What are optimization rules?
Advertising optimization rules programmatically perform the same routine tasks that a brand or account manager would regularly perform to ensure that the campaigns are maximizing efficiency. They are essentially a set of ‘if-this-then-that’ type statements that can be used to automatically decrease spend on poor performing targets, increase bids on high converting keywords, pause inefficient campaigns, harvest keywords, and adjust other parameters that further enhance advertising performance.
Unlike the “black box” of AI-only automation, optimization rules allow strategists to gain flexibility and control over their advertising technology, while still realizing the benefits of scale. Not all campaigns are meant to achieve the same objective, so not all automation should follow the same rule-based logic.
The key benefit is that they enable human practitioners to spend less time on manual tasks and more time on impactful strategy work. Put succinctly, an optimization rule optimizes performance based on predetermined thresholds.
What metrics should define optimization rules?
This depends on the capabilities that are available on the platform that you are working on. Common optimization metrics are: ROAS (or ACoS), Conversion rates, Cost Per Click (CPC), Click-Through-Rate (CTR) and Impression volume, as these metrics are commonly used to measure performance. Some more advance optimizations include target ROAS, average of broader metrics, and optimizations based on target types.
The rules can be as simple as “Decrease bids by 10% if ROAS is less than $1”. Depending on capabilities, rules can also include several metrics, such as “Decrease bids by 15% if ROAS is less than $1 or if conversion rate is less than 30%, and CPC is above $10”. By using optimization rules, you can make granular optimizations that would take hours to enact if they were done manually.
How often should these rules be implemented?
This is in the eye of the beholder; however, it is recommended that the rules are ran at least on a weekly cadence. This allows the rules to stabilize the account, ensuring that keyword bids are decreasing for poor performers and that high performing targets and valuable keywords are prioritized. It is important to consider the lookback window, which is the time period that the rules are considering when implemented. Typically, it is recommended that the lookback window excludes the previous day, or the last two days to account for data attribution.
A rule that increases bids for campaigns performing above target ROAS might run weekly, and use a lookback window of the last seven days, excluding yesterday. This way, you are automatically setting your campaigns up for success week-over-week.
What are the best practices for using optimization rules?
We recommend using an optimization loop or layering optimization rules on top of each other so that the rules work together to drive efficiency.
- Rule 1: Find poor performers, and decrease bids based on predetermined thresholds. (low ROAS, low conversion rates, etc)
- Rule 2: Decrease high bids – a rule that sets a ceiling for bids to ensure that bid inflation, resulting in wasted spend, will not be an issue.
- Rule 3: Find impression volume – a rule that finds keywords that aren’t getting enough impressions and increases bids to see if the target could be a potential winner.
Another recommendation is to set tiered rules, which would make optimizations based on metric parameters. If you set your threshold to be a $1 ROAS, then all targets that deliver a ROAS of under $1 will be treated the same way. However, a keyword that delivers a $0.95 ROAS and a keyword that delivers a $0.25 ROAS are within different ROAS brackets. We recommend setting ROAS brackets and making optimizations based on those brackets, which allows for segmented improvements.
Example:
Rule 1: For keywords that have a ROAS between $0.25 – $0.50, decrease bids by 50%.
Rule 2: For keywords that have a ROAS between $0.51 – $1.00, decrease bids by 15%.
This way, targets are optimized differently based on tiered performance.
Summary
There are many ways to use optimization rules, and they are a value add to an advertising software solution. When setting up rules, granularity is an important consideration, as the more granular the rule set up is, the more granular the optimizations will be. Optimization rules save time by offering an automatic way to make precise optimizations that are time intensive when done manually. It is important to consider the optimization rules that a platform offers, as these are highly valuable to an advertising tool due to time savings, granular optimizations, and performance improvements that the optimization rules provide.
