Myths in contextual advertising: is it possible to influence the operation of auto strategies in Yandex.Direct and Google Ads

The online edition SEOnews and the online advertising automation service eLama disassemble typical misconceptions within the framework of the special project “Myths in contextual advertising”.

We have already found out earlier:

In the new video, eLama, a leading paid traffic specialist Nikita Kravchenko will tell you what affects the work of auto strategies in Yandex.Direct and Google Ads.

Myth: auto strategies cannot be influenced.

Many experts and advertisers think that the work of auto strategies in Yandex.Direct and Google Ads cannot be influenced in any way. This is not true. There are ways to help auto strategies perform better.

Autostrategy in advertising systems is such a “black box”. Specialists cannot influence what is inside this “black box” – the work of the algorithms themselves. But they can influence the data that is submitted to advertising systems, and the effectiveness of the campaign depends on it.

Yandex.Direct has several auto strategies:

• Optimization of conversions;
• Optimization of clicks;
• Target share of ad spend.

Google Ads also has a set of auto strategies:

• Target cost per conversion, which will later move to the “Maximum conversions” strategy;
• Target advertising ROI, which will move to the “Maximum value conversion” strategy;
• Maximum number of clicks;
• Target Impression Percentage.

Most projects work with conversion strategies because they are interested in attracting leads, requests, conversions. Therefore, now we will not consider the strategies “Maximum number of clicks” in Google and “Optimize clicks” in Yandex.Direct Let s talk about how to help algorithms attract more conversions for less money and which strategy to choose if you need to start a new campaign from scratch.

Which conversion strategy to choose

If an advertiser starts a campaign from scratch, does not know the target indicators, then:

• in Yandex.Direct, it is recommended to select the strategy “Conversion optimization by weekly budget” without specifying a specific target price per conversion;
• in Google Ads, you should choose the “Maximum conversions” strategy based on your daily budget.

In this case, the algorithms will be able to work as widely as possible, test, select an audience.

If the goal is one, the target cost of acquisition is known conversions, you can choose to optimize conversions by conversion cost. But at the start of a new project, this strategy is still not recommended.

If there are several goals, then in Direct you can select “Target share of advertising costs”, and in Google Ads – work from value or from profitability.

How a conversion strategy works

1. Indicate a goal or goals for optimization.
2. Choose an optimization method (from budget or target CPA).
3. Feed the data to the algorithm (the more conversions the algorithm receives, the better it will learn).
4. The algorithm is learning.
5. You get the result.

How to help auto strategies work

1. Configure web analytics correctly. This is the base to start with. You cannot launch campaigns until the analytics is built, all lead generation methods are not monitored.
2. Choose the right target for optimization. Initially, it is better to choose one goal – this will make it easier for the algorithms. You can also focus on several goals that are equivalent.
3. Accumulate a sufficient amount of data on the goal (either on the manual strategy, or on the auto strategies for the start, which were mentioned above).
4. Do not limit algorithms at the start.
5. Do not interfere with the learning algorithm – do not pause advertising campaigns, do not change anything for 2 weeks. It is better to make global changes (budget, target CPA, target ROI) approximately once every 2-3 weeks, when the algorithm has already learned.

Golden rule: if everything works, don t touch anything!

How to help the algorithm work as efficiently as possible

1. Consolidate campaigns (especially when it comes to networks – YAN, Google Display Network). It is necessary to enlarge campaigns logically – according to the goal, DDR.
2. Use Google Batch Strategies for Campaign Groups. Using batch strategies, you can combine several campaigns into one group, and more data will accumulate for them.
3. Track conversions using the Google Ads tag.
4. Increase the attribution window so that after a click there is more time to complete the conversion (especially in complex topics, B2B topics, in projects with a long transaction cycle).
5. Select the attribution model for which the maximum data is available (in Yandex.Direct this is usually “Last transition from Direct”, in Google Ads – Data Driven).
6. Leave only the obviously unnecessary negative keywords and negative sites so that the algorithms have more room for testing.
7. Test auto targeting. There is no need to be afraid of automatic targeting, it is worth testing the expansion of reach with their help.
8. Optimize your funnel. Performance is not only affected by the campaign setup, but also by how the site works, how you try to capture leads on the landing page.
9. Follow the recommendations in your account and the status of training. It is desirable that at least 10 conversions come every 7 days, and better – more. If there is no such number of conversions for a global goal – for example, buying or submitting a form, you can optimize for several goals or choose a micro goal on the way to a macro goal (for example, clicking on a button).