Start from the number that matters
Before looking at any Google Ads metric, decide what "down" means for the business. For most advertisers that is return on ad spend (revenue divided by spend) or, where revenue isn't tracked, the number of qualified outcomes: sales, booked calls, signed contracts. Clicks, impressions, and even conversions are inputs. If they move but the outcome doesn't, there is nothing to fix yet.
Pick a fair comparison. Use at least 14 days, compare against the previous period of the same length, and against the same dates last year if demand is seasonal. Then confirm the drop is real in both comparisons before going further.
Split it: order value or acquisition cost?
Return on ad spend can fall for only two reasons, and they need completely different fixes:
- Each customer is worth less: average order value (AOV) fell.
- Each customer costs more: customer acquisition cost (CAC) rose.
Work out which one moved before touching the account. It takes two numbers: revenue per customer and ad spend per customer, for both periods. Here is the whole tree this guide walks through:
Return on ad spend
├── Order value (AOV)
│ ├── Targeting and period
│ ├── Product or service mix
│ ├── Season
│ └── Order or contract length
└── Acquisition cost (CAC)
└── Cost per acquisition
├── Conversion rate
│ ├── Lost-outcome reasons
│ └── High-value segments
├── Cost per click
│ ├── Ad group changes, ad rank, search terms
│ ├── Match types, budget, keyword count
│ └── Targeting, CTR and quality score
└── Landing pageIf order value fell
If customers cost the same but are worth less, the ads are probably not broken. Check, in this order:
- Product or service mix. Are more customers choosing your cheaper options? A campaign that grew its share of low-price searches will lower AOV even while its own numbers look healthy.
- Order or contract length. Shorter subscriptions, smaller baskets, and shorter bookings all show up as lower AOV.
- Season. Compare with the same period last year before treating it as a problem.
- Targeting and period. New locations, audiences, or schedules can bring buyers with smaller budgets.
The fix here is usually budget allocation: move spend toward the campaigns and searches that bring higher-value customers.
If acquisition cost rose
Cost per acquisition is cost per click divided by conversion rate. So a higher CPA has only two causes: clicks got more expensive, or fewer clicks convert. Check which one moved. It is common to lower bids when the real problem was conversion rate, which makes things worse.
Conversion rate: lost outcomes and segments
If conversion rate fell, start with the leads or orders you lost, not the ones you won. If you record why leads didn't become customers (in a CRM, a spreadsheet, or your sales team's notes), group them:
- No response from the start. The lead never answered. Often slow first contact, or low-intent traffic.
- Not related. Wrong service, job seekers, spam. Every account has some; a rising share means search terms or targeting drifted.
- Budget or price mismatch. The lead wanted something you sell, at a price you don't. Either the ads attract the wrong segment or your offer is out of line with the market.
Then check conversion rate for the segments that bring most of your revenue. A drop in one high-value campaign matters far more than a drop spread thinly across small ones.
Cost per click: the eight usual causes
If clicks got more expensive, go through these in order. The first three explain most cases:
- Ad group changes. Look at CPC by day and by keyword, and at the change history. A bid-strategy switch, a new target, or an edited keyword often lines up with the day CPC jumped.
- Ad rank. Compare top-of-page and absolute-top impression share. A competitor bidding harder pushes your CPC up for the same position.
- Search term relevance. Read the search terms report. Irrelevant searches cost money and drag down quality. Add them as negative keywords.
- Match types. Broad match on a small budget can spread spend across loosely related searches. Check for ads limited by policy or low search volume too.
- Budget. A campaign limited by budget, or one that suddenly got more budget, changes how aggressively it bids.
- Keyword count. Too few keywords starve the campaign; too many split data so nothing learns.
- Targeting. Audience, location, ad schedule, and devices. A new location or a wider audience often costs more per click.
- CTR, ad strength, and quality score. Weaker ads earn lower quality scores, and lower quality scores cost more per click.
The landing page branch
If clicks are stable but conversion rate fell and the lost-outcome reasons don't explain it, look at where the clicks land. Check that the page loads quickly on mobile, that prices and availability match what the ad promised, that the product or offer is still on the page, and that there is one clear next step. Check for duplicate pages competing with each other, and whether the page answers the search that brought the visitor.
Also rule out tracking. If conversions fell to near zero on one day with no change in clicks, the tag, form, or thank-you page broke. That isn't a performance problem at all.
A 30-minute checklist
- Confirm the drop over 14+ days, against the previous period and last year.
- Decide whether order value fell or acquisition cost rose.
- For acquisition cost, decide whether CPC or conversion rate moved.
- For conversion rate, group lost leads by reason and check your top revenue segments.
- For CPC, check change history, ad rank, and the search terms report first.
- Check the landing page and conversion tracking.
- Change one thing, write down why, and compare again after enough data comes in.
This is the same tree Boxoo's agents walk on every review. They read your live Google Ads data, find the branch that broke, and file it as a case with the evidence and a fix you can approve. Run a free audit to see it on your account, or read what each specialist agent checks.
