Mini-life hacking to work with Yandex.Direct

    I have been working with contextual advertising for a long time. And I also teach courses for beginners in working with Ya.Direct. For some of them, daily routines seem to be cool "chips." The idea was to share some of them in this short article.


    Customize campaigns for sites with a large catalog of products? When working with semantics, a situation often arises when it is unclear whether “our” request is or needs to be minimized. Especially if it is a specific topic like spare parts.

    In such cases, it makes sense to enter the keyword in the search on Yandex. Pictures and see the issue. If the majority of the pictures correspond to the product, then the request is “our”.

    Another point on the Ya Kartinkam. In the statistics of search queries in advertising campaigns you can see just such an interesting thing:

    At first glance - nonsense.

    But these are just transitions from The search query is now composed of two parts: (initial search query) + (picture header). The "+" symbol can also sometimes catch up with Pictures (in the reports you can see it).

    That is, when a user clicks on a picture, then the name of the picture is added to the search query (which he has already entered), so he can click on several pictures and all the names of the pictures he clicked will be added to the query. From here we see similar requests, seemingly unnatural at first glance.

    Some time ago I decided to evaluate the quality of this site. As a result, it turned out that despite the good depth of views and a small percentage of failures, visitors from the "pictures" do not give conversions. Totally. And although the CTR from the site does not affect the pricing of the phrase, I turned off the display of advertising in this service.

    We save on "refusal"

    For retargeting campaigns, in some cases it makes sense to turn off impressions for visitors, which are marked in “J.Metrica” as “rejection”. Those. spent less than 15 seconds on the site and viewed only one page. This allows you to save on the "random" audience and improve the behavioral site.

    It is very simple to do this:

    1. Create a “Failures” segment in Ya. Metric.

    2. Create a new condition for retargeting and selecting an audience in Ya.Direct, where we indicate the segment of refusals.

    3. In the settings of the advertising campaign, we adjust the rates for the segment of refusers.

    Select the ads

    One of the tasks to increase the CTR of ads is to increase the attractiveness. As soon as the “directors” are not made up. Sometimes on the verge of a foul.

    There is one very simple way to stand out among competitors. Use special characters in headings and texts.

    Available special characters:

    • trademark ;
    • registered trademark;
    • copyright sign;
    • apostrophe ';
    • degrees °;
    • upper case numbers ²;
    • ruble sign

    This will not only highlight the ad, but also help save familiarity.
    By the way, special characters support Direct. Commander and Direct API.

    Pixel Ya. Metrics in the forum profile

    This method allows in rare cases to spy on competitors or simply track at least the effectiveness of their publications.

    Many forums and blogs work on free CMS and allow you to use images in the comments or captions for the post by reference. Ya.Metrika is able to collect some data without embedding JS. This is a gender and age structure and traffic on pages with a pixel.

    1. How it works:
    2. In Ya.Metrik we register a new counter, where we specify the site for which we want to make as the domain.
    3. After registration, you will receive a counter number.
    4. On the forum in the profile / comments / post we add an image by the link of the form (XXXXXX is the number of the counter).

    Adding semantics using competitor analysis services

    We are talking about services that "monitor" how sites are advertised and ranked. And they provide paid information about domains of interest: for which phrases their sites are advertised, the quality of the campaign build, the number of phrases, keywords that are ranked, etc.

    Some of them track 100-120 million requests and up to 30 million domains.
    The only thing you need to remember is that the data that gives us the services is not accurate, because All tools operate on a very limited set of sources and it’s basically impossible to learn all the phrases used by competitors. Those. we are talking about the issue of issuance at some points in time. On the completeness of the samples here is not necessary.

    Why do we need it: we have the opportunity to get new non-obvious phrases and refine our advertising campaigns, increasing our reach and reducing the cost of a click.

    I hope the article was useful for you. Waiting for comments! Share your experience and life hacking!

    Also popular now: