Splits every campaign's results by mobile, desktop and tablet and flags campaigns where one device converts dramatically worse — with the bid adjustment that would equalize CPA as a starting point.
by Tibbe van Asten & Brainlabs (concept), Chiliad (rewrite)·chiliad.io
One table of every campaign's bidding strategy with the two flags that matter: Manual CPC campaigns with enough volume for Smart Bidding, and target-based campaigns too data-starved to learn.
Finds the reallocation hiding in plain sight: efficient campaigns losing volume to budget caps, idle budgets that could fund them, and full budgets funding poor CPAs.
Runs campaigns only inside their date windows — Black Friday starts at 00:00 without anyone staying up, sales end exactly on time, weekend campaigns run only on weekends.
Compares yesterday's average CPC per campaign against its 30-day baseline and alerts on spikes — the earliest visible symptom of a new competitor, broad-match drift, or a bid change gone wrong.
Compares actual spend against your daily budget settings over 1, 7 and 28 days and alerts by email + Google Sheet when Google overdelivers beyond your thresholds.
Compares actual daily spend against the expected pace based on time of day. Logs over- and under-pacing campaigns and can auto-pause overspenders to protect your budget.
Device Performance Analyzer segments each enabled campaign's cost, clicks and conversions by device and compares the devices against each other. Campaigns where the worst device's CPA exceeds the best device's by a configurable factor (or spends without converting at all) are flagged, with the full device table and the mathematically CPA-equalizing bid adjustment shown as a decision aid. Blended campaign CPAs routinely hide a mobile experience problem — this makes it visible.
| Variable | Default |
|---|---|
DATE_RANGE | LAST_30_DAYS |
MIN_DEVICE_COST | 50 |
CPA_GAP_FLAG | 1.75 |
EMAIL_ADDRESS | (empty) |
=== Device Performance Analyzer (LAST_30_DAYS) === Campaigns with 2+ judged devices: 14 Flagged: 2 Generic — Broad MOBILE cost: 1840.20 clicks: 1650 conversions: 18.0 CPA: 102.23 <- WORST DESKTOP cost: 1626.44 clicks: 1102 conversions: 52.0 CPA: 31.27 <- best -> MOBILE: equalizing bid adjustment approx -69% Competitor — Phrase MOBILE cost: 412.80 clicks: 391 conversions: 0.0 CPA: - <- WORST DESKTOP cost: 388.15 clicks: 240 conversions: 9.0 CPA: 43.13 <- best -> MOBILE: consider strong negative adjustment or exclusion Email sent to you@example.com
The suggested adjustment equalizes CPA arithmetically — sanity-check against lead quality and attribution (mobile often assists conversions that close on desktop). On campaigns using Smart Bidding, device bid adjustments are largely ignored (except -100%); there, treat a flag as a prompt to fix the mobile experience, not to stack modifiers.