Your bidding strategy is only as intelligent as the data behind it. A bidding strategy tells Google how to set bids in each auction, ranging from full manual control to full automation[1]. Smart Bidding can optimize toward any goal, but if your site converts poorly or generates sparse data, the algorithm flies half blind. In practice, the strategy you choose matters less than the quality of the conversion signal you feed it.
The Strategy Ladder
- Manual CPC lets you set the max CPC per keyword, giving you full control with zero automation. Use this only when you have fewer than 15 conversions per month or need to test new keywords without algorithmic interference.
- Enhanced CPC adds up to 30% bid adjustments based on conversion probability[2]. Most advertisers should skip this halfway measure in favor of full automation.
- Maximize Clicks gets the most clicks within your budget but does no conversion optimization. It is useful only for brand awareness or building initial data. Turn it off once you have conversion tracking.
- Maximize Conversions automatically bids for the most conversions within budget[3]. This is the default starting point once you have 15 or more conversions per month.
- Maximize Conversion Value optimizes for total revenue rather than conversion count[4]. It requires passing revenue values to Google, which most lead-gen advertisers skip but should not.
- Target CPA targets a specific cost per acquisition and needs 30 to 50 conversions per month to learn effectively[5]. Set targets based on actual historical data, not aspirational goals.
- Target ROAS targets a specific return on ad spend and needs 50 or more conversions with revenue values[4]. This is the most sophisticated strategy and the most data-hungry.
Common Mistakes
- Changing strategies too often triggers learning periods of up to two weeks[6]. Every switch resets the algorithm's understanding, so commit to a strategy for at least two to four weeks before evaluating.
- Low conversion rates produce sparse, noisy data that starves the algorithm. If your site converts at 1% instead of 3%, the algorithm needs three times more traffic to reach the same learning threshold. This is why site quality is a bidding input.
- Conversion lag beyond your attribution window means the algorithm misses signal entirely. If your sales cycle is 30 days but your attribution window is 7, the algorithm concludes that 75% of your conversions never happened.
How Scanner Helps
Better site quality leads to higher conversion rates, which produce more conversions per session, which in turn give the algorithm a richer bidding signal. The end result is smarter bids and lower CPA. Scanner audits the technical factors that directly affect conversion rates.