Set your initial target CPA by calculating the actual average CPA from your historical campaign data over the last 30 days, then start at or slightly above that number. Don't lowball it. Automated bidding needs room to learn, so a target 10-15% above your current CPA gives the algorithm enough conversion volume to optimize before you tighten targets.
Why Your Starting Target CPA Matters
Most teams get this wrong by setting an aspirational target instead of a realistic one. If your manual campaigns have been converting at $80 per acquisition and you flip on Target CPA at $40, the algorithm chokes. It can't find enough cheap conversions, so it throttles impressions and your volume collapses.
Automated bidding systems like Google's Smart Bidding use real-time signals—device, location, time of day, audience—to predict conversion likelihood for each auction. That prediction engine needs a baseline that reflects reality, not your margin goals.
Step 1: Gather Enough Conversion Data First
Google recommends at least 15-30 conversions in the past 30 days before Target CPA performs reliably. Some sources push for 50+ for stable results. If you don't have this volume:
- Run Maximize Conversions first to build a conversion history
- Widen your conversion action to include micro-conversions (sign-ups, demo requests)
- Wait until you cross the threshold before switching to tCPA
Starting tCPA with thin data produces erratic bidding. The system overspends on bad auctions because it hasn't learned your conversion patterns yet.
Generate Proposals with AI in seconds.
Try now
Step 2: Calculate Your Baseline CPA
Pull your cost and conversion numbers from the last 30 days:
Target CPA baseline = Total Ad Spend / Total Conversions
Example: $6,000 spend, 100 conversions = $60 baseline CPA.
Set your initial target at $60-$70, not $45. You can tighten later. The Google Ads documentation on Target CPA confirms that setting targets too far below historical performance limits delivery.
Account for the Learning Period
After you switch bid strategies, expect a 1-2 week learning period. Performance fluctuates. Don't judge results or change targets during this window—the algorithm is recalibrating. Edits reset the learning phase.
Step 3: Ramp Targets Down Gradually
Once the campaign exits learning and hits stable volume, lower your target CPA in increments of 10-15% max per adjustment. Bigger cuts shock the system and tank impression share.
