Google Shopping Campaign Structure for Small Budgets

Small-budget Shopping campaigns fail for a predictable reason: the account is structured so that the cheapest, least profitable clicks eat the budget before the profitable ones get a chance. With 20 a day you cannot afford to let Google explore. You have to force the money into a narrow lane.

Here is the structure we use for budgets between roughly 10 and 100 per day.

Start with the feed, not the campaign

Shopping has no keywords. Your feed is your keyword targeting. Before you spend anything:

• Titles carry the most weight. Structure them as Brand + Product Type + Key Attribute + Size/Colour. Put the words people search first, because titles truncate.
• Product type should use your own taxonomy, five levels deep if you can. This is what you will use to slice campaigns.
• Google product category should be as specific as the taxonomy allows.
• Custom labels are your control panel. Use them deliberately:
  • custom_label_0 = margin band (high / mid / low)
  • custom_label_1 = price band
  • custom_label_2 = bestseller / new / clearance
  • custom_label_3 = seasonality
  • custom_label_4 = stock depth

Everything below depends on those labels existing.

The three-campaign structure

Campaign A — Proven winners (60% of budget). Contains only products with a demonstrated sale in the last 60 days, or high-margin hero SKUs. Manual CPC or Target ROAS once you have 15+ conversions. Campaign priority: low. This sounds backwards, and it is the key trick — low priority campaigns bid last, but because the other campaigns exclude these products, winners still get served, and at the higher bid.

Campaign B — Contenders (30%). Products with impressions and clicks but no sale yet, plus new arrivals. Medium priority, modest bids. This is your discovery lane, deliberately capped.

Campaign C — Long tail / clearance (10%). Everything else. High priority, very low bids (0.05–0.15). It catches cheap traffic and, more importantly, tells you which unknown products actually attract clicks — those get promoted to Campaign B.

Products must be mutually exclusive across the three campaigns. Use inventory filters or custom labels to enforce that; overlapping campaigns will compete with each other and inflate CPC.

Negatives are your real lever

Shopping campaigns bleed money on informational and comparison searches. Add these as shared negative lists on day one:

• free, cheap, used, second hand, wholesale, bulk
• how to, what is, diy, repair, manual, instructions
• Competitor brand names you do not stock
• Job-related terms if your category overlaps with a profession

Then review the search terms report twice a week for the first month. Every term with 15+ clicks and no conversion becomes a negative. This routine alone typically improves ROAS by 30–50% in the first six weeks.

Bidding on a small budget

• Start manual CPC. Smart Bidding needs roughly 30 conversions in 30 days to work; below that it is guessing with your money.
• Set bids from unit economics, not from suggested CPC. If your margin is 18 on an item and your conversion rate is 2%, your break-even CPC is 0.36. Bid 0.25 to start.
• Move to Target ROAS only for Campaign A, and only once it has real conversion volume.
• Set a target that is achievable. Asking for 800% ROAS on day one just switches the campaign off.

Restrict where you spend

With a small budget, everything you do not exclude is a leak.

• Turn off the Search Partner network initially.
• Geo-target only the areas you can profitably ship to, and exclude the rest explicitly.
• Set an ad schedule after two weeks of data; most stores find several hours a day produce clicks and no revenue.
• Adjust device bids. Mobile often produces clicks at half the conversion rate; a -20% mobile modifier can be the difference between profit and loss.

Landing page and price reality

Shopping traffic is high intent and unforgiving.

• Price competitiveness matters more than ad copy. If you are 25% above the cheapest listing in the carousel, expect a poor conversion rate no matter what you bid.
• The product page must load fast on mobile and show price, availability, delivery estimate, and return policy above the fold.
• Show reviews. Shopping shoppers compare listings side by side, and a star rating in the listing lifts click-through significantly.
• Enable free-shipping thresholds and promotions in Merchant Center so the annotation appears in the listing itself.

A weekly routine that keeps it profitable

• Monday: search terms report → add negatives.
• Wednesday: product-level report → promote or demote SKUs between campaigns.
• Friday: check impression share lost to budget and to rank. Lost to budget means raise the budget on winners; lost to rank means raise bids or improve the feed.
• Monthly: recalculate margin bands and rewrite the titles of the ten highest-impression, lowest-CTR products.

What good looks like after 90 days

You should see budget concentrating in fewer, better products; CPC flat or lower; and conversion rate rising because the traffic is more qualified. If instead you see spend spread evenly across hundreds of SKUs, the structure is not being enforced — go back and check your campaign priorities and inventory filters.

Small budgets do not lose because they are small. They lose because they are spread thin.