Eight tips that will help you get the most out of Yandex.Audience

Original Russian article on Yandex.Audience by Oksane Dementieva, head of contextual advertising at Artics Internet Solutions

Yandex.Audience is a service that allows you to configure your ad campaigns based on your own target audience data or data provided by Yandex.Metrica or your DMP. All you have to do to create a segment is upload e-mail addresses and mobile phone numbers or mobile device IDs from your CRM. You can also send MAC addresses using the API. Other options include creating segments based on geolocation or Yandex.Metrica data.
In this article, we’ll discuss some of the latest developments and give you tips on how to make the most out of working with the service.

1. You no longer need to make two separate files to create a segment based on telephone numbers and email addresses.
Now you can upload your clients’ phone numbers and email addresses in one list. This streamlines the process a lot. Now it is much easier to gather the minimum number of contacts required to set up retargeting and audience segmentation.

If you have both phone numbers and email addresses, you will increase your odds of identifying target users. After all, someone might register for certain services using their email address but use their phone number for others. When you create one segment combining phone numbers and email addresses, the percent of users that match based on your CRM data increases. That way it’s easier to put together the minimum of 1,000 contacts that you need for a segment. By uploading contact info in one file, it is also easier than ever to create narrow audience segments (even to the point of targeting those who’ve purchased a specific item).

2. If your coverage decreases when you use a segment based on CRM data, you should check your campaigns instead of re-uploading your list of contacts
If a segment based on CRM data results in reduced coverage, the solution is not to re-upload your list of contacts. Segments that are used in active campaigns get updated every few days. The system constantly checks to see whether it can match data for any more users. If more data appears, it recalculates everything.

So the first thing you should do is check your ads to ensure that you uploaded all your graphic banner options, and make sure you included images in your ads and didn’t make any errors.

Put together various creatives, formats, promotions, banners, and wordings. According to our research, coverage will vary based on banner design and ad wording. Try out different options for your ad material and see what captures the largest audience.
You should use your logo and brand name in graphic ads and show your products or services without labels or your logo in text ads. This technique works best and helps you avoid rendering problems.

But if your coverage is small or rises and falls sharply, check your display settings, including time- and location-targeting, keywords (if they are search ads), and bid adjustments. Frequently these settings account for such problems. For example, an advertiser might put together a segment that covers 500,000 users, launch their ad-network campaign in a relatively small region like Khimki, use adjustments that significantly lower their bids for part of their audience, and then not get many impressions as a result.

3. Data for similar-user segments should be as up-to-date as possible.
The system analyzes the interests and behavior of users in a segment over the last 30 days. You shouldn’t use old data for similar-user segments because it very well may prove ineffective.

For example, let’s say you sell cars and upload phone numbers of people who purchased something from you in the past six months. During that time, users’ interests have already changed, so the system will not select a relevant audience for you based on an analysis of their behavior over the last 30 days. As a result, you will serve your ads to users who don’t need to buy a car and your ad campaign will not perform well. To make sure that your ads based on similar-user segments perform well, use contact information from users who fulfilled a target action very recently (within the last month).

4. When you build a similar-user segment based on CRM data, pay attention to the statistics
In the Yandex.Audience interface, there are segment statistics that allow you to determine how similar users within the segment are based on their behavior on the internet. To see this, just click on the icon next to the segment. If the similarity percentage is high, that group can be used to create an effective segment of similar users. So why is this important?

Imagine that your car dealership gets hundreds of visitors every week. Some of them come to test drive a car, others are interested in A- or B-class cars, whereas still others are close to making a purchase. If you analyze how these users behave on the internet, you’ll see that the ones who are ready to make a purchase progressed quickly from watching car videos to searching for dealerships, special offers, and insurance. The algorithm used to build a segment of similar users analyses such behavior and selects a new audience based on certain characteristic signals.

We recommend that you build your original segment based on users who converted. That way the system can identify characteristics that actual buyers share and find new users who have demonstrated similar behavior.

5. You can use Yandex.Audience to serve your ads to users who’ve seen your banner
Yandex.Audience has a tool for managing display ad audiences called pixel. You can use a pixel to collect anonymous IDs of users who’ve seen your banners and find out how recently and frequently they’ve seen them. Based on this data, you can decide whether to show those users the same creative again or serve them a different option.

To use this option, set the conditions for adding users to your segment, segment your audience, and optimize your expenses. In addition to retargeting, you can use these segments for your bid adjustments in search campaigns. For example, you can raise your bid for users who’ve seen your display ad at some point in the last 15 days.

6. You can use Yandex.Audience segments to serve your ads to very narrow groups of people, such as those who visit a specific business center or are students at a particular university
You can create a segment based on an area that you draw on the map using points and straight lines. If you set this type of segment for your ads, they will be served to users located within whatever area you draw.

You can save this type of polygon-based segment in Yandex.Audience and use it for Direct ads just as you would when setting up a retargeting list. Likewise, you can use set a bid correction for your “polygon audience” on search.

Segments must contain at least 1,000 users and the maximum polygon area is 10 km2. The maximum number of polygons in a segment is 10. There are no other restrictions, so you can create any shapes that allow you to target the right audience.
For example, you can outline an entire residential area and serve your ads specifically to residents who live there. Or you can select an airport or business center and target your ads to people who work there.

7. You can use Yandex.Audience segments to create personalized ads that perform well
Find out what your users want. If you have a lot of data, think about what criteria you can use to divide your users into different segments and what offers they might be interested in.

You can break your audience up depending on what stage of the purchase process they are in. For example, you might serve banners advertising your current promotions to users who contacted your call center for the first time. Or you could serve information about your installment plan to users who’ve already taken a test drive. Likewise, you can show banners describing your maintenance services to users who’ve already purchased a car. That way, you can ensure that your ad campaign will perform well.

8. You can use segments in your settings for both display campaigns and context campaigns in Yandex.Direct
Segments do not increase by themselves; if you only have 1,000 contacts, that number will not go up. Likewise, if you use the same segment for different ad campaigns, there will be no difference in coverage.

If your goal is to increase brand loyalty, you can use Yandex.Audience segments in your display campaigns. If you are focused on achieving certain performance results, however, then it makes more sense to incorporate segments into your context campaigns.

Here are some factors that might indirectly affect your audience coverage:
• Banner formats. In both context and display campaigns, the larger the format, the broader the coverage.
• Impression settings. In display and context campaigns, there are different types of targeting and ways to manage bids. By combining these various settings, you can effectively narrow your retargeting audience.

In theory, you can use retargeting in display and context campaigns simultaneously. However, keep in mind that most businesses do not have an audience of millions and you don’t want to wear out limited attention spans. Small and medium-sized businesses should settle on one strategy for establishing contact with users and gradually lead them down the sales funnel. You can use display ads with maximum target audience coverage for those who are interested in your offers right now, and use retargeting in conjunction with raised bid adjustments to motivate users to complete started orders or come back for more. This approach will help you segment your audience properly, create more effective campaigns, and optimize your costs.

For more detailed information about retargeting and using Yandex.Audience segments, check out our Yandex.Direct online school (which includes videos in English).