Why I hate Yandex turbo pages

Yandex launched its own version “accelerated pages” under the name “Turbo” back in 2017. It was an experiment that we at Rossiyskaya Gazeta (rg.ru) was welcomed, but it’s been 5 years now and I see this technology as more of a problem than a solution.

Next, I share my personal experience of interacting with turbo, which may differ from the opinions of other users, developers and owners of Internet resources whose pages are accelerated by Yandex.

What doesn’t suit me:

User Perspective

  • Moving from Yandex news to turbo pages, I can hardly understand which resource I got to, they all have one “sterile” design, with rather weak functionality, which differs many times from the main site, for example, there are no familiar “Add to bookmarks” or “Subscribe to notifications”. Also, personalization in the form of a personal account, subscription settings and other familiar functionality is completely absent.

  • Turbo navigation is not intuitively clear: some of the links in the text lead to the same turbo pages, and some to the site, by clicking on the link you can only guess where you will end up as a result.

  • Another problem follows from this: when we get to the site itself, we see a completely different interface, different from turbo, which can be confusing.

  • Some content is not available in turbo. It can be annoying when images or widgets don’t load and you end up wanting to scroll down and go to the full site.

I must say that Yandex is struggling with the last point, but, as always, it goes to the developers, who already have a hard time.

From a developer’s point of view

  • Instead of one version of the site, you have to support three: site / turbo / AMP, and each with its own troubles, documentation and requirements.

  • Whatever Yandex says from its sites, that the turbo is simple, but the fact remains that this is a separate version of the site, and it needs to be developed, maintained and monitored. All elements present in the text have to be converted to turbo markup, which means parsing the text of the material, finding images, videos, third-party widgets there and converting it into standard elements / components. And if you missed something, you don’t have to worry, Yandex will definitely remind you, scaring you with an exception from the turbo.

  • Serving content via XML! In principle, I understand why this is so: Yandex is accustomed to working with this format, as well as large media platforms – all Yandex news is exported via RSS. Apparently, the bet was that the Internet media would quickly redo their RSS and not be heavily loaded. At first it was, but then the requirements became more and more, the functionality grew and the burden on developers increased to support this separate XML in a constant mode.

  • To update the content of the page, you need to regenerate the same XML, and if a new widget is added to the site, then if you please send all pages for invalidation, given that we have several million materials, then the task of sending XML via API with a quota of 10,000 elements per one channel (https://yandex.ru/dev/turbo/doc/api/quota.html) does not look so trivial.

  • Errors and warnings are our headache, they need to be constantly tracked, for example, through the Webmaster, which is not the most user-friendly tool in terms of interface, and the errors themselves sometimes lead to a stupor, for example:

    Внутри элемента turbo:content указан URL на Турбо-версию страницы. Чтобы не получать такое уведомление, укажите URL на обычную версию страницы и используйте атрибут data-turbo со значением true

    Here the editor put a link to the turbo page of another publication. If Yandex considers turbos to be full-fledged pages, why should this be banned? As a result, you have to parse the pages for the presence of such links and replace them with normal ones, thanks for the work!

    Here’s another error:

    Слишком много картинок в тексте страницы

    Yes, there is a photo report on this page, which has more than 30 images, there are many such materials, and they need to be removed from the search results, or they themselves are excluded due to quota violations.

  • Custom Components. This is the worst evil. We cannot put a widget on the site that is not in the list of allowed for turbo pages https://github.com/turboext/componentsotherwise we get a warning: “Ваши турбо-страницы отличается от оригинала”and they will be excluded from the search results. As a result, you have to wait weeks for the widget to be available in the turbo components repository, and only after that place it on the site.

    We had a similar story with the voice acting of the text, it was not in the list of turbo components, but we posted it on the site, as a result we received a lot of warnings in Webmaster. The widget, of course, had to be removed, and wait for the appearance in the official repository.

There were also problems in related areas.

In terms of marketing and advertising:

  • Default Analytics! You can add identifiers for Google Analytics (GA), Yandex Metrics, etc., without the ability to configure additional parameters, as a result, we lose a lot of data, and the results of selections from the main site (where a dozen custom definitions are sent to GA) and turbo are difficult to link between themselves. For example: information about the author of the material is sent from the site to GA, with the ability to later build a selection “As the author X was read”, but in turbo there is no such possibility, and it turns out that the author did not work at all.

  • With advertising, everything is simple: Yandex allows only certain formats, no more, no less. If the advertiser pays money for full screen, then we fly by with the turbo and place it only on the site.

  • In addition, direct advertisers refuse to be placed on the turbo, as a result, the pages are clogged with programmatic, and the publisher again receives less income.

  • Targeting of advertising by sections of the site is completely absent, for example, if an advertiser wants to be placed only in the materials of the Sports section, then from the point of view of the turbo it is physically impossible to do this (no Adfox will help)

  • To draw a line under the problems of advertising, I’ll say that the monetization of these pages is generally worse than that of my own resources, I can’t give exact figures, but on average the income is 15-20% lower than that of the main site.

  • I already described the resource identity problem in the section about users, but marketers also suffer from it, it has become more difficult for them to promote a brand that looks like a hundred others and differs only in the logo in the upper corner of the screen.

  • Turbo makes it difficult to share links to original material. Being on such a page, you can share a link only to the turbo page, i.e. the original resource does not receive any backlink, it is received by Yandex! As a result, publishers tend to remove the “sharing” block from their pages.

In terms of resource

  • Turbo pages belong to Yandex, and if the company decides at some point that the project is not profitable and closes it, resource owners will just have to put up with it, and all links will turn into 404 errors. Sooner or later this will happen, and then the problem of “broken links” rise as sharply as possible.

  • Resources are forced to give their pages to Yandex, otherwise there is a risk of losing at least some traffic, and hence advertising revenue. Although there is no official information about pessimization due to the lack of a turbo, no one wants to experiment on themselves, any drop in traffic, even by 5-10%, has an extremely negative impact on media income.

We talked about the negative sides, but are there any positive points? There is! Some of the performance of the turbo is higher – this, of course, the download speed, bounce rates and .. that’s it. The rest, for example, the depth of views and the time spent by users on the page near the site, is greater.

To summarize the above, turbo problems are:

  • placement of content on servers third-party for the publisher that do not support a number of editorial content delivery formats, or their configuration requires significant resources of media developers,

  • diktat in terms of staging only certain ad units,

  • constant problems with counting users,

  • a decrease in publisher revenues, as the monetization of these pages is generally lower than the media’s own resources

  • users are more comfortable and familiar with the main version of the site, and this is confirmed in terms of metrics.

Over the 5 years spent together with turbo, there is confidence that neither users nor site owners need this technology, but only Yandex needs to keep users on its resource.

Similar Posts

Leave a Reply Cancel reply