How do you lower CPC in Google Shopping?

Shopping has no keywords, so the feed is how Google decides which searches your product belongs on. A product identified precisely gets matched to precise queries, and precise queries cost less and convert better than the broad ones a vague listing attracts.

Start with titles. Lead with the words a buyer types, in the order they type them: brand, product, the defining attribute, then size or count. A title that begins with a marketing phrase and buries the product type is competing on generic searches it will lose.

Then fill in the attributes that narrow the match. Google product category and product type place the item in the right part of the catalogue. Colour, size, material, gender and age group qualify it. Identifiers connect it to the same product elsewhere. Each of these removes a class of mismatched query, and mismatched queries are where an expensive click that never converts comes from.

Images belong in the same list. The image is most of the listing, and a listing that earns a better click rate at the same bid is a listing whose effective cost per click falls, because ad rank rewards expected performance rather than spend alone.

A single campaign bidding one target across every query pays the same for a shopper typing your exact model number and a shopper typing a two word category. Splitting them lets each pay what it is worth.

In Standard Shopping, campaign priority plus negative keywords does this cleanly: a high priority campaign with a low bid catches broad queries, negatives push the specific ones down to a lower priority campaign with a higher bid, and the account stops overpaying for browsing traffic. In Performance Max the equivalent is separating brand from non-brand, using search themes deliberately and splitting asset groups by how the catalogue actually behaves.

Custom labels do the same job by product. Label by margin band, price band or rate of sale, then bid each label to what it can afford. A store with a wide price range that bids one target across all of it is systematically underbidding the products that could pay more and overbidding the ones that cannot.

Negative keywords remain the fastest single improvement in most accounts. Run the search terms report, find the queries that consume budget with no orders, and exclude them. Repeat monthly rather than once.

Your landed price sits on the listing, so it changes your click rate before any bidding decision is made. Where you are visibly above other sellers of the same item, the same bid buys less, and stores often read that as a rising cost per click when it is a falling click rate.

Stock depth matters more than it looks. Products that go in and out of stock lose their accumulated performance signal each time, and the campaign has to rebuild it. A catalogue with reliable availability holds a steadier cost per click than one that flickers.

Device, location and time of day carry different prices. Look at where the expensive clicks concentrate before assuming the answer is a lower bid everywhere. A campaign paying a premium for late night mobile traffic that converts poorly is fixable with the reporting you already have.

And remember that competitor behaviour sets the floor. If several new advertisers have entered your category, cost per click rises for everyone, and the useful response is to improve relevance and margin rather than to keep cutting the bid until the campaign stops serving.

The cheapest clicks in any account usually convert the worst. Push cost per click down far enough and you have bought a larger volume of browsing traffic, a lower conversion rate and a worse cost per order, while the metric you were watching improved.

Judge the change on cost per order and on contribution instead. If a lower cost per click came from better feed data and tighter query matching, cost per order falls with it. If it came from bidding down until only the cheap auctions remained, cost per order usually rises.

There is also a floor set by the auction. Below a certain bid a product simply stops showing on the searches that matter, and the campaign quietly loses its best traffic while looking efficient. Watch impression share alongside cost per click so you can see when that is happening.

The question worth asking is not how to pay less per click. It is how much you can afford to pay for a click on this product given its margin, and then whether the listing is good enough to be worth that much to Google.