Most Google Ads accounts aren't failing because the product is bad or the market doesn't exist. They're failing because of specific, preventable mistakes that accumulate quietly until nothing works. I've audited hundreds of accounts over 20 years and the same patterns come up repeatedly. Here's what to watch for, and what to do about it.
I audited a North Wales building firm's account recently where the owner was convinced the problem was their budget. At £800 a month they felt they couldn't compete. When I pulled the search terms report, 38% of their spend was going on searches like "how to point brickwork yourself" and "DIY brick laying tutorial." The budget was fine. The targeting wasn't. We fixed the match types and negatives in a week and their cost per lead dropped by more than half without spending a penny more.
Most advice tells you to focus on bidding strategy when campaigns underperform. In my experience, that's often exactly the wrong place to start. Bidding strategy is downstream of everything else. Fix the search terms, the structure, and the tracking first, and the bidding tends to sort itself out.
1. Using Broad Match Without a Negative Keyword Strategy
This is the most common and most expensive mistake I see. Broad match keywords in Google Ads mean your ad can appear for any search Google considers vaguely related to your keyword. Without aggressive negative keyword management, you end up paying for clicks from searches that have nothing to do with what you're selling.
I've inherited accounts where 40% of spend was going on completely irrelevant queries. The fix: review your search terms report weekly, add negatives systematically, and favour phrase or exact match for your core terms. For a complete guide to building and maintaining your negative keyword strategy, read The Power of Negative Keywords.
2. Ignoring the Search Terms Report
Your keywords are what you bid on. Your search terms are what people actually typed. These are different things, and the gap between them is where budget goes to die.
Open your search terms report in Google Ads at least weekly. Look for queries that are burning spend without converting. Add them to your negative keyword lists at campaign or ad group level. This single habit will improve most accounts more than any bidding change.
3. Sending Traffic to the Homepage
This one kills conversion rates across the board. A visitor who clicked an ad about "emergency plumber Llandudno" should land on a page about emergency plumbing in Llandudno, not your homepage, which makes them work to find the thing they were just promised.
Every campaign should have a landing page that matches the specific intent of the ad. The closer the match between ad copy and landing page, the higher your conversion rate and your Quality Score, which means lower costs.
4. Broken or Missing Conversion Tracking
If Google Ads doesn't know which clicks are converting, Smart Bidding has nothing to optimise towards. I regularly audit accounts where conversion tracking is either absent or misconfigured: it's counting the wrong events, double-counting, or not firing at all.
This matters more than almost anything else. Without accurate conversion data, you're spending money with no feedback loop. Smart Bidding strategies like Target CPA and Target ROAS become active liabilities without it. The full consequences of skipping this step are covered in The Costly Downfalls of Not Using Conversion Tracking in Google Ads.
5. Enabling Search Partners and the Display Network by Default
When you create a new campaign, Google defaults to including both Search Partner websites and the Display Network. These are very different advertising environments from Google Search and typically convert at a fraction of the rate.
Turn both off when launching search campaigns. Test them separately if you want to explore them, but never let them run alongside your core search campaigns without separate tracking, because their poor performance will distort your data.
6. Setting Target CPA or ROAS Too Aggressively Too Early
Smart Bidding works well when campaigns have enough conversion data, roughly 30 to 50 conversions per month per campaign. Before that threshold, setting an aggressive Target CPA or ROAS instructs the algorithm to chase a goal it doesn't have enough data to hit, which often results in either massive underspending (it won't bid where it can't achieve the target) or budget waste.
Start new campaigns on Maximise Conversions with a daily budget cap while building up data. Introduce tCPA or tROAS once the campaign has meaningful history.
7. Targeting the Wrong Geography
A surprising number of UK businesses targeting their local area are also showing ads nationally, or even internationally, without realising it. Google's default location targeting setting includes people who show "interest in" a location, not just people physically present there.
Change the location option from "Presence or interest" to "Presence only" for any campaign where you want to reach people actually in your area. Then check the geographic reports regularly to see where your spend is actually going.
8. Building Campaigns Around Your Products Instead of Your Customers' Searches
Account structure should follow search intent, not your internal product taxonomy. If customers search for "emergency plumber north wales" and "drain unblocking conwy" those should be in separate ad groups with separate ad copy and ideally separate landing pages, not lumped together because both relate to plumbing.
Tight ad group themes lead to higher Quality Scores, better ad relevance, and lower CPCs.
9. Setting Unrealistic Targets Without Historical Data
Targets need to reflect business economics, not aspirations. If your average order value is £150 and your margin is 30%, your maximum viable CPA is £45, not the £20 target you've entered because that's what you'd like it to be. Campaigns constrained by impossible targets either underspend or show in the wrong places.
Base your initial targets on industry benchmarks, then refine them using your own account data as it builds.
10. Making Too Many Changes Too Quickly
Smart Bidding strategies use a learning period, typically 7 to 14 days, to calibrate to your target. Every time you make a significant change to budget, bids, or targeting, the learning period restarts. Accounts that are constantly being tweaked never stabilise.
Identify the changes that matter most, make them, and then let campaigns run for a meaningful period before assessing. Reactive optimisation based on a single day's data is one of the most reliable ways to make a campaign worse.
11. Running Shopping Campaigns Without a Product Feed Audit
Shopping campaigns live or die by the quality of the data in your product feed. I've taken over accounts where Shopping was running on a feed full of missing GTINs, generic titles like "Product 123," and descriptions copied directly from manufacturer spec sheets. Google's matching is only as good as what you give it. Poor feed data produces poor matching, which means you're showing for the wrong searches and missing the right ones.
Before you touch bids or budgets in a Shopping campaign, audit the feed: are titles keyword-rich and specific? Do product types and categories match what people actually search for? Are GTINs populated for all branded products? A clean, well-structured feed often improves performance more than any campaign setting change.
12. Never Testing Landing Page Variants
Most advertisers put significant effort into ad copy testing but treat the landing page as fixed. That's backwards. The landing page is where conversions happen or don't. A campaign sending 1,000 clicks per month to a page converting at 2% would double its leads overnight if that conversion rate reached 4%, without touching the bids, the keywords, or the budget.
I've seen dramatic CPA improvements come entirely from landing page changes: a clearer headline, a phone number in the header, a trust badge above the fold, a shorter form. Run structured A/B tests using Google Ads Experiments or a tool like Google Optimize. The landing page is not finished. It's always the next test.
13. Letting Google Auto-Apply Recommendations Without Reviewing Them
Google's Recommendations tab suggests account changes, and if you've enabled Auto-apply (or left it on from a previous account manager), those recommendations are being applied automatically. Some of them are reasonable. Many are not. I've audited accounts where auto-apply had broadened match types, added keyword variants that had no business being in the account, and enabled Smart Campaigns features that undermined existing campaign structure.
Turn off auto-apply entirely. Review recommendations manually and apply only what makes sense given your specific account context, not because Google says it will improve your optimisation score. The optimisation score is designed to move towards Google's interests, not necessarily yours. I've written a full breakdown of the specific dangers of letting Google auto-apply recommendations if you want to see real examples of the damage it causes.
14. Setting Location Targeting to "Presence or Interest" Instead of "Presence Only"
Google's default location setting includes people who show "interest in" a location, not just people physically present there. For a local business, this is a slow leak. You end up paying for clicks from people researching your area from elsewhere, people who are not going to book your plumber, visit your restaurant, or use your solicitor.
Change every local campaign to "Presence: people in or regularly in your targeted locations." Then open the geographic report and check where your spend is actually going. You will find budget in places you did not intend to target. This is one of the quickest fixes with the most immediate impact on lead quality.
15. Running Ads With No Conversion Tracking in Place
I still encounter accounts spending hundreds of pounds per month with zero conversion tracking configured. No form tracking. No call tracking. No purchase events. The campaigns are running in the dark. Google's Smart Bidding has nothing to optimise towards, and the account owner has no idea whether any of the spend is generating anything.
This is not a minor gap. Without conversion tracking, every decision you make about bids, budgets, and keywords is guesswork. Set up conversion tracking before you spend a penny. For service businesses, that means at minimum: a thank-you page event for form submissions and call tracking via Google Ads forwarding numbers. For e-commerce, purchase tracking with revenue values. There is no excuse for running a campaign without it.
16. Using Broad Match Exclusively With No Negative Keyword List
This is related to point one but different enough to warrant its own entry. Some accounts I inherit aren't just using broad match carelessly. They're using it exclusively, across every campaign, with no negative keywords of any kind. The search terms report in these accounts reads like a tour through every vaguely adjacent topic Google's algorithm could find.
If you're going to use broad match, and there are legitimate cases for it, you need a robust negative keyword list in place before you launch, and you need to be in the search terms report every week without fail. Broad match without negatives is not a campaign strategy. It's a donation to Google.
17. Pausing Ads the Moment a Campaign Goes Through a Learning Phase Dip
Smart Bidding campaigns go through a learning period when first launched or after significant changes, typically 7 to 14 days during which performance can be erratic. In this window, CPAs often look worse than they will once the algorithm has calibrated. Advertisers who panic at this point and pause or restructure campaigns reset the learning period and never get through it.
I've seen campaigns that could have been strong performers killed in week two because the account owner looked at the first seven days, saw a high CPA, and concluded the campaign wasn't working. Give Smart Bidding campaigns time to learn. Make a note of when significant changes were made and resist the urge to react to data from the learning window as though it represents steady-state performance.
18. Treating Quality Score as a Vanity Metric Rather Than a Diagnostic Signal
Quality Score is not a target to hit. It's a diagnostic indicator. A keyword scoring 4/10 is telling you something specific: expected CTR is low (ad copy isn't resonating), ad relevance is poor (the ad doesn't match the keyword's intent), or landing page experience is below average (the page doesn't match what the searcher expected). Each of those is a different problem requiring a different fix.
I've seen clients proudly report that their average Quality Score is 7 without being able to say what was dragging down the keywords sitting at 3 or 4. Those are the ones costing significantly more per click than they should be. Use Quality Score as a map to where the problems are, not as a number to average and report.
19. Building One Catch-All Campaign Instead of Structured Ad Groups by Intent
Dumping all your keywords into a single ad group with one set of ad copy is one of the most reliable ways to keep Quality Scores permanently mediocre. Google calculates ad relevance by comparing your ad copy to each individual keyword. If one ad has to serve 30 different keywords, it can't be closely relevant to all of them.
Tight ad groups with 5 to 15 closely related keywords, each with ad copy written specifically for that intent cluster, consistently outperform the catch-all approach on Quality Score, CTR, and conversion rate. It's more work to structure properly. It pays off in every auction.
20. Scaling Budget on a Campaign That Hasn't Proven Itself Yet
Increasing budget on a campaign before it has demonstrated consistent, profitable conversion performance is one of the most common ways to accelerate losses. I've seen businesses double or triple their ad spend because a campaign "showed promise" or because Google's budget simulator projected more clicks, only to find that scaling tripled the spend and kept the conversion rate flat.
The right time to increase budget is when: conversion tracking is confirmed accurate, the campaign has meaningful conversion history (at minimum 30+ conversions), CPA is at or below your viable threshold, and the campaign is hitting its daily budget limit. If all four are true, increase the budget incrementally, 20 to 30% at a time, to avoid triggering a new learning period. If any of those conditions isn't met, scaling is premature.
Final thoughts
These twenty mistakes account for the vast majority of wasted spend I find when auditing accounts. Most of them aren't obscure or technical. They're the predictable result of launching too fast, defaulting to Google's settings without questioning them, and not building the habit of regular, structured review. The fix in every case is the same: slow down, get the foundations right, and let data drive decisions rather than guesswork. When the foundations are addressed properly, the results can be significant: one restructured account saw costs cut substantially while generating 171% more leads.
Google Ads management in North Wales. If you'd like a free audit, I'll tell you exactly where your budget is going and what I'd change.
