Your score is a number from 0 to 100. It is not a checklist of good practices. It measures how much of your spend is not working.
The bands
-
70 or above is Good.
-
45 to 69 is Needs Work.
-
Anything below 45 is Critical.
What goes into it
The Badger adds up two kinds of inefficient spend and divides by your total spend for the file's date range:
-
Recoverable waste. Spend on keywords and product targets that are enabled, sit inside enabled campaigns, and produced zero orders in the window. This is the fastest and safest money to recover, and it is the figure shown as "Recoverable this checkup" next to your score.
-
Spend above your target ACOS. For each advertised product that made sales, the spend beyond what your target ACOS would allow. The target is 50 percent unless you set your own when you ran the audit or in your profile.
That inefficient share is scaled onto the 100-point range. Roughly, removing 4 percent of inefficient spend moves the score about 6 points. The lowest score shown is 15, because a zero would read as broken rather than as room to improve.
The findings under the score
Every finding carries a dollar figure, and the list is ranked by dollars, not by count. Findings under 50 dollars are dropped as noise. Each finding is tagged Recoverable, Opportunity or Context, and its severity comes from its share of spend: 10 percent or more is critical, 3 percent or more is a warning, and the rest is context.
The Recovery Forecast
The Recovery Forecast is a separate, deliberately simple projection. It takes the Sponsored Products search terms that spent with no sales, assumes about half of that spend can be recovered and redeployed at your target ACOS, and shows the result per period and annualized. It is a back-of-envelope estimate, not a promise.
Why the score can change between runs
The score depends on the date range in the file, on your target ACOS, and on what was enabled at export time. A file covering a launch week and a file covering a quiet month will score differently for the same account. Compare like with like: the same window length, a month apart.