Search engine optimization is the processes and actions taken to affect a websites appearance and ranking in a search engine. SEO is important because traffic from (organic) searches are free, as opposed to the paid traffic from adwords or similar programs. In an increasingly competitive market, the added costumer acquisition cost incurred by paid traffic may be the difference between failure and success.
In search engine optimization there are two partially opposing interests; Search engines, in general, want to show the most accurate and relevant responses to a user’s query, while a web shop owner wants his webshop shown at the topmost position.
This article will primarily refer to Google when discussing search engines, because Google, at least in Europe, has a market penetration of more than 90%. Furthermore, Google seems to be the primary innovator and pretty much defines the standards, which are then adopted by other search engines.
Search engines in general do not disclose their exact ranking and indexing algorithms but rather hint at important parameters in their algorithms. To be absolutely clear, only Google knows the exact implementation of their indexing and ranking algorithms and while they do not disclose the exact nature of the algorithms they have given hints as to what they think is important.
Obviously many other people have ideas and thoughts about the algorithms used by search engines. These ideas are often conflicting, extremely seldom scientific or based on concrete knowledge. We shall politely disregard any opinion, theory and idea concerning SEO unless it originates directly from one of the major search engines and is documented.
That being said, nobody at the eSeller Cloud team are SEO experts nor do we claim to be. There are many good SEO professionals that have great experience and knowledge well founded in documentation and we appreciate their input and knowledge. As we are not SEO experts, we stand corrected and value any well documented input that includes references to official sources.
This article, like the SEO support in eSeller Cloud, is based on the following sources:
- Google Search Engine Optimization Starter Guide
- Google webmasters blog
- How Search Works
- The Anatomy of a Large-Scale Hypertextual Web Search Engine - (Brin, S. and Page, L. (1998) The Anatomy of a Large-Scale Hypertextual Web Search Engine. In: Seventh International World-Wide Web Conference (WWW 1998), April 14-18, 1998, Brisbane, Australia.)
We strongly encourage you to read the above sources and to form your own informed first hand impression.
Search Engine Processes
In order for a Search Engine to produce results at least, two distinct processes take place; Indexing which is the process of gathering information about websites and their contents and Retrieval or Search which is the process of retrieving results for given query.
The following two sections briefly outlines the two processes which will be treated in more detail in the following sections.
Search engines crawl websites by following links (or sitemaps) from page to page while retrieving analyzing and cataloguing the contents of the page. Google builds a vast list of URLs that needs to be crawled and distributes the actual crawling to a number of crawling robots. The crawlers retrieve the contents of each URL, which is then parsed and indexed. The parsing and indexing operation entails counting the occurrence of each word while taking font, capitalization, position and other factors in the document into account. Words that have more prominent font, are featured earlier in the document or have capitalization are considered more important in the indexing.
Many other factors are taken into account when indexing, there are “several hundred” factors or ranking signals. Ranking signals include Page Rank, Mobile access, Load speed and many others.
In general, Ranking signals are dampened, so that rank does not increase linearly by increasing ranking factor but rather drops of when a certain low number of ranking signals have been encountered.
Page Rank is an iterative algorithm that ranks URLs based on the numbers of links from other pages and the Rank of the linking pages pointing to a specific URL.
Mathematically Page Rank can be described as:
PR(A) = (1-d) + d (PR(T1)/C(T1) + ... + PR(Tn)/C(Tn))
Where page A has pages T1...Tn that links to page A, d is a dampening factor and C(T1) is the number of links going out of page T1.
For the sake of example, we assume the three URLs A,B and C where A and B has links to C. In order to determine the Rank of C Page Rank sums the Page Rank of A and B divided by the number of outgoing links from each page of the pages A and B respectively and add a dampening factor.
The importance of the Page Rank in the overall indexing process has decreased significantly because Page Rank initially made it possible for a website to achieve a high rank simply by getting (or even buying) links from other pages.
In addition to the above ranking factors, Google also uses the anchor text of links as a ranking signal and for inclusion into the list of words. For instance in the case of website A containing a link to website B, the link could look something along the lines of:
<a href=”B.com”>performance tuning cell phones</a>
Google takes this to understand that website B contains information about the word and or sentences included in the anchor text.
In the case of Google the entire indexing operation is described in depth in the “The Anatomy of a Large-Scale Hypertextual Web Search Engine” paper.
Based on the above section it is possible to summarize SEO relevant information about crawling:
For a page to be indexed it has to:
- Be accessible for the Search Engine
- Contain contents in a format that allows the Search Engine to parse it
- Be added to the list of URLs that should be crawled.
Furthermore, we have achieved at least some information about how rank is calculated
- Rank is based upon several hundred ranking factors of differing importance
- The value of any single ranking factor is not linear. E.g. overdoing something does not help
- The occurrence, position and font of keywords are important
- Incoming links was previously very important although it is now of less importance.
Search / Ranking
When a user types a query in a search engine, the search engine parses and analyses the query and fetches and displays the most relevant results.
In slightly more detail, the query goes through the following (simplified) process:
1. Parse the query
2. Look up ids of individual words
3. Look up any document that contains all words
4. Compute the rank of documents
5. Sort the documents by rank
6. Return list
In addition to the above, the ranking also takes the currently logged in Google+ user into account.
The results are shown in a list, which is commonly known as a Search Engine Result Page (SERP).
From a SEO perspective, several aspects of Search and ranking are important; Obviously it is important that your page is listed in the SERP and preferably as high in the SERP as possible. It is however also important that the snippet describing your page results in a high click through rate (CTR) and that users entering your shop from a search, stay on the site and are converted.
Googlebot is the piece of software that crawls and retrieves websites for indexing. For Googlebot to be able to do its job the following basic criterions should be fulfilled:
Googlebot must know about the site from one of the following sources
- Sitemap (in google webmasters tools)
- Url added for indexing (in google webmasters tools)
- External links pointing to the website
Googlebot must be able access the site
- The site must be public and accessible and should be in “Live mode” (i.e. not include noindex/nofollow)
- The site or page cannot be password protected
Googlebot must be able to understand the contents
- Reasonably well formed HTML
Googlebot must see contents which is similar to what a human visitor would see
- The site should not contain any hidden contents which is intentionally misleading
In terms of indexing eSeller Cloud shops you should ensure that:
- You shop is live (Check that live mode is enabled)
- You have added or updated the sitemap to webmasters tools
- The pages that needs to be indexed are not password protected
- Search engine indexing is enabled
- That the shop passes w3c validation with a reasonably low number of warnings and error
You should periodically check the status and frequency of indexing in Google webmasters tools.
With the above in place, no technicalities should be in the way for basic indexing but the most important part is of course the contents of your shop.
If your content is not meaningful, unique, coherent and contains the terms users search for it is unlikely that your site will be given a high rank. Writing good quality contents is time consuming but is very important. As a very minimum, you should have a textual description of your products and their properties and meaningful description in your menus.
If you consider this task too time, consuming you should consider enlisting a professional copywriter. Professionals often write good descriptive, informative and well-worded texts much faster than most other people do and their services are not overly expensive.
In general, you should follow these guidelines when writing product texts
- Provide good descriptions of your products and their advantages and USP’s in all target languages
- Consider what users will search for when researching the product and include those terms and sentences if and only if they make sense in the description
- Understand the use cases for a product and describe the products place in the use case
- Use headings sparingly but use them where it makes sense
- Ensure that the text is unique and not copied from another site (e.g. Avoid descriptions supplied by manufacturers as they are often used on the manufactures website or on competing webshops)
In addition to product descriptions, you should also include texts in your menus. These texts can describe the common traits of the products in the menu and or the use case for the products.
All pages in you shop should have meaningful titles – Titles help users understand the contents of a page and as they often feature prominently in search results, it is important to have meaningful and accurate titles.
The default behavior in eSeller Cloud is that titles for product information pages consists of the product name unless a specific title is provided for the product.
The page title is defined in the content design and it is of course possible to alter the title using any available run time variable.
If you choose to manually write page titles you should observe the following guidelines
- Write brief titles that are descriptive of the comments
- Write distinct titles for all pages
- Avoid stuffing keywords into the title
Meta data and mark-up
To make it easier for search engines to understand and analyze the contents of a webpage it is possible to supply a page with meta information and markup which is meant primarily for Search engines.
It is important to stress that there should a reasonable consistency between the human readable content and that presented to the search engine – Major discrepancies could be viewed upon as cloaking and can result in penalties.
The contents of the description meta tag may be used by google as snippets when showing results for search queries and it is as such an important piece of information.
The default behavior of the description is to show the contents of the products meta description field if it is present and to use the product short description otherwise.
This means that you should focus initially on writing good product descriptions for all products as these will also be used for the meta description.
Subsequently when good overall product descriptions are in place, you can turn you attention to writing specialized and optimized meta description.
Google has not published any length restrictions on the tag but as the maximum snippet length is below 70 characters it probably does not make sense to write descriptions that are substantially longer. Furthermore, you should avoid repeating or stuffing the description with keywords.
The default behavior of eSeller cloud is to show the contents of the meta keywords field in the meta keywords tag unless the field is empty in which case it will be omitted. Google does not use the Meta keywords tag and although the tag may be relevant in some very specialized cases for alternative Search Engines, we don’t recommend that you spend too much time worrying about the this tag.
It is possible to alter the default behavior of the field in content design
URLs are shown as an integral part of the result snippets and are important for searches. In eSeller Cloud there are 4 different types of URL’s:
- Product URLS which are automatically formed on the basis of the product name and the product id
- Product menu URLS which again are automatically formed on the basis of menu names
- CMS Pages where the URL are chosen by you
- Fixed URLS such as checkout which are of less SEO interest
Since product URLs are dynamically created based on the product name it is important to use relevant and meaningful product names. While product names should not be overly long it may be relevant to include manufacturer or brand information in the product name. In some use cases for instance for highly technical products it may also make sense to include important technical parameters (e.g. for drills it could make sense to include diameter and length)
When creating product URLs from product name, the following characters from the product name are used directly:
List of allowed characters in product URLs:
All other characters in the product name are replaced with the SEO Space separator (typically a hyphen) which is configured in the shop administration (All settings >> SEO tab) (Notice: Do not change the SEO Space separator without creating 301 redirects!)
Product URLs consists of:
It is often hard to categorize products into distinct menus which is why the same product can be placed in multiple menus. When this is done, the same product will have multiple URLs.
Consider the example of the product named “Bosch drill 7mm x8”. This product may be categorized under the menus Bosch (menu id 1) and Drills (menu id 2).
The product shown in the Bosch menu will link to:
While the same product under the Drills menu will link to
Google does not appreciate the same information or text on multiple pages (duplicate contents) and to avoid duplicate content, one menu is selected as the primary menu and all other URLs for the same product will automatically get a canonical reference to the product under the primary menu. – It is possible to explicitly specify the primary menu for a product, - if none is specified eSeller will automatically select the Menu with the lowest internal Id as primary menu.
When a product is renamed or moved to a different primary menu, the shop will automatically create a 301 redirect from the old product URL to the new product URL.
CMS Pages and product menus should be given names that reflect the contents of the page.
Googlebot scans html contents for links and uses the link for future indexing but also for “understanding” the contents of the linked page.
Googlebot uses the anchor text of a link to better understand the contents of the linked page and care should therefore be taken to write reasonably meaningful and unique anchor texts.
For instance when linking from a specific product (ie. A cell phone) to that products user manual one should strive to use the product name in the anchor text: “User manual for Nokia 3700”.
Anchor texts are important for internal and external links.
When linking to external web sites it is important to note that page rank flows from on page to another page via links. This means that the sites to which you link will be positively affected by your link. If you host message boards or other user generated contents you should add nofollow links to all external links in the user generated contents. Additionally, paid links or ads that contain links to external sites should use the nofollow attribute in order to comply with Googles TOS.
Image search is becoming an increasingly important part of searches and it is thus important that Google is able to index images and to refer a specific image to a specific product.
Google supplies at least 4 different ways of accomplishing this task
- Image alt text
- Image file names
- Rich snippets
There is no evidence or documentation to suggest that using multiple simultaneous methods is better than using one method.
eSeller Cloud supports marking up images with rich snippets and image alt texts. The product name is for the Alt texts while itemprop=”image” is added to all product images.
Rich snippet is a technology that allows google to understand the contents of a page and to show additional data in its results. This data could include review scores or product prices, shown directly in the search results.
As outlined earlier, Search Engines do not like duplicate contents and eSeller Cloud has a number of mechanisms to avoid it. This section briefly covers eSeller Clouds handling of duplicate content under different circumstances.
Products categorized under multiple menus
Products that are assigned to multiple categories will have an URL for each menu. To avoid duplicate contents eSeller Cloud will insert canonical tags pointing to the primary menu. If no primary menu is specified, eSeller Cloud will select the menu with the lowest internal id.
Items with additional query parameters
It is possible to attach parameters to any eSeller Cloud URL. Query parameters can be used to convey information from one page to another page. Googles gclid parameter which is attached to all Google Adwords urls is an example of a query parameter which is commonly used in eCommerce. This parameter conveys information from the SERP to the Google analytics script on the web shop.
Note: This functionality is only active if “UseCanonicalUrl” is enabled (Home > Base setup > Webshops > Shop > All settings > SEO)
The content of the above URL is identical to same URL without query parameters:
When eSeller Cloud is called with additional query parameters it will insert a canonical tag to the same URL without query parameters. This ensures that adding query parameters to a link does not result in duplicate content.
Variants are products that share the same eSeller Cloud product Id and that often have very similar traits. A t-shirt which is available in different sizes and colors is a good example of a product variant. In eSeller Cloud each variant combination have a unique URL and since variants often share the product description and other traits of the master product, this can result in duplicate content unless specifically handled.
eSeller Cloud will automatically setup canonical tags for all variants. The canonical tag of all child variants will point to the parent variant.
Note: This functionality is only active if “UseCanonicalUrlForVariants” is enabled (Home > Base setup > Webshops > Shop > All settings > SEO)
Things to avoid
Gaming the Search Engines
It is fairly trivial to come up with methods that will use false indicators to skew search engine ranking, these methods are often referred to as black hat methods. Although these methods may seem tempting it should be noted that Google is constantly updating their algorithm and that search engines actively punishes users of said methods. Penalties may range from lower rankings to complete removal from index. We strongly discourage any black hat methods or any SEO method which is not conducted in good faith.
- Link exchange and unnatural link building
- Showing content to googlebot which is intentionally misleading
- Showing false reviews or showing rich snippets that are intentionally misleading