This is the format you would receive for your account. To show it with real depth I used a fictional brand: ORVO, an online shop for electric standing desks in Germany, Austria and the Benelux, spending about 21,000 € a month on Meta and Google.
Fictional business and illustrative numbers. No client data used.Nine parts, in the order they are used. The idea behind all of them: let proof and real outcomes decide, not opinions, and spend media budget only on creative that has already won for free.
Last 30 days, read only access. Facts first, then what they cost, then the fix. Ranked by impact, so your team knows what to do on Monday.
Before writing a single new idea, I look for ads and reels that have already proven themselves, from direct competitors and from other markets worldwide. We copy their structure, never their footage, music or claims.
"Put a coffee on it. Now press up."
Structure reused for ORVOConcept B, the wobble test. Same proof moment, our own glass, our own measured 2 mm.
"My office is 4 m². Watch."
Structure reused for ORVOConcept C, fits your flat. Shot in a real small flat, the floor plan replaces the tour.
"Standing all day is wrong too."
Structure reused for ORVOCopy variant 07, the myth. Pre tested as a trial reel before it becomes an ad.
Reference ads on this page are invented for the sample. For your account the cards link to the real ads and posts, with screenshots as proof.
Fewer, bigger units that can learn. Advantage+ gets the reach, but inside clear limits, so it cannot spend the budget on existing customers or junk placements.
Meta · 420 € a day
Google · 280 € a day
With broad or Advantage+ audiences, the video itself decides who sees it. Meta shows it to the people who watch it longest, so the words and pictures in the ad do the job interests used to do. Every concept carries the same signal words in three places.
home officesmall flatback to back callsshared desk
work from home120 × 60 cmone touchafternoon calls
laptop on a deskfloor planvideo calltwo saved heights
Describe the situation: "Sat for 9 hours today?"
Tell a third person story: "Mara works from a 38 m² flat."
Describe the product for a use: "A desk for small home offices."
Implying a health condition: "Is your back hurting again?"
Health promises: "Fixes your posture."
Sensitive attributes in audience names or targeting, which Meta blocks anyway
I read the public reviews per market: Google Maps for every showroom and competitor store, plus Trustpilot and local review sites. They show how far people travel, who buys, what makes them hesitate and the words they use themselves. That sets radius, age range, the objection to answer first and the wording per market.
Berlin showroom, 186 reviews read, buyers come from as far as Potsdam
Vienna showroom, 94 reviews read, buyers from all of Lower Austria
Amsterdam showroom, 141 reviews read, buyers from most of the Randstad
No showroom, 63 online reviews read, delivery only
Examples invented for the sample. Customer wording is used as inspiration only; real customer quotes go into an ad only with consent.
Same video, same claims. Only the words, one local reference and the voice change. A voice from the market sounds like a neighbour; a heavy dialect sounds like a parody and costs trust.
| Market | Standard version | Regional version | Voice |
|---|---|---|---|
| Germany | "Dein Schreibtisch fürs Homeoffice." | Same line, Berlin reference: "vom Altbau bis zur WG" | Standard German, warm |
| Austria | Wording from Germany | Austrian standard: "Sessel" for chair, "heuer", Vienna or Graz named | Austrian speaker, standard with a light colour |
| Netherlands | Dutch translation | Native rewrite: "thuiswerkplek", informal "je" | Dutch speaker from the Randstad |
| Belgium | Dutch from the Netherlands | Flemish wording and voice in Flanders, French in Wallonia | Flemish speaker, French speaker |
Each variant opens with a different hook angle, so the test tells us which reason to buy works best, not just which sentence. Every line is written to pass Meta's personal attributes policy and Google's editorial rules: no claims about the viewer's body or health, numbers only where they can be proven.
Sat for 9 hours today?
Your desk can do the standing part. One touch, four saved heights, and you switch whenever your day allows it. Electric standing desk from 449 €, 30 night home trial.
Headline: Sit less without thinking about it
Why this hook: A concrete number stops the scroll better than a benefit. It describes a day, not a body, so it stays clear of Meta's personal attributes rules. Pairs with Concept A.
We filled a glass to the brim and typed at full height.
Not a drop. Two motors and a steel crossbar keep the top steady at 125 cm, where most desks start to wobble. See the full test.
Headline: The steady standing desk
Why this hook: Proof beats adjectives. "Stable" is a claim everybody makes, a full glass is something people believe. Pairs with Concept B.
Small flat? Measure first.
ORVO Compact is 120 × 60 cm. That is the footprint of an armchair. Check our size guide, then decide. Free delivery across the EU.
Headline: Fits where a chair fits
Why this hook: Space is the most common reason people do not buy (support tickets, comments). Answering it in the first line filters in the right buyers. Pairs with Concept C.
One box. 15 minutes. One Allen key.
That is the whole setup. The frame comes pre assembled, the top clicks on, and the cable tray is already in place.
Headline: Set up before your coffee cools
Why this hook: Speaks to the fear of a weekend lost to assembly. Strong in retargeting, where people already like the product.
Try it at home for 30 nights.
If it does not fit your day, we pick it up and refund you in full. No questions, no restocking fee.
Headline: 30 nights, free return pickup
Why this hook: Moves the risk from the buyer to the brand. Usually lifts conversion rate more than a discount and does not erode the price.
Mara works from a 38 m² flat in Utrecht.
At 2 pm her desk rises, the laptop stays where it is, and the afternoon calls happen standing up. At 6 it goes back down and becomes the dinner table corner again.
Headline: A desk for small home offices
Why this hook: A story about someone else is relatable without assuming anything about the viewer. Specific details (38 m², 2 pm) make it feel real.
Standing all day is not the goal. Switching is.
Most people who buy a standing desk stop standing after two weeks because they overdo it. Start with 20 minutes after lunch. Four saved heights make the switch one touch.
Headline: Sit, stand, switch
Why this hook: Education angle that sounds like advice, not an ad. Also lowers the bar for people who think standing desks are for fitness fans.
Single motor desks wobble at full height. Ours does not.
Two motors lift evenly, a steel crossbar stops the sway. Measured at 125 cm: 2 mm of movement while typing.
Headline: Dual motor from 449 €
Why this hook: For people already comparing models. A measurable spec turns "better quality" into a reason to pay a bit more than the cheapest option.
Less than two months of coworking.
An electric standing desk from 449 € that you use every working day for years. Frame warranty 10 years.
Headline: From 449 €, 10 year frame warranty
Why this hook: Reframes the price against a cost home workers already know. Works best for audiences who have searched or visited before.
Your height, your partner's height, and two more.
Four saved heights, so the desk is right for everyone who shares it. Press once and it moves to the stored height, then stops by itself.
Headline: One desk for the whole household
Why this hook: Opens a second buyer: couples who share one home office. A new angle for retargeting after the first message has been seen.
Character counts checked against the 30 and 90 character limits.
Each concept comes with an animated mock up, so your team sees timing and hierarchy before anyone films. The mock ups are made without stock people; for the real ad you would swap the drawn desk for product footage and keep the timing.
Problem in one number, then the mechanism. The hook asks a question every desk worker can answer, the product answers it without a single benefit word.
Proof demo. Stability is the main reason people pay more than the cheapest desk, so the ad shows it instead of saying it. Dark look to stand out in a light feed.
Objection killer for the biggest reason not to buy. A floor plan makes 120 × 60 cm tangible, the one box beat answers the setup fear in the same ad.
Before a concept gets media budget, it competes against itself for free. Instagram trial reels are shown only to people who do not follow the account, so they behave like cold traffic. After 72 hours the numbers pick the winner, and only winners become ads.
| Variant | Hook, Concept A | Hook rate | Avg. watch | Shares / 1k | Saves / 1k | Verdict |
|---|---|---|---|---|---|---|
| V1 | "Sat for 9 hours today?" | 41 % | 4.2 s | 3.1 | 5.4 | Keep as backup |
| V2 | "Your desk can do the standing part." | 36 % | 3.8 s | 1.9 | 3.2 | Drop |
| V3 | "9 hours. 0 minutes standing." | 54 % | 6.9 s | 6.8 | 11.2 | Run as ad |
| V4 | "One touch. Four heights." | 33 % | 3.5 s | 1.2 | 2.7 | Drop |
| V5 | "Mara stands up at 2 pm." | 47 % | 5.6 s | 4.4 | 8.9 | Retest as story ad |
Sample numbers. Hook rate on Instagram = 1 minus skip rate, the share of views that stay past the first 3 seconds. Rule: highest hook rate wins if its watch time is above the median of the five; ties go to saves, which signal real interest.
The test runs as an automated routine on the official Instagram API, so your team does not cut or post variants by hand.
Next to online sales, ORVO books test sessions in its showrooms. Those leads only pay off when someone visits and buys, so Meta should learn from visits and purchases, not from cheap form fills.
Tracking is fixed first, otherwise every result below would be judged on wrong numbers. Then free organic tests before paid ones, one variable at a time, with the decision rule written down before the money is spent.
| Days | What runs | Why | Judge by |
|---|---|---|---|
| Day 0 | Event ID deduplication, purchase only primary conversion, Enhanced Conversions, customer lists. Instant form, CRM statuses and Conversions API live | Clean signal before anything is judged | Platform purchases within 10 % of shop orders |
| Days 1 to 3 | Trial reels: Concepts A, B, C in 5 variants each, no budget. New Meta structure live, old campaigns paused. Google brand exclusions and negatives live | Finds the strongest hook for free, gets ad sets out of learning | Hook rate, watch time, shares, saves. No CPA judgement yet |
| Days 4 to 7 | Trial winners as ads in the testing campaign at 30 € per concept per day. Regional A/B ads per market at 10 € each | Only proven openings get paid reach | Link CTR, cost per add to cart, cost per lead |
| Days 4 to 14 | Google RSAs per intent group, Performance Max with new feed titles | Message match with the search lifts CTR and conversion rate | CTR, conversion rate, cost per new customer |
| Days 8 to 14 | Best ad per concept against the current winner, retargeting with copy 4, 5, 9, 10. Instagram winners go to TikTok as Spark Ads | Turns attention into sales, tests risk reversal against discount | New customer CPA, cost per showroom visit |
| Day 14 | Review call: graduate winners into Advantage+, write the next concepts from the next research round | Keeps creative fresh before fatigue | All KPIs below vs baseline |
Short answers are enough. They make sure every claim in the ads is true and the tests can start on day 1.
Delivered within 5 working days after read only access: audit and research on day 2, copy, concepts and market wording on day 4, pre test plan, lead loop, test plan and walkthrough on day 5. Strategy only, your team implements, exactly as in your brief.
I use AI to draft many copy variants and animate the concept mock ups, which is why you get more options to test for this budget. Every line and every number is checked by me before it reaches you.