The account is not underperforming because of its creative. It is spending in the wrong places and sending Google the wrong signals — and both are structural, which means both are fixable in days rather than quarters.
Bibado ranks strongly on its own name. Paying for the majority of those clicks means buying the same visit twice.
Four numbers describe the whole account. Three of them are problems with a fix measured in minutes or hours; the fourth is the campaign those fixes should be funding.
Moving brand from 61% to a 10% ceiling frees roughly half the monthly budget without asking for a penny more. That is the largest single lever in the account, and it funds everything else on this page.
Read the four campaigns side by side and the correction more or less writes itself — money is sitting where it is least needed and absent where it is working hardest.
| Campaign | What we see | State |
|---|---|---|
| Brand | 61% of total budget, against a strong organic presence on the same terms | Over-funded |
| NB Search | £397 spent on search terms; only £35 of it on terms that converted | Leaking |
| PMAX Coveralls | 61 new customers at £16.73 on just £40 a day | Starved |
| BN PMAX Bundles | Only two products generating; the rest never get impressions | Concentrated |
Some leakage is normal in non-brand search. At 91% it is not leakage — it is the campaign’s default state.
Sixteen search terms converted, and they account for £35 of the spend. Everything else went to queries that brought no order at all. The fix is a negative keyword pass, then a weekly routine so the list does not refill.
Meanwhile PMAX Coveralls returned 61 new customers at £16.73 each on £40 a day. That is a campaign limited by its budget rather than its performance — and the natural home for money recovered from brand.
Smart bidding is only as good as the instructions it is given. Two of those instructions are currently wrong, and neither costs anything to put right.
An Android app install action and one recording zero conversions are both set as primary, so both are pulling the algorithm away from revenue.
Keep one revenue action primary. Move the Android install and the zero-conversion action to secondary, or remove them. Ten minutes of work with an outsized effect on what the algorithm chases.
Audience signals describe the ideal customer. Four are ineligible, so most of that guidance never reaches Google — and the relevant in-market audiences are not in use at all.
Fix or rebuild the four, then layer in the in-market audiences that match a weaning customer. There is a long list the account has never touched.
Everything beyond a single primary action is noise in the signal Google optimises against. These two fixes land before any budget moves, so the account is already pointed at the right outcome when the money follows.
The top sellers over the last thirty days were the Happy Hares and Arctic Amigos long-sleeve coveralls. Because they perform, the algorithm keeps giving them the spend.
That is correct behaviour — and also the reason no other SKU gets a chance to prove itself. A campaign that has found two winners stops looking for a third.
Move those two into their own feed-only asset group inside the same campaign, so the rest of the range has room to generate revenue. The same applies to BN PMAX Bundles, where only two products are performing. Against this much retailer competition on generic terms, this is also how the revenue stays with Bibado rather than going to a reseller.
It currently points at a single product page. The coveralls collection page would give more choice and put the bundles in front of the customer, which is where the higher order values sit.
Test the collection page as the destination — but after the asset group split, so the two changes can be read separately.
One long headline is in breach and restricting exposure. It was disapproved under what Google calls a past policy — the account likely triggered the same rule repeatedly, the rule has since been retired or rewritten, and the flag stayed.
Replace the headline, or appeal on the basis the policy no longer exists in that form. Replacing is faster; appealing is cleaner if the flag suppresses more than this one asset.
On generic terms Bibado is bidding against resellers carrying the same product. In Shopping, the first words of the title and the clarity of the image are most of the battle.
Titles should lead with what customers search, then colour and size, with the brand last. Bibado’s currently open with the brand and then repeat variations of the same phrase — three ways of saying bib, where the words could be describing the product instead.
Rewrite the feed titles to that pattern across the range.
Some products use a standard ecommerce shot — best practice, and generally the stronger performer in Shopping. Others use a baby model shot, which is a good image and works well on the website, but Shopping is a fast-scanning surface where the product needs to read instantly.
Both images are from Bibado’s own photography and are shown to illustrate the two styles. They are not pulled from the live Shopping feed.
Standardise the feed on ecommerce shots. The lifestyle photography keeps its place on site, where the scene is the point rather than the obstacle.
Ordered by effect against effort. The first three cost nothing but attention, and they change what the account optimises towards before any budget moves.
| Change | Why | Effort | |
|---|---|---|---|
| 01 | Fix the conversion actions | One primary action tied to revenue. Everything downstream optimises better for it. | Minutes |
| 02 | Rebalance brand spend | From 61% toward a 10% ceiling, with the recovered budget moved to non-brand and PMAX. | Hours |
| 03 | Clear the disapproved headline | Replace or appeal, so the campaign stops running with its reach restricted. | Minutes |
| 04 | Cut the non-converting search terms | £362 recoverable — then a weekly negative keyword routine so it does not come back. | Hours, then weekly |
| 05 | Repair the audience signals | Four ineligible lists fixed or rebuilt, then in-market audiences layered in. | Hours |
| 06 | Split the PMAX asset groups | Top sellers into their own feed-only group, so the rest of the range gets a chance. | Hours |
| 07 | Rewrite the feed titles | Keyword first, brand last, across the range. | Days |
| 08 | Standardise the feed images | Ecommerce shots throughout the Shopping feed. | Days |
| 09 | Test the collection page in PMAX | Run it after the asset group split, so the two can be read separately. | Set in minutes, read in weeks |
Nothing above asks for additional budget. The first six changes are paid for by the spend already going to brand terms and non-converting search queries.