Merchant Center & Feed Rebuild
Disapprovals cleared at source, GTIN, MPN and attribute gaps closed, and titles rewritten to lead with searched terms. Custom labels for margin, stock cover and seasonality.
San Francisco, CA
Google Ads Management in San Francisco, CA
Google Ads here is about capturing the demand that already exists — and refusing to pay Bay Area prices for demand you would have got free.
Delivered remotely for brands across San Francisco and California.
Grow · San Francisco
Why San Francisco brands come to us for this
- Merchant Center rebuilt first: disapprovals cleared, GTINs and attributes closed, titles led by searched terms
- Branded search isolated so non-brand performance is finally measurable on its own
- Custom labels by margin, stock cover and seasonality driving Shopping and PMax asset groups
- Alcohol certification and supplement claim disapprovals handled at the catalogue level
- Weekly query mining and negatives, because a wasted click costs more in this auction
Somebody in this metro has already decided they want a pour-over grinder, a magnesium supplement or a Sonoma cabernet; the account's only job is to be there when they type it. That framing matters more here than elsewhere because auction prices in this DMA are set by advertisers with enormous tolerance for loss, so any dollar spent on a query that was not going to convert is a dollar lost at a premium rate. The work is therefore mostly subtraction: negatives, query mining, and structure that keeps cheap demand from subsidising expensive guessing.
The single biggest lever in almost every account we inherit is the Merchant Center feed, and it is the least glamorous. Titles that lead with the brand name instead of the searched term, missing GTINs, unpopulated attributes, no custom labels for margin or stock cover, and a batch of silent disapprovals nobody has opened in months. Shopping and Performance Max spend your budget according to that feed, which means feed work is bid strategy whether or not anyone calls it that. For pre-order hardware, the availability attribute has to be set correctly or you are advertising a product Google believes is in stock and shipping tomorrow.
Then there is brand. A well-known Bay Area brand often finds that its impressive blended ROAS is mostly people searching its own name, which the store would have captured anyway. We split branded demand into its own campaign with its own budget and target, so non-brand performance becomes visible for the first time — and that number is usually the real conversation. Sometimes the honest recommendation is to reduce brand spend and stop counting it as growth.
Local context
Category approvals and feed compliance that this city's catalogues actually trigger
The product mix San Francisco brands sell runs headfirst into Google's restricted categories more often than most. Alcohol advertising requires certification, restricts where and to whom it can serve, and Shopping for wine has its own eligibility rules by destination — worth solving before a holiday allocation release rather than during it. Supplements and skincare draw disapprovals for health claims lifted straight from the product description, so the fix usually lives in the catalogue rather than the ad copy. Connected devices with a battery or a cord trigger their own attribute and safety requirements. On top of that, California privacy rules mean a meaningful share of your visitors opt out of sale and sharing, which affects remarketing list sizes and enhanced conversions match rates; we set consent mode correctly and read the resulting numbers as reality rather than as a tracking bug to be worked around. Geographic targeting also deserves thought here, because a nine-county Bay Area radius and a national campaign behave nothing alike and should never share a budget.
Scope
What Google Ads includes
The same standard of work we run for every client — applied to a San Francisco brand’s realities.
Full service detailBrand, Non-Brand & Conquesting
Branded demand isolated into its own campaign, budget and target so non-brand performance becomes visible. Competitor conquesting runs as a separate line, judged separately.
Shopping & Performance Max Structure
Asset groups split by margin band and product type instead of one catch-all, with listing-group bids, product exclusions and brand-term controls applied wherever PMax still allows them.
Query Mining & Negative Hygiene
A weekly pass over search terms and PMax category reports, with a maintained shared negative library so budget stops leaking into research, DIY and job-seeker queries.
Promotions & Merchant Programs
Merchant promotions, sale price annotations, shipping and returns policy setup, product ratings, and local inventory ads where you have stores worth feeding.
tROAS Bidding to Contribution Margin
Targets set from margin per product group rather than a platform default, moved in controlled increments, with a written reason attached to every bid and budget change.
Scoped and quoted for your San Francisco store
We do not work off a rate card. Every San Francisco engagement starts with a fixed statement of work — named deliverables, named dates, one number — written after we have looked at your store, not before. If a smaller first step would serve you better, we will say so.
Get this scopedHow it runs
From kickoff to results
Feed & Account Audit
Merchant Center diagnostics, attribute coverage, campaign overlap and wasted spend scored against ninety days of search terms. You get the findings whether you hire us or not.
Fix the Feed
Titles, attributes, product types and custom labels rebuilt before any campaign work. A perfectly structured account on a bad feed still shows the wrong products to the wrong queries.
Restructure
Brand, non-brand, Shopping, PMax and Search rebuilt with hard budget boundaries and a shared negative library, so each line answers a different commercial question.
Bid to Margin
tROAS targets derived from contribution margin by product group and applied gradually, so the account keeps its learning instead of resetting it every Monday.
Mine & Expand
Weekly query mining, monthly feed reviews, then expansion into the categories the search data says you can profitably win. Nothing scales before the query set is clean.
Proof
Google Ads results
Anonymised under NDA. Figures pulled from the client’s own analytics.
Performance Apparel (DTC)
~$6M/yr DTC, 900+ SKUs across size and colour variants, US · Shopify Plus
Returns ran at 31% and refund cost consumed the entire paid media margin. One size chart image served 40 different fits, and 62% of add-to-carts started on a collection page that never showed variant availability. The named constraint: no new product photography budget, so every fix had to come out of the existing asset library and the review corpus.
+29%sitewide conversion rate1.71% to 2.21%, five-month average
+14%average order value$84 to $96 once the threshold bar and cross-sell shipped
31% → 22%return ratenine-point drop, roughly $310k/yr in recovered margin
1.9x → 2.4xblended MERwith paid spend held flat throughout
Consumer Electronics & Accessories
~$9M/yr, 210 SKUs, US + AU · Shopify Plus (migrated from BigCommerce)
Meta ROAS had slid from 3.6x to 1.9x in a year and the team had spent twelve months buying new creative to fix it. The real cause was measurement: the BigCommerce checkout dropped 22% of purchase events and the Conversions API had never been installed, so both ad platforms were optimising on incomplete data. The named constraint: peak season was 14 weeks out, and the replatform had to be live and stable well before Black Friday traffic arrived.
1.9x → 3.4xMeta ROAS, once the 22% event gap closed60 days after server-side tracking went live, spend up 18%. Most of that is signal we recovered, not performance we invented — the honest number is the blended CAC below, which is measured against Shopify orders
-32%customer acquisition cost$44 to $30 blended across Meta and Google
4.1s → 1.7smobile LCPdesktop went 2.9s to 1.2s over the same window
+47%peak-season revenueBlack Friday through Cyber Monday, year over year
In their words
Clients on this work
Non-brand search revenue ~2x at flat spend
“Six thousand products and a Shopping feed nobody had touched since it was first generated — a third of it was disapproved and we had no idea. They rebuilt the feed off our real product data, fixed the GTIN and size attributes, and split brand off from non-brand so I could finally see what we were actually paying to acquire. They also cut the broad 'baby clothes' terms that were eating a quarter of the budget on people who were nowhere near buying. Spend is roughly flat and non-brand search revenue has close to doubled.”
Director of OperationsKids & baby brand, ~$5M/yr · Atlanta, GA
Verified via Shopify Partner referralFAQ
Google Ads in San Francisco — your questions
More than most brands assume, once certification and destination eligibility are in place. Alcohol advertising requires approval, serves under age-related restrictions, and has Shopping eligibility that varies by market, so the setup work is front-loaded and worth doing outside your holiday peak. Search for varietal, region and gifting terms is generally the productive ground. Your licences and shipping permissions remain your counsel's territory; we build campaigns to match them.
Nearly always, and nearly always in the catalogue rather than the ad. Disapprovals usually trace to claim language in product titles and descriptions that Google reads as a health or efficacy claim, and because the feed pulls from your Shopify product data, the fix has to happen at source or it returns on the next sync. We rewrite the offending fields, resubmit, and keep a documented list of language that triggers review so your team stops recreating the problem.
With the availability attribute set to pre-order and an accurate availability date, not as in-stock. Getting that wrong risks the feed and, worse, sets a delivery expectation you cannot meet. We usually keep pre-order products in a separate campaign or asset group so their very different conversion behaviour does not distort bidding on shipping inventory, and we tie the ad copy and landing page to the same stated date.
If you have real inventory on the shop floor, yes, and separately from national. Local inventory ads and store visit reporting can work well for a Valencia Street or Hayes Valley shop, but the buying behaviour, conversion values and geography are so different from national ecommerce that mixing them in one budget makes both unreadable. We split them and judge them on different measures.
Yes, once the feed and the exclusions are right. PMax on a clean feed with margin-segmented asset groups and brand controls is the strongest Shopping vehicle Google has shipped. PMax on a neglected feed is an expensive way to sell bestsellers to people who were buying anyway.
Because Shopping never sees your keywords, only your feed. Title, product type, GTIN and attributes decide which queries you are even eligible for. We routinely find a quarter of a catalogue disapproved or serving on the wrong terms, and fixing that moves revenue before a single bid changes.
Usually yes, but as a separate campaign with its own target and its own conversation. Branded clicks are cheap and convert well, which is exactly why blending them into non-brand hides whether the rest of the account works. Where the spend is material we run a holdout to see what is genuinely incremental.
Next step
Google Ads for your San Francisco brand.
Thirty minutes with the strategist who would actually run your account. We screen-share your store, read your data live, and tell you the three highest-value things we can see from the outside.
Shopify or Shopify Plus stores doing $150k/mo or moreFounder, CEO or eCommerce lead on the callNo deck and no pitch — we open your store instead
Prefer to write it out? [email protected] gets a real reply the same business day, Mon-Fri, 9am-6pm MT.