Are You Paying for the Wrong Search Keywords?
Your Google Ads dashboard says clicks are up. Your enquiry inbox disagrees.
That gap — busy ads, quiet phone — usually means one thing. You are not buying too little traffic. You are buying the wrong searches.
The click looked cheap until you asked what it bought
A low cost per click feels like a win. Until you look at what each click was actually shopping for.
A plumber bidding on "bathroom fitting" can collect clicks from people hunting for £40 tap washers, DIY guides and showroom brochures. Each click costs little. None of them can become a £3,000 refit. The campaign reports success. The business feels nothing.
This is the first check, and it takes five minutes. Open the search-terms report and read the actual phrases people typed — not your keyword list, their words. If a third of them describe something you don't sell, you have found the leak. Cost per click never told you this. Cost per qualified enquiry would have.
That economics question matters before any keyword fix. Our guide to whether your marketing buys revenue or just reports is the right lens: measure what each search term produces, not what it costs.
Broad match is wider than most advertisers expect
Broad match does what it says. It matches broadly.
Google's own guidance on keyword match types describes broad match as covering synonyms, related searches and other relevant variations. In practice that net is wide. Bid on "office cleaning" and you can appear for "office cleaning jobs", "office cleaning checklist printable" and "domestic cleaner wanted". Related to cleaning. Useless to a contractor selling monthly office contracts.
As an example, take a small IT firm bidding broad on "server support". One week brings "server support engineer salary", "free server monitoring tools" and "synology server support forum". The firm fixes none of those problems for money. Every click still bills.
Exact match is the opposite discipline. It limits you to searches with the same meaning or intent as the keyword — slower volume, far closer fit. A pattern worth checking is whether a campaign relies heavily on broad match without regularly reviewing its search terms. The traffic may feel random, but it can be broad match doing its job on a keyword that needed a narrower one.
Start your highest-spend ad group on exact for the two or three phrases that describe paying work, and keep broad only where you actively mine its search terms each week. Unwatched broad match is not a strategy. It is an open tap.
Close variants quietly move the goalposts
Even exact match is not literal. Google folds in what it calls close variants — misspellings, plurals, paraphrases and same-intent searches.
Sometimes that helps. "Plumber hyderabad" and "plumbers in hyderabad" are the same person with the same problem. Sometimes it relocates you. "Company registration" can stretch toward "company registration documents download" — a researcher, not a buyer. "Annual maintenance contract" can stretch toward "annual maintenance contract format" — someone writing a document, not signing one.
The shift is small per search and large per month. Ten near-misses a day, every day, at ₹80–₹200 a click, is a salary spent on people who were never customers.
You cannot switch close variants off. You can watch for them. Sort the search-terms report by cost, not clicks, and read the top twenty. The expensive mismatches sit there in plain sight. When a variant keeps appearing and never converts, exclude it with a negative — which is the next section's job.
Negative keywords are where the waste goes to die
Nobody opens an ad account excited about negative keywords. That is exactly why they save money.
Negatives are the phrases you refuse to pay for: free, jobs, salary, course, DIY, download, second hand, cheap — plus whatever your search terms keep surfacing. A training provider selling paid certification keeps paying for "free course with certificate" until "free" becomes a negative. A B2B supplier keeps fielding retail bargain hunters until "cheap" and "price list download" are excluded.
Build the list from evidence, not imagination:
- Export the last 90 days of search terms, sorted by spend.
- Mark every phrase describing someone you cannot serve — the wrong budget, the wrong city, the wrong intent.
- Add those as negatives at account level when universal ("jobs", "free"), campaign level when topical.
- Repeat monthly. New variants arrive constantly.
One caution. Negatives are sharp tools. Excluding "review" can block a buyer comparing options; excluding a suburb can starve a good campaign. Check each candidate against real intent before banning it. The goal is fewer wrong clicks, not fewer clicks.
Branded searches hide behind generic costs
Here the numbers lie in the friendlier direction, and it still costs you understanding.
Your business name converts beautifully. Of course it does — those searchers already chose you. Mixed into the same campaign as generic terms like "web design company", the branded clicks lift the average and make the whole campaign look healthy. Pause the brand terms and the generic side stands exposed: expensive, low-converting, possibly loss-making.
When branded terms help
Keep bidding on your own name when competitors bid on it too, or when your organic listing sits below ads and map packs. Defensive brand spend is often cheap and sensible.
When they mislead
Never let branded and generic share a campaign and a verdict. Split them. Judge generic keywords on their own cost per qualified enquiry, with brand stripped out. That is the number that decides whether your website, SEO or Google Ads deserve the next rupee.
As a hypothetical example, imagine a clinic campaign reporting a 4% conversion rate. If a review of its own data showed that 80% of conversions came from searches for the clinic's name, while generic treatment searches converted at under 1%, the blended figure would hide a very different picture. Those numbers are illustrative, not results from a MakerWeb client or a benchmark; use your own conversion data to compare branded and generic searches.
Quality Score is an intent-mismatch detector
Advertisers treat Quality Score as a technical grade. Read it instead as a relevance complaint.
Google's explanation of the ad auction says placement weighs bid, competition, context, expected asset impact, and the quality and relevance of the ad and landing page. When your keyword buys searches your page cannot satisfy, expected click-through falls, ad relevance falls, landing-page experience falls — and the next click costs more. You pay a penalty for the mismatch, then pay again on every click after it.
That is why keyword cleanup and page work belong together. Pointing a mismatched keyword at a better page helps a little; pointing a matched keyword at a page built for it helps a lot. If your pages also greet paid visitors with the wrong message, read why Google Ads get expensive when the landing page is the problem alongside this piece — then decide which side bleeds more.
And if several fixes feel bigger than keywords — slow pages, confusing offers, no enquiry tracking — hold the ad budget where it is and work through the website problems to fix before more ad spend. Keywords decide who arrives. The site decides whether arriving matters.
A keyword check you can run this week
No audit needed. One hour, one account, honest reading.
- Pull search terms for 90 days, sorted by cost. Read the top thirty like customer statements, not data rows.
- Label each: buyer, researcher, jobseeker, bargain hunter, lost. Be strict. "Might convert" is not buyer.
- Negative the patterns that never buy — at the right level, account or campaign.
- Move your money phrases to exact. Keep one broad experiment running only if you review it weekly.
- Split brand from generic into separate campaigns so each reports truthfully.
- Match the survivors to pages. Each remaining keyword should land where its promise continues word for word.
If the same keyword keeps buying the wrong people after all that, the issue may sit wider than keywords — the offer, the audience or the channel itself. Why campaigns survive every fix and still don't convert covers that deeper layer.
MakerWeb · Build. Secure. Grow.
Spending on clicks that never enquire? Send us your top search terms. We'll tell you which ones buy customers, which ones burn budget, and what to change first.
The wrong keywords don't announce themselves. They arrive as clicks, spend as budget, and report as traffic. Read what people actually searched, and the account starts telling the truth.
