Last updated November 2nd 2023
To set your campaign up for success, setting the correct CPC (Cost-Per-Click), CPA (Cost-Per-Acquisition), and ROAS (Return on Ad Spend) targets is essential.
When deciding where to set your targets, it's important to balance what is realistic with what is profitable. In the following we go through factors important to finding the sweet spot for optimal campaign performance.
Stay close to your historical performance, this is the biggest indicator of the results you can expect from a campaign. When switching a campaign to smart bidding, review the performance of your campaign over the last couple of months and use that as a starting point.
Start by setting a target close to your historical performance
The CPCs, CPAs and ROAS you have been able to achieve so far is a good indicator for where to set the initial targets of your campaign. Consider adding a +20% to allow the smart bidding algorithm flexibility to find your audience. Once your campaign gets running you can gradually steer your campaigns towards more aggressive goals.
There is an inverse relationship between cost-effectiveness and volume, when looking at a campaign's incremental performance. Consider for a moment the marketing funnel in relation to search campaigns. At the top, we have a large audience of cold traffic. In the middle, we have a lukewarm, medium-sized audience. At the bottom, we have a smaller audience of warm traffic.
The cheapest conversions, or the best return on ad spend (ROAS), can be obtained from bottom-of-funnel users. Getting a conversion from the top and middle of the funnel is more expensive because these users are less likely to convert upon seeing the first impression of our ad. However, the bottom-funnel audience has a limited size. Therefore, if we are looking for volume, we will need to bid on searches from mid- and upper-funnel users as well.
This means that when you scale a campaign, there is often a tipping point where your conversions become more expensive, or your ROAS dips. Finding the balance between volume and profit is key.
An online electronics store is currently running a campaign selling desktop monitors. For each sale, the store earns a profit of $60. The campaign is achieving a target CPA of $10. Within a 30-day period, Google brings in 10 conversions for a total cost of $100. After deducting their ad costs, the store has made a $500 profit.
The retailer decides to scale the campaign by increasing the target CPA. As a result, Google bids in more auctions and now brings in 50 conversions within a 30-day period but at double the price of $20 per conversion. The total cost of the campaign is now $1,000, but after deducting ad costs, the store has made a $2,000 profit.
While the retailer is now paying more per conversion, they are getting more of them, and still at a price which is profitable for them. They could obtain more cost-effective conversions, but at the cost of volume.
To inform both the scaling process and initial target setting, take the time to calculate your break-even cost per acquisition (CPA) or break-even return on ad spend (ROAS), considering the average margins for the products or services targeted in your campaign. This information should guide your goal-setting without determining it, as it is important to take historical data into account to set a realistic target which will allow the campaign to run.
Therefor, if you're running a new campaign or switching bid strategies, you may need to set less aggressive goals to get it started. On the other hand, if you're looking to scale a mature campaign, your break-even values will show you how big you can go and while maintaining profitability.
Resources & Further Reading
"3 ways to think about return on ad spend when growing a business with digital ads" Mike Rhodes delves into the inverse relationship between the likelihood of conversion and volume.
Continuously evaluate and adjust your targets to ensure you are getting the most out of your campaign. Note that Google does not necessarily achieve the exact targets you set for your campaign.
Instead, it uses them as a guideline. This means that sometimes you need to set more ambitious targets than what you are actually aiming for, in order to land on the results you want for your campaign.
Be gradual in changing targets
When adjusting targets up and down it is recommended to do it incrementally, with no more than approximately 20% at a time, giving the algorithm enough space in between changes to catch up. How often you can make changes without causing disruption largely depends on volume. The more volume, the faster the algorithm adapts to changes.
Consider seasonality and plan ahead for expected fluctuations, such as changes in customer behaviour during summer, or events that impact the market, such as Black Friday, Christmas, and other relevant events for your business.
Anticipation is key, so studying seasonal trends in the market can help you plan ahead. This ensures that you don't miss out on traffic during peak times when competition is higher and you might need to relax targets.
Suppose your client runs an e-commerce store that sells gardening supplies. They earn 80% of their profits between May and September, and their competitors follow a similar pattern. During these peak months, competition will be high, but so will sales. Therefore, this retailer should increase their bids during the peak season and decrease them during the remaining months.
When you first change a campaign to a smart bidding strategy, you will see the status 'learning' in the bid strategy status field. This indicates that the bid strategy is in the learning period, which usually lasts between one and three weeks.
During this time, the algorithm is in its initial phase of learning where to bid in order to reach its goals. Since it does not know which searches to bid on in order to perform, it experiments and takes more risks. Performance is likely to be lower during the learning period. It is advisable to discount the data from this period when evaluating the performance of the bid strategy.
Resist making changes when in the learning phase
During the learning phase, it's important to avoid making frequent changes to the bidding strategy or significant modifications to the campaign. Doing so can disrupt the learning process and delay the algorithm's ability to learn to optimise bids effectively.