Campaign performance and wasted spend

Two analyses read your Google Ads reports. Campaign performance compares two periods. Wasted spend finds search terms and keywords that cost money and brought no conversions. Both name candidates for you to check. Neither changes your account.

Campaign performance

What it answers

• Is my ad spend paying off?
• What changed since the last period?
• Which campaigns or ad groups did best and worst?

What you get

For each campaign or ad group, MeasureScale shows these numbers for the period and for the comparison period:

• impressions, clicks and cost
• click-through rate and cost per click
• conversions, conversion rate and cost per conversion
• conversion value and return on ad spend

It also shows how much each number changed, in amount and in percent.

Periods

• With no dates, the period is the 28 days ending yesterday in the account’s time zone.
• The comparison period has the same length. By default it is the period just before. You can ask for the same dates a year earlier, or for no comparison.
• A comparison period that starts before the account’s available history is dropped. The answer then has no comparison.
• You can ask for campaigns or ad groups, for some campaign types such as Search or Shopping, and for enabled or paused campaigns.
• Rows come sorted by cost, highest first.

How to read the label

The label judges the change in conversion rate:

A change counts as real only when chance would produce it less than 5 times in 100.

A campaign is marked when its cost rose and its cost per conversion worsened beyond normal variation.

Confidence

Confidence follows the smaller of the two periods.

Best and worst

The summary lists the best and worst three campaigns by return on ad spend and by cost per conversion. It ranks only campaigns with enough data. A campaign never appears on both lists.

Notes on the numbers

• A ratio with nothing to divide by is empty, not zero. Cost per conversion is empty with no conversions. Return on ad spend is empty with no cost.
• A campaign with conversions but no conversion value has no return on ad spend. The answer says the value is missing.
• A campaign with spend and no conversions has a return on ad spend of zero.
• A campaign that had no impressions in one period shows zero for it, so a change is a change.
• When a campaign’s average daily cost reaches 90% of its daily budget, the answer points to the impression share analysis. That is a pointer, not a diagnosis.
• Amounts are in the account currency, rounded to its smallest unit.

Limits

• MeasureScale does not say what caused a change. A bid change, new ad text, a budget change and a season look the same.
• Google credits conversions to the click date, sometimes days later. The answer warns you when the period ends within 3 days of today.
• A fractional conversion count comes from data-driven attribution. The check on the conversion rate treats it as a whole count, so it is approximate.
• A change in cost per conversion that comes only from cost per click is not tested.

Wasted spend

What it answers

• Where am I paying for clicks that do not convert?
• Which search terms might be negative keywords?

What it reads

• The search terms report, with each term’s clicks, cost and conversions.
• Your keywords, when you ask for them.
• Your existing negative keywords, so it can mark a term you already block.

How terms are judged

A term needs enough clicks to be judged. The default is 20 clicks. You can raise it, and you can set a minimum cost. Terms below the minimum are counted and left out.

A term that Google served under several match types counts once.

MeasureScale compares each term with a normal conversion rate. The normal rate comes from the term’s own campaign when it has at least 30 conversions in the period. Otherwise it comes from the whole account when that has at least 30. With neither, no term is labelled likely waste.

The answer shows the chance for each term. It says “no conversions in 40 clicks” and does not say “never converts”.

Negative keyword candidates

A likely waste search term that you do not already block is marked as a candidate for an exact-match negative. A term you already block shows as blocked and is not counted again. MeasureScale offers no phrase or broad candidates. It never adds a negative keyword for you.

Chance flags

MeasureScale tests many terms at once. About 1 in 20 could be flagged by chance alone. The summary shows how many terms it tested and how many flags chance alone would give. Read a long list with that in mind.

Summary

• The cost of likely waste and the cost of unproven terms, as amounts and as a share of total spend.
• The number of candidates and of terms you already block.
• The share of campaign clicks that the search term report covers.

Limits

• Google hides rare search terms. The answer shows how much of your clicks the visible terms cover. MeasureScale never scales a total up to cover the hidden ones.
• The normal rate assumes a term would convert like its campaign. Intent differs by term.
• The cost shown is the spend at stake. It is not a saving. Blocking a term does not return that money.
• Conversions arrive late, so recent clicks look worse than they are.
• Shared negative keyword lists are not read. A term blocked only by a shared list does not show as blocked.
• Close variants are not applied when MeasureScale checks whether a negative blocks a term.
• A campaign with no conversions at all may have no conversion tracking. The answer says so.

Related

• Check how your Google Ads perform
• Impression share, quality score and budget pace