What automated bidding actually does
Strip away the jargon and automated bidding does one thing: it lets Google's own system set your bids for you, using signals it can see in the moment that you can't, things like device, time of day, location, and how similar this particular search looks to previous ones that converted. In theory, that beats a human guessing a single bid and applying it everywhere.
The catch is that this only works if the system has enough real conversion data to learn from. Without it, it's not making smart, informed guesses, it's making the same kind of guesses you would, just with more confidence attached.
The five main strategies
Manual CPC. You set every bid yourself. No automation, full control. Rarely the right long term choice for most accounts, but sometimes the right starting point for a brand new account with no data at all, or for tightly controlling spend on a small, specific set of keywords.
Maximise Clicks. Automated, aimed at getting as much traffic as possible within your budget. Doesn't care about conversions at all, just volume. Often the best fit for a new or low volume account, especially with a bid cap in place so it doesn't chase clicks at any cost.
Maximise Conversions. Automated, aimed at getting as many conversions as possible within your budget, regardless of what each one costs. Needs a bit of conversion history to work with, but is more forgiving of low volume than the strategies below it.
Target CPA. Automated, aimed at hitting a specific cost per conversion you set. This needs a genuinely steady flow of conversion data, since the system is constantly adjusting bids to hit that number, and it can only do that reliably with enough signal to work from.
Target ROAS. Automated, aimed at hitting a specific return on ad spend. The most data hungry of the five, since it needs not just conversions but revenue values attached to them, tracked accurately, before it has anything meaningful to optimise against.
The conversion volume threshold
A rough rule of thumb worth knowing: most automated strategies want to see somewhere around 30 conversions in the last 30 days before they have enough data to bid well. It's not an exact science, and Google doesn't publish a hard number, but accounts below that threshold consistently struggle with Target CPA and Target ROAS in particular.
Below that volume, the system simply doesn't have enough recent examples to spot genuine patterns, so it ends up reacting to noise instead. That's when automated strategies can behave erratically, hiking or dropping bids in ways that don't obviously connect to anything you'd expect.
Quick check: Look at your conversions over the last 30 days in Google Ads. If that number is comfortably above 30, most automated strategies have enough to work with. If it's well below, that's usually the actual problem, not the strategy itself.
Why new or low volume accounts often do better on manual
This feels counterintuitive to a lot of people, since automation is supposed to be the upgrade. But a new account with barely any conversion history handing bidding decisions to a system with almost nothing to learn from is a bit like asking someone to navigate a city they've never seen, with the map removed. Manual CPC, or Maximise Clicks with a sensible bid cap, gives you a steadier, more predictable base while the account builds up the data it needs.
Once conversions start accumulating consistently, that's the point to reassess, not before.
Signs your bidding strategy is mismatched
A few patterns tend to show up when the bidding strategy doesn't fit the account's actual data volume. Performance that swings wildly week to week for no obvious reason. Cost per conversion that's inconsistent even though nothing else about the account has changed. A "Limited by budget" or learning status message that never seems to clear. Or a nagging sense that bids are moving around in ways you can't explain even after checking the obvious causes.
None of these guarantee the bidding strategy is the issue, but they're worth checking against your conversion volume before assuming something else is wrong.
How to know when to switch, and what to expect
The safest time to move to a more advanced automated strategy is once you've got a consistent run of conversions well above that 30-in-30 threshold, not right after a single good week. When you do switch, expect a short learning period, usually one to two weeks, where performance can be a little less stable while the system recalibrates around the new target. That's normal, and it's worth resisting the urge to switch again immediately if the first few days look rocky.
If you're moving from Maximise Conversions to Target CPA, start with a target close to what you're already achieving rather than an aggressive one, and tighten it gradually as performance holds steady.
The bottom line
There's no universally "best" bidding strategy, only the one that fits how much reliable data your account currently has. Match the strategy to the data, not the other way around, and revisit the choice as your conversion volume grows rather than picking one and leaving it untouched indefinitely.