Smart Bidding Explained: Which Google Ads Strategy Should You Choose?

Open any Google Ads campaign in 2026 and the platform will nudge you toward Smart Bidding. The pitch is straightforward: let Google's machine learning decide how much to bid for each auction, in real time, based on user signals you cannot see. In theory it gets you more conversions for the same budget.

In practice, Smart Bidding works brilliantly for some accounts and badly for others. The difference is whether you have the right data and the right strategy. Pick wrong and you can burn a month of budget before realising it.

What Smart Bidding Actually Does

Smart Bidding strategies use Google's algorithms to set bids at auction time. They take into account hundreds of signals you cannot manually adjust: user device, location, time of day, browser history, prior site interactions, day of week, weather in some cases, audience match, and so on.

The trade off is control. Manual bidding lets you set a fixed maximum cost per click. Smart Bidding sets it dynamically, sometimes 10x your manual bid for a user who matches a pattern Google thinks will convert.

The Five Smart Bidding Strategies (And When to Use Each)

Why Smart Bidding Fails

Smart Bidding is data hungry. Without enough conversion data, the algorithm is guessing. Three signs Smart Bidding is failing in your account:

1. Your cost per click suddenly doubles or triples. Google thinks one user is worth a lot and bids accordingly. If those users do not actually convert, you are paying premium prices for nothing.
2. Conversion volume crashes after switching. Common in the first 7 to 14 days as the algorithm "relearns". If it does not recover by week 3, switch back.
3. Performance is great for two weeks then collapses. Usually means the algorithm overfitted to a small data set, then the pattern changed.

The Conversion Volume Rule

This is the single most important number to know:

Conversions per Month	Recommended Bid Strategy
0 to 14	Manual CPC. Smart Bidding will not have enough data.
15 to 29	Maximise Conversions (no target). Let it learn.
30 to 49	Maximise Conversions with optional Target CPA constraint.
50 plus	Target CPA or Target ROAS, depending on goal.

How to Set a Target CPA That Works

If you switch to Target CPA, do not start with the number you wish you could pay. Start with your actual current average cost per acquisition. If you are paying £40 per lead today, set Target CPA to £40. Once it stabilises, lower it gradually by 10 percent at a time and watch what happens to conversion volume.

Drop the target too aggressively and Google simply stops spending. The algorithm cannot magically find conversions at half the price.

The Learning Period

Every Smart Bidding strategy goes through a 7 to 14 day learning period when you first turn it on or change targets. During this time the algorithm tests different bid levels to figure out what works. Performance is volatile.

Rules during learning:

• Do not make budget changes (more than 20 percent up or down)
• Do not change the conversion goal
• Do not pause ads or add lots of new keywords
• Wait the full 14 days before judging

What About Manual Bidding?

Manual CPC is not dead. It is the right choice when:

• You have fewer than 15 conversions a month
• You are starting a brand new campaign and need to build data first
• You want predictable click costs (such as a strict daily budget you cannot exceed)
• Your conversion tracking is unreliable (Smart Bidding amplifies bad tracking)

We move clients off Manual CPC once they have 30 plus conversions in 30 days. Until then, manual gives you tighter control while data builds.

Enhanced CPC: The Middle Ground

Enhanced CPC sits between Manual and full Smart Bidding. You set the bid, Google adjusts it up or down by up to 30 percent based on conversion likelihood. It is a reasonable compromise for accounts with 5 to 14 conversions a month, building toward full Smart Bidding.

Common Mistakes With Smart Bidding

1. Switching too soon. The campaign needs at least 15 conversions in the last 30 days before any Smart strategy will work.
2. Changing too often. Each switch triggers a new learning period. Pick a strategy and let it run for 30 days minimum.
3. Bad conversion tracking. If form submissions are tracked but bad form submissions (spam, junk leads) are counted equally, the algorithm optimises toward them.
4. Target CPA set too low. Algorithm responds by stopping spend, not by finding cheaper leads.