As Google AdWords evolved into Google Ads, their machine learning has evolved too since the last time we looked at automation and the sheer volume of automated bidding strategies and Smart campaigns for each campaign type out there can be quite intimidating. Not to worry, we’re here to breakdown how automation works in Google Ads, what automated strategies there are, recommended strategies for Search, Display and Shopping campaigns and if they will be useful for your business.
Automated bidding essentially takes away the stress and guesswork out of setting manual bids and bid adjustments to meet your KPIs. This is all based off the keywords in your account and history of your account performance combined with Google’s machine learning (don’t ask how machine learning works at Google, no-one knows. Just see it as a black hole filled with robots and unicorns). Google’s machine learning algorithms can process 70 million signals within 100 milliseconds, and can access information from billions of users from Google, Gmail, YouTube, Google Maps and more (https://www.ppchero.com/automated-bidding-strategies-by-the-numbers/). In bidding, machine learning algorithms allow you to make more accurate predictions across your account about how different bid amounts can affect conversions. These algorithms factor in a wider range of parameters that affect performance than any human could.
Although these both set your bids for you, there is a slight difference between the two. Smart bidding is an automated bid category that only includes conversion-based strategies that use machine learning to optimise for conversions or conversion value (revenue) in each auction.
Smart bidding combines machine learning and contextual signals such as geographic location, time of day and user device to optimise bids at each and every auction. Smart bidding allows you to set performance targets and customise settings to your unique business goals and KPIs.
Target CPA – helps you get as many conversions as possible at the target CPA (Cost Per Acquisition/Conversion) that you set.
To use Target CPA bidding, we would recommend that the campaign has received 30 conversions in the last 30 days so that there is enough data for the machine learning to do its work.
Google will use historical data about your campaign and finds an optimal CPC bid for each auction that your ad is eligible to appear in. The Target CPA bidding strategy adjusts bids on devices, location, times of day and remarketing audiences for you. When setting the Target CPA, Google will recommend a target that is based on our CPA performance over the last few weeks. You can look at further historic performance to find an average CPA, and we would recommend not setting your target CPA too much lower than this, as you may miss out on clicks and can restrict Google’s optimisation of your bids.
Target ROAS – helps you get as much revenue (conversion value) as possible at the target ROAS (Return on ad spend). ROAS is calculated by dividing your revenue (conversion value) by the cost times by a 100 (to get a percentage). For example, if a £500 cost generates £2,000 in revenue, the ROAS will be 400%.
To use Target ROAS bidding, you will need to ensure you are tracking revenue from your e-commerce website correctly. Also, your campaign must have at least 20 conversions in the past 45 days, however, Google recommends that you have at least 50 conversions in the past 30 days to give the machine learning algorithms enough data to make informed bidding decisions. To identify your target ROAS, look back at your historic data from the past 30 – 90 days to see your average ROAS for that respective campaign and use the average ROAS to begin with. Setting a target ROAS higher than your average can hinder your overall performance and is likely to result in your ads gaining less traffic and fewer clicks. Be realistic with your target. Smart bidding is great, but it’s not magic.
Maximise Conversions – helps you get the most conversions for your campaign within your budget.
This smart bidding strategy automatically finds the optimal CPC bid for your ad each time it is eligible to appear in an auction by using historical information about the respective campaign. This strategy should be used over maximise clicks when you have a smaller budget and want to get as many conversions as possible. We would not recommend using this strategy if you are working to a specific KPI, whether that be a CPA or ROAS goal. It is worth noting that you cannot choose “maximise conversions” for Smart Shopping campaigns, however, you can choose “maximise conversion value”. We would recommend using this option if you get varied results on your existing shopping advertising and aren’t sure how to set your bids or don’t have any immediate KPI targets.
Enhanced cost per click (ECPC) – adjusts your cost per click (CPC) to help maximise conversions.
To use ECPC, you will need to set up conversion tracking. When setting up this bidding strategy, select Manual CPC from the drop-down menu and select the Enable Enhanced CPC option. Google will look for ad auctions that are more likely to lead to a conversion, and then raises your max CPC bid. If a conversion or click is less likely, Google will reduce your bid. Your daily budget may be exceeded, however, your overall monthly budget (daily budget x 30.4) will not be exceeded.
Maximise clicks – helps you get as many clicks as possible within your budget.
You can set a cap on the amount you are willing to pay per click. If you don’t set a Max CPC limit, Google Ads will bid as much as it wants in order to achieve the bid automation goal of maximising clicks. This strategy works well on brand campaigns where you want to increase site visits and ensure you are maximising the number of clicks as possible within the budget when someone searches for your brand. If you are working to a certain target or have a limited budget, the maximise clicks bidding strategy may not be for you.
Target Impression Share – helps achieving a goal of showing your ad in a specific percentage of eligible search auctions at an ad location that you set. You can choose the following ad locations – Anywhere on the results page, Top of the results page or Absolute top of results page.
You can also set the percentage impression share to target for that specific campaign. We would recommend targeting either the Top of results page (above the organic search results on the first page) or Absolute top of results page (first ad on the first page) at a realistic target between 50-80%. Only brand campaigns are where we would recommend setting a target of 100% impression share for Absolute top of results page. You can find out more on using Impression Share metrics to your advantage here.
Target outranking share – a portfolio bid strategy that helps your ads outrank ads from another domain that you set.
Portfolio bid strategies are automated, goal-driven bid strategies that help you optimise bids across multiple campaigns. To set up target outranking share bidding, you select the domain whose ads you’d like to outrank in the search results, enter the percentage of auctions you’d like to target to outrank and set a max CPC that Google Ads will not exceed. Google Ads will then raise and lower bids as needed to outrank the domain that you have selected. This strategy works well when there is a domain name you have identified as a direct competitor and you want to ensure that you are ranked higher when they show in an auction.
Target search page location – a portfolio bid strategy that helps show your ad on the top of the page or on the first page of the ad auction. We would recommend using the Target search impression share strategy ahead of this, as that also allows you to choose a target Absolute top impression share.
To maximise results from smart bidding, Google and their machine learning algorithms need enough data to make informed bidding decisions. Google recommends accounts have at least 30 conversions in the past 30 days before using Target CPA bidding, and 5 conversions in the past 30 days to use Target ROAS bidding.
For lead generation companies, we would recommend using a Target CPA bidding strategy. If you are considering online advertising, you must have an idea of how much you are willing to pay for a lead within your monthly budget. When setting a Target CPA, begin with Google’s recommended CPA target or your average CPA of that campaign over the past 90 days. If you hit this Target CPA, you can then reduce your target by 10% and then 20% and so on. Do not set bid limits on your Max CPC as this can prevent necessary bid adjustments that Google is trying to do in order to reach the CPA target. Be careful when using device bid adjustments, as any bid adjustments made in a Target CPA campaign modify the value of your CPA target, rather than adjusting the amount you bid for click. For example, if your target CPA is £20, a mobile device bid adjustment of +25% will increase your Target CPA to £25 on mobiles. We would recommend removing any manual bid adjustments when moving to Target CPA bidding for best performance.
When creating a Smart Display campaign, you should be using Target CPA bidding when focusing on conversions. Smart Display campaigns use these three optimisation technologies: Automated bidding, Automated targeting and Automated ad creation. This bidding method will set bids according to the likelihood of a conversion in each ad auction. In terms of targeting, Google Ads will automatically optimise your targeting for conversions using historic performance data. The two targeting methods Smart display campaigns use are automatic remarketing (website visitors) and automatic targeting – the latter of which uses machine learning to catch customers ‘mid funnel’, meaning they are early in the buying process for services which relate to yours. If you already have an existing Remarketing Display campaign, Google will tend to not shift customers away from that campaign to the Smart one.
When creating your Smart Display ads, you will create your headlines and descriptions and upload your logo and images. You can add up to five headlines (30 character limit each) and five descriptions (90 character limit each). Google will automatically optimise your ads by selecting the best combination of headlines and descriptions to give the best performance.
If you have a Google Shopping feed, you can enable dynamic prospecting. Not to be confused with dynamic remarketing, which focuses on getting the most value out of people that have previously visited your site and shows them products that they have viewed. Dynamic prospecting uses machine learning to predict which products potential buyers are looking for from the items you have in your feed.
This is pretty much what it says on the tin – you can choose to pay only when users convert on your website, and this option is only available on Target CPA bidding. The only caveat to using this option is that your account must have more than 100 conversions in the last 30 days, and 90% of these conversions will have to have occurred within 7 days after someone clicks on your ad. To ensure that you are eligible for Pay for conversions, click on all your Campaigns on the left window, change the date to the last 30 days, click the Segment icon and select Conversions > Days to conversion. You can then check how many of your conversions occurred within 7 days.
When setting Pay for conversions up, you still use the Target CPA, but you will only be billed for conversions received at the Target CPA that you set. So if you set your Target CPA to £20 and get 50 conversions in a month, you will be billed £1,000
For your shopping campaigns, the only automated bidding strategy you should be using is Target ROAS (Return on ad spend). To use this bidding strategy, you will need to set up conversion tracking with transaction-specific values, and you’ll need at least 20 conversions over the last 45 days across existing shopping campaigns. You will also need a remarketing list of at least 100 active users associated with your account in order to allow for your ads to be used for dynamic remarketing. When setting a Smart Shopping campaign up from scratch, we would recommend setting your first strategy to “Maximise conversion value” and not setting a Target ROAS limit. Smart Shopping will use machine learning and use your product feed to show relevant ads across the Google Display and Search Networks, YouTube and Gmail, as well as automatically choosing your Shopping CPCs to meet your ROAS target. Google will create automated ads and placements to appear in the most relevant places, which includes dynamic remarketing ads across the various networks. Dynamic remarketing shows personalised ads for people who have interacted with your website before (also known as stalking).
Set a realistic ROAS target when enabling Smart shopping. If you have been achieving a 2:1 or 200% ROAS for that past year, setting a target ROAS of 10:1 or 1000% could result in some of your budget not being spent and your overall sales could decline. We would recommend setting an initial ROAS target based off your average ROAS over the past 90 days, and slowly increase this once the target has been reached for optimum results. If you have an existing manual Shopping campaign and Display remarketing campaign, the Smart Shopping campaign will take priority. It is recommended that these other campaigns be paused to minimise the interference on Google’s machine learning capabilities.
Target CPA (Cost-per-action) bidding and Maximise conversions are the only available bidding strategies when choosing automated bidding on a video campaign. For Target CPA, Google Ads uses historic conversion data from your campaigns to predict the likelihood an engagement with your video ad will lead to a conversion and will set your bids automatically to get as many conversions as possible at your targeted cost-per-action. Engagements are defined as clicks on the ad or 10 seconds of video-viewing time (if both happen, only the click is counted). We would recommend choosing Maximise conversions as an automated bidding strategy as Google will just help you get as many conversions as possible within your budget
Unlike Display ads, the automated bidding does not automate your targeting, so you will still need to choose your demographics, audiences, topics and placements when creating your video campaign.
Don’t worry, machines aren’t stealing our jobs… Well not for now anyway. Although you don’t set your bids or target audiences manually, there is still a lot you will need to do to give Google’s machine learning as much assistance as possible.
Even after reading this blog, you may still be apprehensive of handing over much of the control over to Google, and understandably so. However, Google Ads have been trialling automation for two years now, and their machine learning technologies are unrivalled and can comprehend data and trends better than any human can. At Anicca Digital, we have been trialling automated bidding and Smart campaigns for a while now and can say we have seen some great results. Our advice would be to test automated bidding on a few of your campaigns, allow Google at least two months of learning, then you should start to see the results that you want.
Last week I had the pleasure of travelling down to the glorious Brighton for 2 days of jam-packed learning at…
Anicca Digital was proud to play a part in an innovative public/ private partnership to train the marketers of the…
Anicca Digital won three awards at The Drum Search Awards last week, making it the agency's biggest awards haul to…
Anicca Digital was in sunny Cambridge on Monday talking to an audience of advanced marketers about Chatbots, Programmatic TV, offline…
When it comes to your digital marketing strategy, paid media is likely to play a large and important role in…
Staying connected with your customers and starting the conversation with new ones is becoming increasingly difficult in today’s digitally complex…