Search engines are unique tool that provides wonderful and one-of-a-kind possibility to communicate with people exactly at that moment when they need your goods. Only while searching users show their current needs by entering corresponding queries into the search bar. And that makes organic search traffic the most efficient.
At the same time, search traffic (both, organic and paid traffic from contextual advertising and price aggregators with search option) accounts for a great share in online shop traffic sources mix. In most cases it accounts for 30%–70% of all visits to the site. So, how to increase the efficiency of search engine traffic?
Let’s divide all the traffic from search engines into groups and take a closer look at each of them.
1. Product search (query example – “ Samsung Galaxy tab 2 best buy “)
A very large number of people look for specific items and such people are usually converted the best. And it is no wonder as they know exactly what they need. Such traffic should be delivered to product pages and the task of a seller (subject to the availability of competitive prices and acceptable conditions of delivery) is just the only one – do not prevent a user from making a purchase. That is not to force people to pass the quest while trying to place an order.
And what if the product is not available any longer? Indeed, the pages of not in stock products are in search engine index and some traffic is still delivered onto those pages. Also, contextual advertising with a large number of goods is not turned off immediately after the goods are sold off.
2. Category search + filter (query example – “buy tablet”, “buy samsung tablet”, “buy cheap tablet”)
Unlike the first group, these people are yet to make up their mind about what item to buy. Traffic on such requests lands onto the pages of product categories (product listings). The conversion for this traffic is lower than for product search traffic due to two reasons: you may not have the goods users look for and, what a pity! But users may not find the goods which will meet their needs.
At the same time, very often users suggest what they are interested in when entering a query into a search bar (“cheap tablet”, “Samsung tablet”, etc) but they are delivered at the same listing pages (tablets page in our case) and see the same goods. Until recently there was only one solution for the situation like this: the formation of a huge number of pages by means of combining filters and complex patterns of headers, titles and content. See how it’s done on Zappos:
When you choose multiple parameters in a filter (type – sneakers, brand – Adidas, color – black) a separate page with unique URL, title and h1 header is formed. Search engines see that page and deliver the traffic there for the queries like “black Adidas sneakers for men.”
The creation of such functional is a rather complex operation both for online stores database load (it is not so much easy to filter a few thousand / tens of thousands of commodity items by several parameters on a virtual hosting / VPS, and to make it 30-40 times per second), programming and content management. No wonder that only the largest retailers have such possibilities, and even not all of them do have it actually.
What are the alternatives? After performing keyword research and analysis you may learn building connection between keywords and goods through the behavior of users. Such system may take into account what are the most popular goods (viewed, added to a shopping cart, wishlisted, bought, etc.) among users who came from search traffic for each query.
Getting back to the tablets, people who came by query “buy Samsung tablet” will be most likely interested only in Samsung goods, and those who were looking for “cheap tablet” will most likely choose less expensive goods. The system will see that and will recommend exactly those products that better suit a particular visitor.
Do not know what goods to show to those who were looking for “wall decor“? Here you are!
Placing recommendations for users’ search queries with the help of such system is really easy. A special widget can be created for this purpose.
3. Information search
Almost all online shops try to promote by means of feature articles giving advices on how to choose right goods, how to use them, how to store them, etc. One of the most important goals of creating such content is certainly the desire to sell particular products that have been just recommended to visitors. Except that the goods themselves are rarely placed in these articles. Standard CMS systems do not allow doing anything more than inserting images and links to the goods. What if the price for the goods has changed? Or what if it is sold out? Search traffic that was delivered to the page with the article is simply wasted. And here you may still consider search recommendations. Having implemented search recommendations on the inner pages that are specially created for search traffic you may be confident about prices and availability being up to date if the system regularly updates this information. Such system will still show current offers even if the initial goods are out of stock.
It may look like this:
4. On-site search
One of the most efficient place on the website for search recommendation is the on-site search results page. Standard CMS systems use quite straightforward full-text search, and provide search results with goods that contain a search query in their titles.
According to some researches up to 30% of all visits for online shop refer to using the site search. Searching a particular item is a more or less smooth procedure (not taking into account synonyms, typos, transliteration, etc.). But “category filter” search may not be really efficient. Obviously, the built-in CMS standard search algorithms will never be able to better than Google, so the number of problems described above increases by many times. And the solution is still the same – to complete the on-site search results page with recommendations formed on the basis of user behavior analysis. Especially this is vital for online shops that use the internal search results page as a landing page for contextual advertising.
5. The product on-site search
Also, we would like to tell more about the product on-site search. While working on search recommendations, you may notice that users often look for the products that are not available on your website. Isn’t it a gold mine?! All you need to do is to analyse the on-site search and arrange with a purchase department adding to the stock those goods that are in so much demand but yet to be available on your site. This information is provided by Google Analytics after the internal site search is configured (for some reason this option is not offered after you configure it for the first time).
1. Go to the “Admin” (top right corner)
2. Click the “View Settings”
3. Enter into the “queries parameter” field the GET parameter of the internal search results page URL on your site containing the search query that a user entered.
When users search on your site their queries are shown in the URL of the search results page. For example, when a query “text” is searched in the Google the URL of the search results page displays the word “text”, followed by a search term: https://www.google.com/search?hl=en&gl=US&biw=1276&bih=853&q=text&oq=text
After the Google Analytics is set up as described above it will begin collecting the data on the internal search and you will see the following data in the Content -> Site search -> Search terms report:
Highlighted in red color product groups are groups that are not available in a store.
The search recommendations influence
First of all, adding search recommendations positively affects the conversion. It becomes easier for users to find the goods they need on a website and a great part of visitors to your site becomes your customers. What is more, adding search recommendations lowers bounce rates (by 5% -15%), allows to increase the average depth of view and the average duration of the session (the average growth is 8% -10%). As a result, this certainly allows to increase website sales.
In fact, we can take it that the solution described is an absolutely “white” automation of work with behavioral factors and, consequently, that may be used to improve ranking in the search engines, which in turn will drive additional traffic and additional sales.