Smart Bidding Explained: Choosing the Right Google Ads Bid Strategy

Manual CPC bidding is now the exception rather than the rule. Google’s Smart Bidding strategies use machine learning to set a bid for every single auction, taking into account signals no human could process in real time: device, location, time of day, browser, audience lists, query wording and much more. Used well, Smart Bidding is one of the most powerful levers in your account. Used carelessly, it can chase the wrong goal very efficiently.

The main Smart Bidding strategies

• Maximise Conversions spends your full daily budget to get as many conversions as possible.
• Target CPA aims to get conversions at or below a cost per acquisition you set.
• Maximise Conversion Value spends your budget to get the highest total conversion value.
• Target ROAS aims for a specific return on ad spend, for example 400% (four pounds or dollars back for every one spent).

Maximise Clicks and Target Impression Share are automated strategies too, but they do not optimise for conversions, so we treat them as specialist tools rather than everyday options.

Choosing the right strategy for your stage

We usually think about it as a ladder. A new campaign with little or no conversion history starts on Maximise Conversions (or Maximise Conversion Value for ecommerce) without a target, so the algorithm can collect data. Once a campaign is getting roughly 30 or more conversions a month and costs are stable, we introduce a Target CPA or Target ROAS set close to what the campaign is already achieving. From there, targets are tightened gradually.

If every lead is worth about the same to you, optimise for conversions. If order values or lead quality vary a lot, optimise for value, because a cheap conversion is not always a good one.

Smart Bidding is only as good as your conversion data

The algorithm optimises towards whatever you tell it counts. If you count a contact form view as a conversion, it will find you lots of people who view contact forms. Make sure your primary conversion actions are the outcomes that actually make you money, assign realistic values, and import offline results such as qualified leads or closed deals where you can. This is the fastest way to improve Smart Bidding results without touching a single bid.

Respect the learning period

After you change a bid strategy or make a significant change to a target or budget, the campaign enters a learning phase that usually lasts one to two weeks. Performance can wobble during that time. The common mistake is to panic and change things again, which restarts the clock. Make one meaningful change, wait, then evaluate.

As a rule of thumb, adjust targets by no more than 10 to 20 percent at a time.

Use seasonality adjustments and data exclusions

Smart Bidding learns from recent history, so it can be caught off guard by short events. For a flash sale or a big promotion, a seasonality adjustment tells the system to expect a temporary jump in conversion rate. If your tracking breaks for a few days, use a data exclusion so those broken days do not teach the algorithm the wrong lessons.

Portfolio strategies and shared budgets

When several campaigns share a goal but each has limited volume, a portfolio bid strategy lets them pool data and hit a combined target. This works well for smaller accounts or for campaigns split by geography that sell the same thing.

How we judge Smart Bidding performance

We look at trends over at least two to four weeks rather than day to day, and we always check conversion lag. Many businesses see conversions arrive days after the click, so yesterday’s numbers will look worse than they really are. Judge the campaign on the period after conversions have had time to come in.

The bottom line

Smart Bidding works best when you feed it accurate conversion data, choose a strategy that matches your business goal, and give it room to learn. The job of an account manager has shifted from setting bids to setting the right goals and protecting the data the system relies on. That is exactly where we spend our time for clients at WeWebMedia.