Beginners should switch from manual bidding to automated bid management once their campaign has steady conversion data—typically at least 15 to 30 conversions in the past 30 days—and a stable account history of 2 to 4 weeks. Automated strategies need this volume to train their models, so switching too early starves the algorithm and tanks performance.
The short answer: wait for conversion volume
Most beginners get this wrong by flipping to Smart Bidding on day one. Google's automated strategies (Target CPA, Target ROAS, Maximize Conversions) are machine-learning systems. They predict the value of each auction based on historical signals. With no data, those predictions are basically guesses.
The practical thresholds Google documents:
| Strategy | Recommended conversions (last 30 days) |
|---|---|
| Maximize Conversions | 0–15 (works with low data) |
| Target CPA | 15–30+ |
| Target ROAS | 50+ (with conversion value) |
| Maximize Conversion Value | 15–30+ |
You can read the full guidance in Google's Smart Bidding documentation. Microsoft Advertising follows similar logic for its automated strategies.
Signals you're ready to automate
1. You have enough conversions
The single biggest gate. Aim for 30+ conversions in 30 days per campaign before moving to Target CPA or Target ROAS. Below that, start with Maximize Conversions, which tolerates thinner data.
2. Your conversion tracking is clean
Automation amplifies whatever it learns from. If your tracking double-counts, fires on the wrong page, or misses offline sales, the algorithm optimizes for garbage. Verify tags fire correctly before you hand over the wheel.
3. You know your true target CPA or ROAS
Don't set a Target CPA blindly. Pull your last 30–60 days of manual data, find the actual average cost per conversion, then set your target at or slightly above that number. Setting it 50% lower than reality will choke delivery and the campaign stalls.
4. Your account has stable history
Google's models pull signals across your account, not just one campaign. A 2–4 week track record of consistent spend and conversions gives the system context.
Generate Proposals with AI in seconds.
Try now
When to stay manual
Manual bidding still wins in a few cases:
- Brand-new accounts with zero conversion history
- Tiny budgets (under ~$10/day) where there's not enough auction volume to learn
- Highly volatile niches with seasonal spikes you want to control by hand
- Learning the fundamentals—running manual CPC for a month teaches you how keywords, match types, and quality score actually behave. That intuition pays off later.
Think of manual bidding as your training wheels. It's not inferior; it's the phase where you generate the data automation needs.
How to make the switch without tanking performance
- Don't change everything at once. Keep budgets, keywords, and ad copy stable when you flip the bid strategy. Change one variable so you can measure impact.
- Expect a learning period. Google flags a 1–2 week "learning" status. Performance often dips before it improves. Resist the urge to revert.
- Start with Maximize Conversions, then graduate. Once you have volume, layer in a Target CPA. Later, if you track revenue, move to Target ROAS.
- Avoid frequent edits. Each significant change (budget swings over 20%, new target CPA) can restart the learning phase.
