Russian SEOs became much more interested in user’s behavioral factor or user-factor in Yandex nowadays.
Search engines had been analyzing users’ behavior all the time, so why’d SEOs started speaking about it only now? Maybe because it’s time for it. After Yandex introducted MatrixNet, bringing websites to the top became harder, especially for new sites, SERP quality raised, SEOs started to look for new factors that can influence site positions. This article describes a group of user-factors that can make influence on site’s position in SERPs, in one way or another, for different types of search queries.
Before we start categorizing user-depended factors, let’s look once again at Yandex’s main objective. It says: “Provide users with maximum quality search”. To provide user with high-quality exerpt, search engine must first analyze user’s query, by doing so, understanding what is he searches for and what he expects to get in response for his question.
Search queries types once implemented by Andrey Broder are working fine enough. They helps identifying user’s motives, but there is place for some other categorization. I.e. “what queries users search when looking for high-quality website?” Say, you’re looking for website, you’re know, you want to bookmark for future, but you don’t know it’s name yet. Maybe it’s about one of your hobbies or interests or provides study material, etc. Say, someone’s looking for “jquery manual”. Most likely, that if that user finds high-quality website with manuals for jquery, he will bookmark it and occasionally will come back to it, looking for information that he needs. It’s hard to classify this kind of queries, but there is formidable amount of them, so let’s call them “interest-resembling queries”. Such queries are:
- salad recipes
- php manual
- movie stars
- movie reviews
- desktop wallpapers
What’s interesting in those queries? First of all, that users, most likely, are going to view large number of pages, since they shown their concern for the subject. It’s obvious, that each query shows user’s interest at given moment, but try and find the difference between such queries as:
- bikers site
- how to shift gears on the bicycle
- sell bicycle
- bicycle photo
- where to buy bicycle in <city>
First one looks like navigational query, but it still resembles user’s concern, so when found a good enough site, user will go through many pages to get more acquainted with site itself, with it’s participants, rules(like if it’s forum, or smth.), watch through the gallery, etc. In other words, by very first query, Yandex or Google expecting user to view large amount of pages. If he left site after only one or two pages viewed, then, probably, this particular site may be not fitting this particular query, nevermind how high it’s numeral rankings are, especially, if that user had not returned back again in, say, a month.
By second query, people would most likely search some specific answer. It may be on the SERP landing page, as well as in 1-2 maximum clicks away from it. Now here search engines would not expect many views at all. In fact, the less pages user seen, the faster he have found information he was looking for, and if he never comes back that is also a good sign.
Fifth variant is pretty much like the second one – user is looking for specific information, where to buy, how to get somewhere, etc. He wants to make as least clicks as possible to get the right answer. The same effect would have queries like “Putin visited Washington” or, sticking to bicycles, “tour de france 2012”.
In other words, for every query, search engine expects some certain behavior from users, thus classifying and categorizing those queries.
So, what user-dependent factors do we know?
1. User experience – how satisfied user will be after visiting given website and his behavior on site. It can be measured by such metrics as:
- number of pages viewed
- bounce rate
- time spent on page and on site at whole
- page loading speed (yep, this one too, including all those scripts, pictures and other media of yours)
- most popular leaving page (the page, users most likely to see last)
- number of returned visitors for certain period
Everything said before about classifying queries was mostly said about this particular user-factors group analysis.
2. User’s behavior in SERP results – what sites are most click-able, how many pages user sees through before finding necessary result. Includes:
- Snippet clickability (CTR). If nobody clicks on it for long enough time, it is probably unattractive and Yandex will exclude it from TOP to improve search quality.
- Number of pages viewed. This has some indirect impact on top-sites’ CTR.
3. Traffic volume – it’s more of quantitative factor than qualitative, but still can be counted as user-dependent. I.e. users can come to your site not only form search, but from facebook, twitter, forums, or pretty much any other website. Analyzing the fact, that the site is being visited, search engines can take proper decisions. Most likely this factor to have meager importance, because of all user-depended factors this one is easiest to manipulate.
Behavioral factors, influencing search ranking could be divided in two groups: qualitative and quantitative. Probably, the former got even more importance for Yandex after introduction of MatrixNet technology. While earlier SEOs used to rely on quantitative factors, such as linkbuilding, number of keywords’ entry in text, title, etc, now it’s time to move on to qualitative characteristics, improving site’s design, it’s usability and making them more friendly for users, instead of search engine crawlers. How would it affect SEO-companies and freelancers is not known yet. Maybe the hour of glory for services like usability and user experience improvement consulting is coming.