See how your campaigns are performing

Your .bid files describe what the campaigns should be. This recipe is about the other direction: what the campaigns did — impressions, clicks, spend, conversions.

The tool is bidsmith query. You give it a question written in Google Ads Query Language (GAQL), and it prints the answer using the same sign-in your plan and apply already use. Reading stats is completely safe — this command physically cannot change anything in the account.

Campaign overview, last 30 days

bidsmith query "SELECT campaign.name, metrics.impressions, metrics.clicks, metrics.cost_micros, metrics.conversions FROM campaign WHERE segments.date DURING LAST_30_DAYS ORDER BY metrics.cost_micros DESC"

campaign.name             metrics.impressions  metrics.clicks  metrics.costMicros  metrics.conversions
------------------------  -------------------  --------------  ------------------  -------------------
[W1] magazyny-energii     48211                1322             8540000000          41
[W2] klimatyzacja         21704                611              3120000000          12

2 row(s)

What's a micro?

Google Ads reports money in micros — millionths of your account currency. Divide by 1,000,000: 8540000000 micros is 8,540 PLN (or dollars, euros — whatever the account currency is). It’s the same unit your .bid files use for amount_micros.

What did people search for?

The search-terms report shows the real queries that triggered your ads — the single best source of new keywords and new negatives:

bidsmith query "SELECT search_term_view.search_term, metrics.clicks, metrics.conversions FROM search_term_view WHERE segments.date DURING LAST_30_DAYS ORDER BY metrics.clicks DESC LIMIT 50"

Spot a search term that converts? Add it as a keyword in your .bid file. Spot one that burns clicks without converting? Add it as a negative. Either way the change goes through the normal plan → apply loop.

Which keywords earn their keep?

bidsmith query "SELECT ad_group_criterion.keyword.text, metrics.clicks, metrics.average_cpc, metrics.cost_micros, metrics.conversions FROM keyword_view WHERE segments.date DURING LAST_30_DAYS ORDER BY metrics.cost_micros DESC"

Pick your date range

Swap the WHERE clause:

... WHERE segments.date DURING LAST_7_DAYS"
... WHERE segments.date DURING THIS_MONTH"
... WHERE segments.date BETWEEN '2026-05-01' AND '2026-05-31'"

Get it into a spreadsheet

--format tsv writes tab-separated values that open cleanly in Excel, Google Sheets, or Numbers:

bidsmith query "SELECT campaign.name, metrics.clicks, metrics.cost_micros, metrics.conversions FROM campaign WHERE segments.date DURING THIS_MONTH" --format tsv > performance.tsv

Let an AI assistant do the analysis

Because query is a plain command with --format json output, an AI agent working in your campaign repository can read performance data, relate it to the .bid files, and propose changes as an ordinary diff — which you still review with bidsmith plan before anything is applied. Asking your assistant “pull last month’s search terms and suggest negatives for the Warsaw campaign” is exactly the workflow bidsmith is designed for.

Stats flow one way

query reads; it never writes. Every change you decide to make goes into a .bid file and through plan → apply, so the account stays in sync with the files and the Git history stays a complete audit log.

See also

• bidsmith query — full command reference, output formats, more starter queries.
• Google’s GAQL overview — every field and report you can ask for.
• Audit who changed what, when — the Git side of “what happened in this account?”.