Why Most Publishers Waste Half Their Google Ads Budget?

Search ads look like the easiest channel in the world. Someone types the name of a subject, your book appears, they click, they buy. That is the theory, and it is the reason so many small publishers start with Google Ads before anything else.

Then you open the search terms report for the first time and find out what you have actually been paying for.

The gap between keywords and search terms

This is the single most expensive misunderstanding in paid search. A keyword is what you told Google you want to bid on. A search term is what a real person actually typed. With broad match — which is the default, and which Google’s recommendations will keep pushing you back towards — those two things can drift very far apart.

You bid on “military history books”. Google decides that is close enough to “military history documentary”, “history of military uniforms”, “army recruitment”, and “military discount”. Every one of those is a click. Every click is money. None of them are ever going to end in a purchase.

Open your account, go to the search terms report, set the date range to the last ninety days, and sort by cost. Read the top fifty. In almost every audit we run, somewhere between a third and half of the spend sits on terms the publisher would never have chosen deliberately.

Negative keywords are the actual work

The fix is unglamorous and permanent: negative keyword lists, maintained weekly, not once.

Start with the obvious commercial mismatches. If you sell hardbacks at £35, add “free”, “pdf”, “download”, “torrent”, “epub free”, “read online”. If you sell new books, add “used”, “second hand”, “ebay”. If your books are for adults, add “for kids”, “for children”, “ks2”.

Then add the intent mismatches specific to your list. A publisher of naval history should be excluding “navy jobs”, “navy uniform”, “navy blue”, and every model kit query in existence. These are not edge cases. They are where the money goes.

Build these as shared negative lists at account level so a new campaign inherits them automatically instead of repeating the same education at your expense.

Match types, and why you should start narrow

The advice you will get from inside the Google interface is to open up match types and let the system learn. That advice is written for advertisers with the budget to absorb the learning period. A publisher spending a few hundred a month does not have that budget.

Start with exact match on the terms you are confident about — your author names, your series names, your exact book titles, and the two or three subject phrases that genuinely describe what you sell. Add phrase match once exact is converting. Only introduce broad match when you have a mature negative list and enough conversion volume for the algorithm to have something to optimise against.

Yes, this limits your reach. That is the point. Reach is not the constraint on a small press; margin is.

Brand terms are not a victory

Every account we look at has a brand campaign posting a spectacular return on ad spend, and it is almost always doing less than it appears. People searching your imprint name or an exact book title already know who you are. Some of those clicks are genuinely incremental, particularly if a retailer is bidding on your name. Many are not those people would have reached you through the organic result immediately below your ad.

Keep a brand campaign if competitors or aggregators are bidding on your titles, but do not let its numbers flatter your overall reporting. Look at your non-brand campaigns separately. That is where you find out whether paid search is actually acquiring readers.

Track the right conversion

A depressing number of publisher accounts are optimising towards a conversion action that costs nothing to complete: a page view, a newsletter signup, an add to cart. Google will faithfully find you the people most likely to do that free thing.

Set purchase as your primary conversion, with value tracking so a £45 illustrated edition and a £9 paperback are not counted as identical wins. If your order volume is genuinely too low for the algorithm to learn from purchases — under roughly thirty a month — then optimise towards a high-intent secondary action such as reaching checkout, and review manually rather than pretending the automation has enough signal.

What good looks like

A small publisher’s search account that is working usually looks boring. A handful of tightly themed campaigns. Exact and phrase match. A negative list that grows every week. Purchase conversions with values attached. Brand separated from non-brand. Spend concentrated on the two or three subjects where the list is genuinely strong rather than spread thinly across everything ever published.

It will not spend your budget as fast as the recommendations tab wants it to. That is the improvement.