions | Above 25% is healthy. Below 20% after 8,000 impressions the first three seconds are wrong. Recut the opening and change nothing else |
| Hold rate ThruPlay ÷ 3s views | Above 20% is healthy. Below 12% the middle sags. Cut fifteen seconds out of it |
| Cost per ThruPlay | Above £0.10, pause and shift budget to the better performer |
| Frequency | Above 2.5 in seven days, rotate even a winner |
| Stage 2, warm | |
|---|---|
| Link CTR | Above 1.0%. Below 0.6% after 8,000 impressions the headline is not matching the video. Test the variant |
| Cost per landing page view | Under £0.60. Above £1.20 the audience is too narrow or too tired |
| Pool contribution | If the stage 3 pool is not growing week on week, stage 2 is not doing its job regardless of CTR |
| Stage 3, hot | |
|---|---|
| Cost per booked Blueprint | Under £150 for weeks one to three. At £500 a month that is three to four bookings, so judge the rate, not the volume |
| Cost per attended | Under £400 once HubSpot feeds attendance back. This is the real target |
| Show-up rate | Below 65%, the problem is the confirmation sequence, not the ads. Deploy Explainer 03 and the SMS reminders before touching the campaign |
| Lead quality | If more than 40% of leads say "still researching", tighten stage 1 creative rather than adding form friction |
| The audience gates | |
|---|---|
| Open stage 2 | When the 50% video-viewer pool passes 1,000 people |
| Open stage 3 retargeting | When the combined engaged pool passes 3,000 people |
| Open lookalikes | When 100 attended consultations or 50 won projects have been uploaded from HubSpot. Seed from customers, never from leads |
Any stage 1 ad that holds above 25% and produces cheap ThruPlays gets duplicated into stage 2 against the warm audience. Winners are promoted up the funnel rather than simply left running. This is the mechanism that keeps the system compounding instead of decaying.
What the next shoot needs to produce, in priority order. Regenerated monthly from whatever the loop shows is missing.