AdQuantum’s User Acquisition Team Lead, Alexander Gordeev, explains how automated rules help media buying teams react faster, scale what works, stop what doesn’t, and why automation is a must in daily workflow.
AdQuantum is a performance-driven marketing agency that has been scaling subscription apps, mobile games, and AI-powered tools across global markets for over a decade.
In media buying, managing multiple campaigns at once is just part of the daily routine. And some days, it feels like circus jugglers would be impressed by how many things you’re keeping in the air. That’s exactly when automated rules come into play. They handle routine checks, keep campaigns on track, and save time as media buyers can focus on more important moves.
At AdQuantum, we turn to automated rules only when there’s a real need for them. They’re not a magic pill, but simply a tool, and like any tool, they work only at the right time and in the right context. If you use it at the wrong moment, it won’t make a media buyer’s job easier. It does exactly the opposite, creating unnecessary chaos in the ad account.
Key takeaways
- Automated rules help you run your campaigns smoothly, scale quickly, or handle situations where results are looking good but still uncertain.
- You can pause campaigns that are not performing well, monitor high-budget campaigns at different spending milestones, give campaigns another chance if they meet KPIs, increase budgets for high-performing campaigns, and end creative tests that are not effective.
- You can also use situational rules, such as reducing the budget for proven campaigns or pausing them based on audience activity in different time zones.
- Overall, automated rules support marketers in their work, making it easier and more effective while relying on their expertise and judgment.
When and how to apply automated rules
The primary goal of automated rules is to enhance the campaign system's strength. That’s why a media buyer first needs to сhoose the acquisition strategy, understand the campaign setup and structure: which campaigns are running, how the ad sets are organized, and what logic drives testing and scaling.
The strategy can go in different directions, for example:
- Focus on long-running campaigns: grow them by rotating GEOs or refreshing creatives inside.
- Go for a more flexible setup with an aggressive stop-and-scale approach. This is where automated rules play a key role, and in today’s article, we will break down the automated rules designed in this case.
Most of the time, campaigns begin with a long-term strategy. If this approach fails to produce the desired results, or when there is a pressing need to grow quickly, we switch to the aggressive stop-and-scale approach and bring automated rules into play.
In which scenario do automated rules help?
Automated rules help media buyers in several common scenarios:
- Daily buying control. Monitoring spend and results during the day.
- Restart and launch support. Automatic restarts and, in some cases, creating duplicates.
- Scaling control. Systematic increase in budgets and bids.
- Creative testing. Stopping weak creatives in time, giving more spend to promising ones, and moving successful creatives into the next testing stage.
Once we understand the workflow and scenarios, let's look closely at each rule and see how it works in practice. Before we start, one small note: in the text, we use CPA-based examples, not ROAS.
Stop rule
A stop auto-rule is one of the most frequently used rules for media buyers. It stops underperforming campaigns that don’t meet the KPIs, helping quickly identify weak combinations, save time, and budget.
It’s important to note that stop rules are tested and fine-tuned for each specific project. If the conditions are too strict initially, campaigns risk being turned off too soon and too often, potentially damaging overall performance.
It’s worth repeating: buying performance is not built on automated rules, but on the smart selection of creatives and combinations. If weak material makes it into the campaign, no automated rule will save it. That’s why we initially give a campaign some freedom and only tighten the conditions as spend accumulates, bringing them closer to our KPIs. This way, campaigns that perform well keep running, while underperformers are automatically eliminated.
We apply stop rules to all types of campaigns: low-budget and high-budget.
Low-budget campaigns are new campaigns that haven’t earned trust yet.
High-budget campaigns are those with high trust. They aren't campaigns where we spend a lot of money right away. Instead, they usually start as low-budget, go through the testing phase, prove their results, and only then receive more trust and bigger budgets.
Let’s look at how the stop rule works using the example of a new low-budget campaign with no trust.
Low-budget new campaigns without trust
This approach is used for the fresh campaigns that don’t have enough data and, therefore, get little trust. Here, we allow for a bigger margin before stopping the campaign. For example, with a CPA target of $50, the campaign can spend up to $100 (two CPAs) and must generate at least one purchase. If there’s no result, the campaign is stopped. At $150 spent (three CPAs), there should be at least two purchases. Sometimes the media buyer can set the threshold not at $150 but, for example, at $140, gradually making the rules stricter and reducing the risk of extra spend. In this way, there is almost no margin left, and we don’t let the campaign burn the budget without results. As spending grows and purchases increase, we apply stricter rules and put more pressure on the campaign.
At the same time, we try to keep flexibility. For low-budget campaigns, the rules are usually strict, but there are cases when a campaign is overall performing well, but temporarily goes above the KPI. For example, with $240 spent and 4 purchases, the cost per purchase would be $60 against the KPI of $50. If one more purchase comes in, the cost drops to $48, bringing the campaign back into KPI. That’s why, with smaller budgets, we give them a chance to catch up before shutting them down.
Along with stop rules, we also use start rules. We use it for campaigns in the middle zone that are hard to rate clearly – they are not bad, but not strong either. In such cases, we may restart them with a start rule, hoping they will enter a different auction and perform better. Working with campaigns of medium trust is always risky, and here the media buyer makes their own decision, depending on their own experience. Once again, everything strongly depends on the campaign, its behavior, and the final decision always belongs to the media buyer.
Budget-percentage rule for higher-budget campaigns
When working with high-budget campaigns, we often apply an automated rule based on a percentage of the budget. Campaigns don’t become high-budget right away. They start from low-budget campaigns, show results, and gradually grow. But not all low-budget campaigns reach this level, only those that prove their efficiency.
The automated rule based on the budget percentage is universal. It helps to control large spends and account for data refresh delays.
How do we use this rule? We set several checkpoints at certain percentages of the spent budget and check whether it stays within the acceptable CPA at each stage.
For example, for a campaign with a $60 CPA target, we might set the following thresholds:
At 30% of the budget spent, CPA must be below $75
At 50%, CPA can be slightly higher – for example, up to $67,5
At 75%, we should already be hitting our KPI, for example, CPA < $60
These percentages may be adjusted: you can remove some checkpoints or tighten the CPA threshold. Everything depends on the project, the target country, the time zone, and the accumulated performance data. Different accounts have different country pools and different audience behaviors.
The budget-percentage automated rule helps avoid killing a campaign too early and gives it a chance to ramp up. If at each of these checkpoints the CPA is above the set threshold, the campaign stops automatically.
In addition, for high-budget campaigns, it’s important to account for the time delay between an ad being shown and the event being recorded in the stats. Large projects typically track performance via a BI system that pulls data from multiple sources. Data doesn’t arrive all at once and is aggregated into reports with a delay, usually within about two hours. This is the buffer period after which a campaign can be evaluated.
Moreover, in different time zones, user activity is distributed differently. That’s why the budget percentage for KPI checks should be tied to audience activity, not just the total spend.
Alongside this rule, we usually use a mirrored auto-rule for restarting. In cases where data from the systems took a long time to load, and the campaign was paused due to underperformance, it gets reactivated once it meets the KPI again.
Restart rule
We use automated rules to restart a campaign if its average performance over the past few days meets the KPI, even if it was stopped for underperforming on a single day.
This way, we give setups a second chance and avoid situations where a good campaign is turned off forever due to one bad day.
How do we use this rule? We look at the average CPA over the past few days and compare it to the target KPI. For example, with a KPI of $60, a campaign could perform like this:
Based on the 3-day average, the CPA is still within the KPI (around $56), which means the campaign gets a chance to restart the next day. The logic is simple: we don’t ruin our results today, and tomorrow we try again in case the bad performance was just a fluke.
We usually use the Last 3 Days period for such checks, but in some cases, we take the Last 7 Days. For example, to restart before the weekends if they perform better than midweek. This way, we can give a boost on days with cheaper traffic and reduce spending on “expensive” days.
This rule is especially useful when it’s important to balance between aggressive stopping and keeping potentially viable campaigns. It automatically selects the setups that, based on recent days’ results, are still alive and worth restarting.
Scaling rule
For stable campaigns, we use an auto-rule for scaling to increase budgets only for those setups that perform with a comfortable margin from the KPI. This way, we avoid situations where we scale a campaign right up to the KPI threshold, and then it quickly goes negative.
How do we use this rule? We set a CPA threshold below the target and check that the campaign not only meets the KPI but also has a minimum number of purchases. For example, with a KPI of $60, we might set CPA > 1 ≤ $53 and at least 3 purchases (floating threshold, depending on a project), so we scale only those campaigns that are performing well.
Our standard setup: increasing the budget by 50% once per day. Sometimes we use +25% or +50% every 2-4 hours, but we avoid scaling at the very beginning of the day due to insufficient data, and at the end of the day, since the budget won’t have time to spend in full.
If a campaign has been growing for three days in a row while maintaining a comfortable margin from the KPI, we trust it and set softer conditions for further scaling.
On small budgets, this rule works very precisely – 3 purchases are already a signal to scale. On larger budgets, the event threshold can be increased for greater accuracy. But on average, this is enough, since the lifespan of a campaign under aggressive automated rules is only 2-3 days.
For web projects, we take into account that closer to the evening, scaling may not work as intended – even if we increase the budget, the actual spend will be below the limit. Therefore, once again, we adjust the trigger time and the budget increase step for each niche, country, and audience behavior.
Сreative testing rule
For creative testing, we use a separate automated rule that automatically stops the test if it doesn’t deliver the required result within the allocated budget. It helps to eliminate unnecessary manual stops and helps quickly filter out weak creatives, keeping the focus on those that have real potential to perform. But in general, we apply the same automated rules to creative testing as we do to campaigns. It all depends on the project.
There are different approaches to creative testing. In our case, we give an example of individual testing, within a single task.
How do we use automated rules in this case?
For example, we have three creative variations. All three are uploaded into one campaign with a budget roughly equal to 3 CPA. If, during the test, a creative generates 3 or more purchases at the target CPA, the test is considered passed, and the creative moves into active scaling. If the result is even better (for example, more than three purchases), that’s an excellent signal. If the creative gets only 2 purchases or fewer within the test budget, it’s automatically stopped.
Sometimes, before making a final decision, we manually check if the system limits a creative’s potential by splitting the budget too evenly. For example, with a $150 budget and 3 creatives, each gets $50, which might not be enough to reach 3 purchases. As a result, none of them technically passed the test, even though one could have performed well with more budget.
Additional situational automated rules
In addition to our basic rules, we sometimes use some automated rules that aren’t mandatory for every project but can deliver good results under certain conditions.
Budget reduction instead of a stop
For campaigns with large budgets that have already proven effective for several days, we sometimes use an automatic budget reduction instead of a complete stop. This helps maintain delivery and the campaign’s presence in the auction while reducing risks if performance temporarily drops. This approach is especially useful when the decline in efficiency is short-term.
Temporary pauses based on audience activity
For projects targeting a single geo audience, if the ad account’s time zone differs significantly from the users’ time zone, we use an auto-rule that pauses the campaign during night hours for users and restarts it 2-3 hours before peak activity. Such a rule is applied only after analyzing hourly breakdowns on a solid volume of data to ensure its effectiveness.
Basic workflow for implementing automated rules
The process usually starts with this logic:
- Start with the stop rule. This helps us watch how the traffic works. From the start, it is better to monitor campaigns manually to understand the dynamics.
- Start rule. If we have a CPA margin, we can add a start rule. This gives ad sets in the middle zone more chances to perform.
- Restart and scaling. Once the setup is stable, we add restart rules and begin scaling. With time, we analyze how aggressively we can scale campaigns. It can be +25% every 2-3 hours or +50% once a day. These are just two possible approaches. In practice, each campaign requires its own scaling strategy.
In the end, let’s go back to the main point: automated rules are a trusted helper for a media buyer, but never a replacement. They support decisions and make the work more effective. The key skill is knowing when to give a campaign space to grow and when to tighten control; when to scale fast and when to keep it steady.
It’s important to choose the automated rules wisely. The examples in the article are just one possible strategy for how they can be applied. In other cases, the logic may look completely different: for example, for value optimization, the focus shifts to ROAS, and the structure of the rules is built differently.
Ultimately, in such a collaboration of media buyers’ experience and automated rules, the magic happens. You keep your performance and protect budgets under control, and find the right balance between speed and long-term growth.
