The Rise of AI Campaigns: Part 2
Our first post of this series dissected AI campaigns and their benefits and challenges. One of the large disadvantages is that the default position for these campaigns tends to be to deliver less incremental conversions - but as bad as that may sound, it doesn’t mean you should avoid AI campaigns entirely.
What’s most important is leveraging incrementality in your performance assessment, whether that metric is incremental Cost per Order - CPO(i) - or incremental ROAS - ROAS(i). These are functions of vendor-reported conversions multiplied by the observed incrementality percentage. A campaign with a very high conversion rate and low incrementality can still perform very well - this is common in email, for example. What’s crucial is to implement some type of testing that allows you to understand this tradeoff for AI campaigns. Keep in mind: a 50% drop in incrementality requires a conversion rate that is 2x better just to get back to where you began. Therefore, AI campaigns must deliver a larger percentage increase in conversion rate than the percentage decrease in incrementality.
How to Increase Incrementality in AI Campaigns
Generally speaking, if these AI campaigns represent an amount of spend that’s similar to what you’re spending on traditional campaigns, you are at a higher risk of audience overlap and lower incrementality.
However, if you decide you still want to use AI campaigns, which may especially appeal to you if you have a smaller team and are looking to leverage the automation features, there are a few actions you can take to increase incrementality.
- Apply brand keyword exclusions
- Optimize campaigns for new customer acquisition
- Minimize overlap between PMAX and other “normal” Google campaigns. The more you spend on each of these, the more you risk advertising to the same audience, negatively impacting incrementality
- Upload more comprehensive “customer lists” that include audiences beyond past purchasers (past purchasers such as site visitors, add to cart, etc.)
- Assign a percentage of your budget that can be spent to this group - but keep in mind that this needs to be tightly controlled. At Measured, we have personally seen cases where only 1% of the budget was allowed to go to this “customer list,” but more than 25% of conversions occurred within that group
- Similar to Google, try to minimize overlap between audiences. The easiest way to do this is to avoid spending similar amounts in ASC and “normal” campaigns. We’ve found you to be better off when you’re more significantly invested in either ASC or traditional campaigns. Though Facebook removed the audience overlap tool, you can check frequency as a proxy. Take a look at all campaigns/ad sets; if the total frequency across all of these is higher than what you are seeing on average at the individual level, this indicates the same people are being targeted across multiple campaigns. The bigger the difference, the bigger the overlap, which will result in a lower incrementality.
Should AI Campaigns Be A Part of Your Media Mix?
What do AI campaigns mean for the future of marketing? While they may be here to stay, whether or not you should incorporate this style of campaign in your media mix depends entirely on your own team and level of sophistication.
Here’s a look at how Measured brands invested in AI campaigns alongside traditional campaigns in Q3 of 2023. Clearly, they’re here to stay.
Here’s the question to ask yourself: How comfortable are you, the pilot, with taking your hands off the yoke (basically the steering wheel for planes)? Adopting AI campaigns is like making the move from flying a plane manually to operating on autopilot. You’re still in charge of the final destination, but allow these tools to get you there most efficiently.
There are a few main questions you have to ask yourself:
- Do you feel comfortable trading off control for performance?
- Is your business relatively simple or your team relatively unsophisticated, such that an automated campaign will likely deliver a result you would be looking to achieve manually?
Alignment of incentives
- By default, the platforms have different incentives than you do. They want to sell ads that lead to conversions they can take credit for, while you, as a marketer, want to grow the revenue/profit of the business. Have you taken steps to ensure that the ad dollars being spent on these campaigns are delivering incremental conversions? The steps outlined above can help you with this.
- Do you have a testing program in place to check the incremental performance of these campaigns?
The more of these questions you answered yes to, the more comfortable you should be incorporating AI campaigns into your mix. At the end of the day, there is no right or wrong answer - it all depends on your personal circumstances.
If you would like to speak to someone at Measured to understand your options better, reach out to us today.