A Guide to Bid Adjustments: Google Best Practices
Create a system for managing your bids
Being thoughtful and thorough with your bidding and adjustments takes time. That’s why we recommend using automated bidding (in addition to the performance increases many advertisers see). If you opt to manage bids manually, though, it’s important to set up a system for making changes that is manageable over time.
Use base bids to hit your overall portfolio goals, and use bid adjustments to account for segment-level performance differences (by device type, time, location and audience).
The most specific way to set your bids is to use the targeting level. Keyword-level bidding is the most common form of targeting-level bidding, but there are others like product groups, topics or dynamic ad targets. Adjust bids for your keywords or targets to hit your desired performance goals at the aggregate level of your account or bidding portfolio.
The key is to have an acceptable return on your investment while still getting as much volume as possible. Efficiency is not the same thing as profit, so fully consider what your performance goals should be.
Hitting a CPA or ROAS target becomes less imperative when a slight bid increase gets you to the top slot and a big bump in conversions. Remember that as you go through this process you can take a look at the bid simulator to see how changes to your bids or automated bidding targets will affect your conversion volume.
Bid adjustments are intended to optimize performance across various segments of traffic, such as device, location, time-of-day and audience. We recommend setting bid adjustments to try to equalize performance across your different segments. For example, if mobile is performing significantly better than your campaign's average CPA, and tablet is performing significantly worse, it probably makes sense to increase bids on mobile and desktop, while decreasing bids on tablet. By bringing each segment's performance closer to the overall average, your campaign should become more efficient and yield more total conversions.
One way to do this is to calculate bid adjustment changes by dividing the total CPA for the campaign by the CPA of a specific segment of that campaign. For example, for mobile bid adjustments you can divide your total campaign CPA by mobile CPA. You can then apply that figure to the existing mobile bid adjustment. It's very important to note that this is an ongoing process; after applying the new bid adjustments you should monitor your campaign's performance and make further changes as needed to base bids or bid adjustments.
As you apply adjustments, it will be another place where you can think about the full value of different segments. In this example, mobile devices drive more conversions than you can directly measure in AdWords. So you’ll update your conversion and CPA figures before making adjustment calculations.
The device bid adjustment calculation would be:
Device bid adjustment ratio = (Total full value CPA/Device full value CPA)
Then, to get that number to a percentage:
Device bid adjustment change = (Device bid adjustment ratio - 1) * 100
|Device||Cost||Reported Conv.||Full Value Conv.||Full Value CPA||Bid adj. change|
*Note: If you already have bid adjustments in place, remember to apply these new calculations to your existing adjustments.
If this campaign had default bid adjustments of 0% for all three devices, you would apply +27%, +6% and -43% to the mobile, desktop and tablet bid adjustments respectively. It’s straightforward the first time through, then you’ll need to change your modifiers at the conclusion of this process each time in the future. For example, if this campaign had a +50% adjustment in place for mobile, meaning your mobile bids are 150% of your base bids, you would then multiply that 150% figure by 27% to get a further increase of 41%, resulting in a new mobile bid adjustment of +91% (50% + 41%).
You can use the same logic to calculate bid adjustments for other segments such as location, time-of-day, and audience:
Location bid adjustment ratio = (Total full value CPA/Location full value CPA)
Time bid adjustment ratio = (Total full value CPA/Time full value CPA)
Audience bid adjustment ratio = (Total full value CPA/Audience full value CPA)
To get each of these ratios to a percentage you can use the same formula as the device formula above:
Segment bid adjustment change = (Segment bid adjustment ratio - 1) * 100
Here’s an example showing how to set location bid adjustments:
|Location||Cost||Reported Conv.||Full Value Conv.||Full Value CPA||Bid adj. change|
As you put positive bid adjustments in place, keep an eye out for diminishing returns. In some situations you may already be participating in most relevant auctions in a prominent position, so increasing bids may not lead to significant increases in clicks or conversions. Check out your impression share and bid simulator to get a sense of where you might run into a situation like this in your account.
Wait for enough data to make the right decision about your bid adjustments.
You can miss out on opportunities if you neglect to account for conversion delays or have limited conversion data. Give yourself enough time to allow users to complete a conversion cycle before making any changes. You also want to have enough data to be able to make changes based on clear performance trends instead of temporary fluctuations.
With bid adjustments, it’s important to choose the right level of detail for analysis. You’ll want to think about how much detail to use when making changes. For example, when looking at time you’ll need to decide if you want to adjust bids by day of the week, by eight-hour increments or by the hour. While it may be tempting to optimize to the hour, it’s important to ensure that you have sufficient data to make logical decisions for each time period.
Drafts and experiments will tell you when an experiment has received enough traffic to conclude a test. Use experiments to test with confidence.
Develop a strategy for overlap across different bid adjustments.
The intersection of location, time of day, audiences and device can have an effect on what a potential user could be worth to you. On the whole, mobile users may be worth a certain amount to you, but those users might be more or less valuable depending on where they are located.
While automated bidding looks at the combinations of signals like this, it’s not possible to manually bid on the intersection of different signals. If you think combinations of signals drive meaningful differences in performance that can’t be addressed by bid adjustments, automated bidding is the best way to bid on those combinations.
If you want to use manual adjustments, focus on a single bid adjustment first for segments where you have the most conversion data to make decisions with. By setting a single bid adjustment first, you can understand the impact of the change before applying other bid adjustments. When making your adjustments, stagger them, waiting at least a couple of days between each. That way you can first understand what impact the earlier adjustments have on performance. To help you decide which bid adjustment to set first, think about your business operations and your overall goals, and which adjustment will make the biggest impact. It’s a good idea to apply basic business sense to this process:
- You may want to set a location adjustment first if you see different advertising performance across countries, cities, states or zip codes (or if you want to bid higher for users who are close to your stores).
- You may want to set a device adjustment first if consumer response varies widely between mobile, desktop and tablet devices.
- You may want to set a time adjustment first if you have a store or call center that’s only open during certain hours of the day.
As a general rule, we recommend not adjusting bids for dimensions that have fewer than 1,000 clicks and 30 conversions. Increasing the date range for your reports to the past 90 days or more can help. You should revisit your bid adjustments from time to time to make sure you’re taking into account seasonal changes, changes in customer behavior, and your return on investment.