Executive summary

This is a summary of Region Västra Götaland’s strategy for enterprise search. Discussing the overall points for orientation of the strategy’s contents and be able to take a decision on its establishment.

Vision and mission statement

The vision is that search “will deliver the right information when it is needed.” Not necessarily is the user forced to seek to find, but also to receive feeds and be notified if anything of importance occurs. The information need to become more intelligent. The business goal for search is primarily to avoid the risk of someone acting on outdated knowledge. Secondarily to offer a complete picture to the information, not to unnecessarily develop or purchase information or things that already exists somewhere within the organisation. We should strive to make the benefits of search known and the platform well-used to realise these goals.

Application requirements and usability

We will focus on user experience and work user-oriented. This is because search is there to help finding information and knowledge – even when there to the user’s not obvious where to find it, or that it exists. Two fundamental challenges to adhere to is that users have varying digital maturity and varying expertise in the material they are seeking. Therefore, search has to take advantage of both the novice and the expert way to express themselves, but also support the usage pattern that distinguishes the digital novice from the digital savvy. Search has to live up to high standards of accessibility and usability, both from a desktop computer and a more mobile scenario.

Management and organisation

Search should be continuously improved. By reflecting on operational goals, feedback from the users as well as by insights gained working with search analytics. In addition, the organisation shall have annual priorities of the activities to be carried out and follow up on the efforts made. The organisation intends to gradually move towards a more structured information architecture which naturally affects the curation of information sources. Also, it is important to provide opportunities for personalised search results, which, when appropriate, is also displaying access-restricted material.

The search strategy

About the strategy

This is a strategy for organisational search technology, commonly referred to as Enterprise Search. That is, the tools and systems that work together to search the organisation’s own information sources, while external sources can and should be searchable in a well-functioning internal search platform.

Vision

“Get the right information when it is needed.”

• “Get” means that the information reaches the user automatically through feeds, notifications or through manual search.
• With the “right information” the accuracy, validity and relevancy is implied.
• By “when it is needed” we pinpoint that the information reaches the user in the context the user is currently at and with the tools that are available (PC, mobile, etc.)

Business goals

The primary business case for an organisation’s internal search is to help with finding information and knowledge when there is not obvious to the user where to find it, or that it exists in the first place.

Search is the tool that is offered when editorial priority or lists of content does not solve the user’s specific problem. Since search has a unique overview of multiple sources of information it can also be used to convey information between different information systems.

Reasons why it is difficult to recoup investment in search mainly revolves around that its usage is seldom part of a controlled business process. Even if were, the primary metric is time saved. Time saved is not a particularly good indicator of success. In other words, it is difficult to develop Key Performance Indicators (KPI) that define the success of search.

Industry expertise highlights that the success factors for search is to avoid risks, such as the risk that someone is acting on outdated knowledge. This can be measured quantitatively.

Another example of business goal is to look historically what opportunities were not exploited fully because of inadequate information. Such can occur during major changes in the business environment or internal reorganisation. This can be measured qualitatively.

In other words, it is not enough that there is an enterprise-wide search. It must be known among those who need it but also perform well for those who choose to use it.

Common business goals for search is to simplify access to information from known sources. For instance by increasing the reuse of information and knowledge, and to enable collaboration by making it easy to find colleagues with relevant knowledge.

Enterprise search has to bring operational benefits in the following scenarios:

1. A user on a website performs a site search, hoping to find specific information.
2. An employee is searching in what is perceived as the organisation’s intranet, ie targeted information, published documents, etc.
3. A user need an overview of what multiple sources of information contains on a certain topic, the sources may be both internal to the organisation and external provided by its partners.
4. When a specific application’s internal search is not efficient, the organisation’s strategic search platform should to take its place.
5. When information systems need access to content already in search’s index. This might happen because it may be better structured than the original source.

Strive for consistency

A big advantage with enterprise search occurs when it is established, well known and have few competitors for the attention of the users. It is common to strive to replace accompanying search features of business applications. Examples of such can be document management systems, project management, intranets, and many more applications which main focus is not the information’s findability.

A pre-condition is that the enterprise search performs at least as well as the individual application’s inherent search it intends to replace. The feature refers not only to the user interface, but also how search is managed. For instance, if someone evaluates zero results, common search queries, what is actually clicked, etc.

In a large organisation, enterprise search need to comply with the search conventions users have already learnt, but perhaps also establish an complementary internal convention. It is important that the user interface for search is consistently designed, peculiar relevance or a user experience that is puzzling. At the same time, you need to offer a logical autonomy for the divisions of the business that has a good reason to differ. If there are good reasons to deviate, those need be considered wisely to not lose the big picture.

In other words, there are a number of inherent requirements for subsequent search capabilities, whether it is a new or specialised application based on the existing search platform or an entirely different search engine. Whether or not we complement our common search platform, the basic requirements are at least to be:

1. Collecting usage data comparable and compatible with the common search’s insights. Otherwise, one cannot conduct enterprise-wide search analytics and will little by little make search less impactful.
2. Adhere to the organisation’s design-conventions for search. Consider design patterns such as the aesthetics of query-completion and its behaviour to the color the search button has. The point is recognition, of the visual and functional. A user is not supposed to ponder who the sender is.

Requirements regarding usage

Search ought to be the primary option, regardless of whether the user is an expert or novice in the she’s looking for, whether the user is digital native or not.

Search must be responsive towards integration with other information systems that follows pragmatic standards, in order to support other systems that themselves is not capable of great search functionality.

To describe the users of search we can divide them into four principal groups – in a quadrant. Describing how familiar they are in the subject matter they are seeking, and how much experience they have of the opportunities digital technology offers.

The quadrant is inspired by the book Designing the Search Experience.

Search has to consider the double novice’s needs but also offer the efficiency double experts expect without worsen the double novice’s experience. To the greatest extent possible, avoid creating new specialised search interfaces in order to not cause confusion among users about “which search engine” to use when.

Examples of characteristics that distinguish subject matter experts and novices is what words they use when searching. The expert believes that they know what something is called and may have difficulty using layman’s terms. This behaviour requires that the terminology is supporting both the expert and the layman’s way of expressing themselves.

Examples of things that differ users based on their digital maturity is that the expert can take advantage of advanced tricks she has learned from other technical environments. For example, to use complex search queries or actively control acceptable misspelling. The novice’s needs rather are met with a introductory and educational way in which she is not caught off guard by too many possibilities at once, rather, she get contextual tips when they are meaningful and the benefits are obvious.

Four modes when searching for information

1. Known knowledge. To search for already known knowledge is easy to understand, because the user:
   • Know what they want.
   • Can express what they want.
   • Have an idea of ​​where to start looking
2. Exploring. When in this mode, the user has an idea of ​​what it want to know. But the user may have difficulty expressing it, or cannot use the correct terms. The user often know when it has found the right content, but has no knowledge if the amount of information is sufficient.
3. Do not know what they need. The key concept behind this condition is that users often do not know exactly what they need to know. They may believe that they need to know one thing, when in reality it is something else. Sometimes, they visit an information source without any specific purpose.
4. Retrieve. In this mode, the user is looking for information that it has prior knowledge of. They can remember where they saw it recently, which source of information it were or they have an idea where to find the content.

Some common use cases for enterprise search is to:

• Find a specific document.
• Finding specific skills / roles / persons.
• Quickly become an expert on a topic-
• Find out everything that the organisation knows about a topic-

Usability and accessibility

To ensure good accessibility and usability search’s user interfaces has to be developed in accordance to the accessibility guideline WCAG 2.0 at least at level AA. In addition to this, user interfaces should at the very least comply with priority 1 of webbriktlinjer.se and never deliberately deviate from their guidelines of secondary priority.

Design principle to adhere to is mobile first since those connecting with a mobile device should not have a worse experience.

In most cases, when search highlights or present structured data it shall also, technically, be disclosed in a structured manner. Such as described by schema.org, where content is marked up with the appropriate level of structure.

Extremely good performance must be delivered to the user. Measured according to Google Pagespeed ​​Insights, from a mobile user scenario, never less than 80 / 100, as well as for desktop computers never below 85 / 100, unless stricter requirements exists within the organisation.

Management and governance

The management of search is important and since it’s often regarded as both everyone’s and no one’s concern it’s not easy to manage. Managing search requires a range of skills which does not make things easier. Here are the most strategically important to partake in the management.

Key figures

The groups identified below are those that have the greatest impact on search’s continuous management and development to provide the conditions for good business and user benefits.

Those who budget

A search platform need ongoing development and adjustments to deliver what users expect. To constantly develop to meet the changing user needs requires that resources are allocated accordingly.

Content creators

Everyone that may affect the searchable content are within the primary group interest. The most easily identified group is the web editors, working with content on the intranet and external websites.

Another such group is people who produce documents, such as word processing, spreadsheets or presentations. Documents are often searchable either from the website’s upload folder or document management system indexed by the search platform.

This group of key people need to have an understanding of how the results of their work affects how search is either successful or failing.

Content curators

Search is in great need of an orderly process for informatics in the organisation. The reason for this is that search is user-driven. Because the users themselves are phrasing their search queries in hope that it will deliver good results.

Search is dependent on us working with informatics – at least to some extent – for subject matter experts to use their terminology, without excluding the related information created by the laity, and vice versa.

Those who support the search platform

To not be reckless on technology-driven initiatives, a business representative (business owner of search) is to control the activity plan and its priorities yearly. Connecting the activity plan, budget and the quality of delivery to the business. It should be clear what benefits were gained.

This action plan should then be monitored and events are regularly reported to the business owner of search.

Those working as search editors need acceptance from line managers for their work with editorial tasks – manually controlling the results, deciding on keymatches, improving important metadata on others’ content, working with search’s instructional texts, and more.

It has been shown in studies that search is more appreciated if there is someone responding to user questions and managing feedback channels. This need to be formalised so users are receiving help or feedback in a reasonable time.

To have a good technical availability we need a dedicated technical support and an active IT operations, also that servers work satisfactorily.

Organisation

The most important effort an organisation can do to improve its search is to appoint a owner of search! It is an absolute minimum requirement.

This means that a owner of search must have time set aside to work with search. A few hours a week is much better than nothing. And even more important: to work with search is a long-term work, certainly not a project.

The roles and competencies in search’s management should consist of:

• (Business) owner of search
• Search technician
• Search editor and/or Information specialist
• Search analyst
• Search support

Further reading: Enterprise Search Team Management by Martin White.

Evaluation of search

Search fill its purpose when it deliver the right information, is fast about it and always available. To satisfy these requirements, the function of search is to be tested regularly and tests should be documented in test plans. Below are some of the tests that are appropriate:

• Search loads quickly, tested with Google Pagespeed Insights, with a minimum of 80/100.
• The response time of a query should be about 0.1 seconds, but never longer than 1 second, measured at the user interface.
• Search will be available 24/7 (around the clock seven days a week). Monitored by, for instance, Pingdom or Uptimerobot.
• Size of search indexes. Among other things, to see if more or fewer documents are indexed, which can provide warning signs in advance, help being proactive.
• Search’s user interfaces are accessible, tested with the W3C Validator.
• Search’s user interfaces are usable, tested against webbriktlinjer.se and W3C:s WCAG 2.0 at level AA.
• Survey the satisfaction of users.
• Reviewing search statistics and/or performing search analytics, to gain insight into how users are searching.

Search analytics

Search statistics must be collected and need to be continuously analysed to find out what works and what does not. Search analytics is used to improve search results. At the very least, frequently analyse zero-results and the two hundred most common search phrases.

By analysing the search queries that did not give any result at all (zero results) it is possible to identify what content is missing, find synonyms to use, understand which abbreviations are used and discover alternative spellings.

The two hundred most common search queries need to continually be reviewed, probably manually, to get an insight into how the experience of search is for a large part of the users. And also, if the relevance model can be improved and what content is most in demand.

Helpdesk and support

User support (1st and 2nd line support) should be handled directly by the search editor(s) and the owner of search. The reason for this is to give the search management direct contact with users and thereby gain a better understanding of what works and what does not. Users should be able to directly report their perceived problems with search in general and search results through a feedback form in connection with the search interface.

The technical support is handled like any other system.

Training

Probably everyone who use search are in need of some form of training in the offered features. At least the following user training needs to be actively disseminated and be available when needed:

• All users need to understand how search works and be able to supplement their knowledge with new handy tricks. An educational approach is to let users be activated by taking advantage of a interactive guide, where they “practice” their new skills while they seek. This interactive help in the search interface can be supplemented with links to more extensive documentation in the form of text, audio or video, for those interested.
• Content creators need to understand how they markup the information properly. It is worth to introduce an organisational quality quota regarding metadata which at least would be mandatory for professional content creators and communication professionals. In addition to this, the search technology ought to make it easier to get feedback on content’s potential findability directly when saving it – regardless of which system the creator is using.

Advantageously, a large part of the training content and documentation is available on the intranet. Also, use screen recordings to display operations that involves several interactions.

Search workshops – assistance on the user’s own premises

Search workshops should be organised regularly, a meetup where both content creators and users can met to get help and learn more about search. A search workshop can be held in a conference room, an afternoon a month.

In addition to attend this meetup, it should at the same time, for the convenience, also be possible to participate remotely. With screen sharing, voice calls or chats, remote user’s should be offered a similar support.

Information

The most important for a properly functioning search is the quality of information indexed and that the organisation has a good information hygiene. Good information hygiene pinpoints that old and out of date information is regularly culled following established procedures. The procedure is the lifecycle management of information, sometimes referred to as “the content strategy”. In public organisations the lifecycle / strategy is often documented in a so-called document management plan.

The paradox of search is that large amounts of information potentially offers great business benefits, but also hampers the ability for search to be precise – to find the right information.

Lifecycle management of information is very important for good findability. A prerequisite for findability is to avoid ROT, that is, content that is Redundant, Outdated or Trivial.

In practice this means that information available need to be evaluated in accordance with established principles, which can be summarised in the question: Should the information be archived, deleted or updated?

By categorising all the information it’s easier to decide how information should be handled. Categorisation is dictating the information’s lifetime. A guideline may be considered valid for 10 years, while a press release is archived in six months. Event and other time-based content can be archived after it occurred, and so on.

To get an overview of the sources that are important for the organisation to be searchable, a list of sources need to be compiled and maintained. The list should contain the already indexed sources, and also a priority of the sources of information to be indexed in the future.

Metadata and thesaurus

As part of a successful information management, there must be a list of the organisation’s common concepts and terms. A common ground on terms contribute to good findability. This is called different things, depending on how far the work has progressed: glossary, vocabulary, nomenclature, terminology, taxonomy or thesaurus. These concepts must be documented, made available and used by all sources of information in the organisation.

Metadata

By establishing a specification for metadata, both categorisation and use of a common terminology is facilitated. If categorisation and annotation of information is consistent across all information sources indexing becomes easier and findability is improved greatly. The use of metadata ensures good findability. It is advised to require a minimum of metadata in each source of information, for example the following metadata for each document / page / information service:

• A title.
• Description of the contents.
• A number of descriptive keywords.
• Some timestamps, highlighting the content’s lifecycle, such as when it was created, published, updated, revised and finally possibly archived.
• Status of availability. For example, if it is public, access-controlled, valid, outdated, archived, etc.
• Its canonical address. That is the original and primary URL. A so-called Canonical URL .

By using the schema.org specification and similar standards, metadata requirements can be met.

Design

Design is important. Especially today, when most people refer to Google as their point of reference when it comes to how a search should look and behave. It is important that search behaves similarly to established design conventions – for the convenience and efficiency of users who have already learned certain skills to get the job done. Of course, search’s design patterns and design principles cannot differ from users’ preconceptions. The most important thing is that the search function design provides value for the users and business.

When some change affects search’s design, a usability professional have the final say in how the design will end up.

Example of a design prototype created for VGR, person-specific search , a keymatch and the wizard to introduce the new search-GUI .

Technology

The infrastructure of search should be handled by the organisation and is one of the organisation’s strategic services.

The technical platform for search is a very important part of a well-functioning search. It should be emphasized that all the technical platforms for search (commonly referred to as “search engines”) that is currently on the market are good enough for most of us. The choice of platform(s) of search is an important strategic one. The choice should reasonably be affected by economic, technical and operating concerns.

The technical infrastructure of search cannot only consist of a search platform, but also a well-functioning integration of supporting services, at least:

• Centralized metadata management that supports the users’ language – whether it’s layman terminology or not – to connect the expert’s terminology with the novice. In addition to this you need to manage the unstructured and more agile through an orderly folksonomy management – possibly linking it to controlled vocabularies to bridge the gap between different users. The metadata management should also provide support to other information systems, or directly to the content creator, to provide up to date feedback on how findable the information will be once it is published.
• Master Data Management (MDM), Product Information Management (PIM) and reference data in order to identify the business’ physical and digital items from the abundance of information available in search.
• Directories to find people, traverse the structure of the organisation and other commonly used directories, both internal and external. Most obvious is that search should be able to compile a complete profile of every employee regardless of whether the complete picture is scattered in several specialised systems.

Collecting the information

When chosing a strategic platform for search, the range of data-connectors is one of the most important things to consider. The question to ask is: is there connectors for the organisation’s most important sources of information? The collection of information is resource intensive and to configure and develop custom connectors can be expensive, error-prone and difficult.

A connector’s purpose is to fetch (or receive) information from sources of interest for the search platform. Each connector is tailored to a specific type of information, such as databases (one connector for each different database type), file system, the Web, business applications, etc.

One method to gather information for the search index is for a crawler to visit the source of the information and see if anything has changed since the last visit. This is called crawling, a method of pulling content towards indexing. The disadvantage of this type of indexing is that it takes longer to notice recently created or updated information, making the search index less timely.

Another method of collecting information to search’s index is when the information source informs search when something have been changed or added. So-called push-driven indexing. The advantage of push indexing is that recent information quickly becomes searchable. It can be difficult to implement, though.

The perhaps most strategic solution is a combination of both push and pull. That information sources offers sitemaps, according the sitemap protocol. Search’s connector inspects the sitemap to determine what should be indexed, then the crawling connector does not need to visit all the information to find out if there is any new or modified.

Federated search – to search in several search applications simultaneously

The reason that not all information is searchable in a single search index is often about rights to source material, sometimes there are regulatory restrictions. For example, commercial journal repositories, research databases, etc. Another complication might be an individual user’s desktop search. With desktop search the material the user has stored locally on their machine, smartphone or whatever is perceived to be right by the user’s fingertips.

The search results of a federated search can be displayed in various ways. Often the results are displayed side by side, visualising that there are multiple repositories contributing.

Federated search is difficult to implement in a usable manner and the recommendation is to refrain for the time being, at least until very clearly business needs and goals allows for evaluating the potential benefits in an isolated pilot.

Security

Being able to search for access-controlled materials is important! There are two main principles for how search can make access-controlled materials presence known (for users without the necessary authorisation):

1. 1. The search result displays a content summary, that is, a normal search result, even the restricted information, for all users. To access more than the summary, the user must provide the authorisation needed. If the user does not have access to the information, she’s granted no access. This is suitable for using search for exploration, to make users aware that there is information on the subject.
2. 2. Search only shows results that the user has access to according to her authorisation. This is easier to implement.

No matter which of the principles chosen, the authorisation system should be based on the same directory service controlling each and everyone’s authorisation. Also, the roles need to be described in the same way in all the information systems. The reason is that the documents have to be accompanied by an access control list (ACL) that controls who is granted access to the information. Otherwise, the access control is certainly complicated to implement.

If there is no common directory service, avoid indexing information that need access-control until the users’ permissions and roles are widely implemented.

License – CC-BY-SA

This material has been produced by the Region Västra Götaland (VGR) in collaboration with several other public actors. It is not considered to be finished. We disseminate this information to interact with external stakeholders. The material is licensed under the CC-BY-SA , which among other things allows for further use.

Translation made by Marcus Österberg.

The post Example: Strategy for Enterprise Search appeared first on Web Strategy for Everyone.

This material was left over when I wrote the book Web Strategy for Everyone. Rather than throwing it away, it is now published with some editing – though not quite as ambitious as it would’ve been in print

Search engine optimization is all about optimizing one’s web presence for search engines to take a liking to your site. It is often abbreviated SEO. Something quite different is keyword optimization – the editorial part of the process of SEO.

Search Engine Optimization ≈ Make your website’s content great according to search engines

Search engine optimization, also known in short as SEO, is your attempts to attract visitors to your website by making your site as good as possible in relation to search engines and their users. It is possible to divide SEO in lots of niche terms, but I will restrict myself to make a difference between search engine optimization and keyword optimization because, in its basic form, it’s about either technology or linguistic. This post will therefore not take up the subject about search engine marketing (SEM).

Search engine optimization is basically to have great content. What‚ then, is good content? The following features are a good start for the content you wish to attract users to:

1. Unique – same content is not available on other sites (such as through syndication / RSS, etc.).
2. Suitable length – short product texts are regarded as facts.
3. Appreciated by others – has links to itself, and many visitors (Google has access to your visitor statistics if you have Google Analytics).
4. Posted on a website that has subject-related authority – to more information about similar things are.
5. From a trustworthy source – the author is an authority in the hen write about?
6. Accessible – well put, not long chunks of text, and preferably marked up with RDFa/Microdata techniques such as Schema.org so even machines understand what the content is about.

Search Engine Land has an excellent list of the SEO’s periodic system, where you can see more positive factors.

There are also bad signals, among other things:

• Many different authors with few posts per site. At least Google does not like guest posts or suspected sponsored posts. Many writers can be an indication of a suspicious behavior.
• Outlinks to pages with low credibility, or links that lead to error 404, or hacked websites.
• Too many links. Offering hundreds of links in your content exhibit an inability on priority, everything cannot be a priority.
• Slow page views. Both search engines and users have better things to do than to wait for a slow website.

Sitemaps

The structure of the website needs to be good enough for a crawler to be able to look around, but there is still a point to submit a sitemap to the search engines. This is done by Google through their tools Search Console and Bing with it is their Webmaster Tools that apply.

SEO best practice changes over time

A common mistake among those who do not work with SEO very much is to believe that it does not change very much – it does, a lot! We notice that especially more and more. For instance on the books released about SEO that nowadays are named with the year of release, and the authors trying to delist their books a couple of years after they are released. This is the reason why I did not include a chapter on SEO in my book Web Strategy for Everyone, the subject-matter is to agile for print.

A common misconception is that it is beneficial to enter keywords, so-called meta-keywords. It is still widespread, but there are many more examples that it is important to keep on top of the updates in SEO.

The reason is attempts to spam Google…

For example, Google is fighting a constant battle against search engine spammers attempting to lure in visitors. The problem Google is trying to solve is to display only relevant hits from the original source of that which is sought after. For instance, it’s not supposed to be meaningful to build a site which largely contain materials snatched from Wikipedia, then Wikipedia is to be displayed in the list.

If you want to learn more about these updates, you can visit SEO forums, or search for animals names Panda and Penguin which were the first two major upheavals in SEO. While you’re out there and googling you can also read about the Google algorithm Hummingbird, its aim is to understand a search query’s context and provide sensible synonyms and related searches. This is what the book Web Strategy for Everyone referred to as Web 3.0, or, the Semantic Web. An attempt to make machines understand content and context, and to bridge the different data sources differences in structure.

Everyone needs to know some SEO!

The reason why all website owners, editors and bloggers need to learn a lot about search engine optimization is that otherwise they avoid customers and visitors. Compared with the physical reality a optimized site shop is placed on the street where its customers are frequently visiting. The website that is not optimized at all is like having a department store on top of Mount Everest, surely a cool concept store, but there are not many customers in the vicinity.

Keyword optimization is to make your website attractive to search engine’s and to clearly position the keywords your audience uses so the website will be easy to find. Besides some initial technical measures it is eventually those who handle the content the website shows up who have the greatest impact on the site, making for it to be keyword optimized.

Somewhat simplified, one can say that only text is searchable because it is difficult for search engine’s software to understand the content of image or video without it offering descriptive text. Although the machines are getting better at understanding images, video and speech, it will be a while before we can take a pass on the production of texts.

Glossary of SEO

• Keyword Optimization, SEO (Search Engine Optimization) – The activity and know-how on how to improve websites to be easily found through a search engine.
• Search querys & keywords – Usually one or more words that a user types into a search engine in the hope of finding something relevant.
• Search results & SERP (Search Engine Results Page) – SERP is often used by those already initiated into the search engine industry and construed as Search Engine Results Page. On the Swedish search results page, that page of the search engine in response to a search term.
• Visitors – When referring to a visitor of a website, it’s about all the visits provided. Much like the visitors to a business, some are recurring others are not.
• Unique visitors – A unique visitor is different from a visitor by that we are able to identify them as an unique individual, able to perform multiple visits to a website.
• Page views – Number of times a visitor viewed a page on the site, or the number of views for a certain page.
• Unique page views – Number of unique visitors who viewed a page. For example, it is not unusual for a visitor to visit a single page multiple times, which is a single unique page view.
• Index – The search engine’s index is what it knows about and what its users can search within.
• Bot, crawler and spider – Is the search engine software searches the web for new, changed, and deleted material as pages, images and other types of documents.
• Inbound links – the links pointing to your site. So other sites that link to yours.
• Keyword Density – The percentage of text that consists of relevant keywords.

Editorial keyword optimization

Wording and editorial

The visitors you want to reach are those who are actively looking for something your website offers. Therefore, it is very important that you are familiar with how your audience express themselves and what search queries they use.
Are they looking for ‘gore-tex hiking boots’ or ‘The North Face Hedgehog GTX 2016 edition’? While both search queries perhaps are describing the same product, someone searching for one of them does not necessarily get to see the other one in their search results.

It is paramount to be aware of the words used so there is no risk that extremely rarely get a match between the words you have chosen to use and a search engine user accidentally enter to search.

Usually it is a combination of several possible words that constitute a search term. Try to naturally get the most commonly used word for each sub-page of a product. Later in this section is useful tools like Google’s keyword tool you can use to find out how many people use certain search criteria.

Heading! = Headline

Keep in mind that there is a big difference in a heading and a headline. A heading is more of a name for something while a headline is descriptive of what you can expect from the content.

Sample heading:
Tale of Little Red Riding Hood and the Wolf

Sample headline:
Wolf ate grandmother

Of the two examples above, only the last one is wise to use a large extent on the Web. The heading indeed talks about the kind of information – a fairy tale – and who is involved – Little Red Riding Hood and someone / something called the wolf – but it summarizes in no way what is to come. The point of instead writing that a wolf ate grandma is to summarize all the text has to tell. This way of obfuscating the content is all too common.

Writing for the Web is to write what is most important first. A descriptive title, a summary preamble when necessary and that the most important in the body comes first and then go into the details and references to learn more.

Writing page titles

The page title is the name of the page and is often reflected in the top of the browser window (at least on computers). The page title is also the clickable text in the search engine’s result page and the name in a browser’s bookmarks.

Often, the page title is the same as you choose to name the webpage. In some web content management systems you can manually write whatever you want for the page title and some other phrase for the main header of the page.

There are two common variations of how page titles are written. What divides them is whether the site’s name comes first and the unique in the page’s name comes first. To be unnecessarily obvious, two examples below:

1. Weekend trip to Prague – Travel Company Inc
2. Travel Company Inc – Weekend trip to Prague

In example 1 above, the unique information comes first, which means that the words weekend trip to Prague is more keyword optimized than the company name.

Example 2 focuses on to clearly tell who the owner of the site is – at the expense of the actual content of the individual page.

Before choosing the syntax of example 2 it is good to think through whether it is worthwhile to compete to appear on one’s brand. It is often enough the case that users are looking for what a business does or offers rather than the company itself. Moreover, it is common to be ranked really high in search engines if users are searching for one’s brand. So I would absolutely recommend example 1.

In fact, some take this to such an extent that they only write the page name in the page title. After all, there are URL nearby on a search engine and it usually specify who the sender is.

It is worth bearing in mind that the length of a page title, if possible, should be a maximum of 70 characters. You will not get much more space in the search engine result page and, also, it is probably not that common to read much more than the first few words. Yes, users are skimming, almost all the time on the Web.

Headlines

In order to clearly divide a body of text and have a good readability, we as readers need information structure. For this we use headlines that describe the body of text below the headline. It is not enough to format a header with bold large text – the headlines must have the correct HTML code to remove all doubt that it is a headline. The same applies to the headings.

If possible, it is recommended that all important subpages has a headline and at least one sub-heading of level 2. These titles will come in the right order, sizes, in descending order, without any level is missing.

If I had a dime for every time a web editor explained why they did not use the correct heading levels, or real headlines at all, it would at least paid for a family pizza. The most common argument is that the text becomes too large, the different color, change the font or is in uppercase. It is a problem they should rather take up with their web developers.

Words in a headline are worth more than those in the subsequent body text. This is because it is more prominent. Actually, it’s logical. When sighted, blind and machines skim the text through the headlines as they act as entrances to the text which is between headings.

Alternative text for images (and other media files)

Alternative texts are for describing media files content for those who cannot perceive the content, such as those with deficit sight, the blind, deaf – and that includes search engines!

Putting alternative text on images is really for the visually impaired and those not loading pictures. The text is meant to explain the image, which for those not loading pictures can explain which of the pictures are worth to load at all.

The reason why some choose not to load images in their browsers are often due to a dubious internet connection, or perhaps that it costs money on most cellular connections, or to avoid advertising.

Although alternative texts will be used primarily to explain the images’ content for those who cannot take part of the image you have the possibility to use other terms when writing those almost invisible phrases. Consider using synonyms and suitable words that is not present in the body.

Body text

That a page contains more than images is important because otherwise there is no text to search for. It is considered by many SEO professionals that there is a critical minimum amount of text – which of course is difficult to put a static figure on. Then there is the problem search engines are increasingly trying to avoid the so-called content farms. A content farm is a website containing pages with either very targeted text and commonly content that is not unique, for example snatched from Wikipedia.

In addition, the body is the ultimate place to have a good breadth of searchable keywords. You can also put some effort with something called keyword density, thus using the same words many times over.

However, website copy is for the benefit of the reader in the first place. It ought to be a legible and comprehensible text, to please the search engines is secondary. So, when you achieved legibility you can care about SEO. Such as the body text being at least 300 words, a level that is constantly raised according to most SEO in recent years.

Bullet points and numbered lists

Words that appear in a list on a web page is considered to be of greater value than that which lies in the body text, at least compared to text in ordinary paragraphs. Besides lists visually guides the reader in a body of text – regardless of whether the person uses a screen reader or not – also for search engines lists stand out, often with containing relevant facts or summaries. This for the simple reason that lists visually stand out from the other content.

If you have a text that lists things also format the content as an ordered or unordered list. It also makes text easier to absorb, and lists are a great way to not scare the visitor away with a massive wall of unstructured body text.

Inbound links and URLs

Having many quality inbound links, especially from known or trusted sites, is important to show that the website’s content is worthwhile linking to.

Many tend to register their sites in different link directories, exchange links on webmaster forums or email their partners to get links. To a limited extent, it is still worthwhile to get the website listed in these contexts. But remember that words in the clickable text of the link is associated with the site. Therefore, send a thoughtful suggestions so the anchor text don’t become ‘Check their site here …’

Descriptive URLs are also relevant in SEO, only since they sometimes become the clickable text when some post the URL in forums, or commenting on the Web. Not to mention that it is good for those who stumble across an URL to be able to figure out what is referred to.

Example of a descriptive URL, also called “friendly URL”:
webstrategyforeveryone.com/performance-for-intranets/

Example to the contrary:
www.gp.se/sport/1.565873

The addresses of your pages are worth trying to keep as brief as possible – but still descriptive. This is something that many Web Content Management systems (WCMs) struggle with. Such as the URL syntax in Episerver, which my employer, Region Västra Götaland, uses, becomes absurdly long sometimes:

www.vgregion.se/en/Vastra-Gotalandsregionen/Home/
Healthcare/Public-Health-/Public-health-policy-Vastra-Gotaland/
Living-conditions-based-on-equity-and-equal-opportunities/
Region-Vastra-Gotaland-is-developing-a-system-to-monitor-health-
inequalities-/

Try typing that correctly when read to you over the phone…

Some addresses are so long they become uncomprehensible. Also there is the risk they do not work when sending e-mail, or when posted in forums since many system cuts “words” that are too long and inserts spaces (in order not to break their design).

Website speed – Web Performance Optimization (WPO)

That a website is loaded swiftly has become a more and more important argument in recent years. Above is a picture from Google’s tool Search Console, one of many ways to keep track of your websites’ performance. In the example, it takes an average of 0.61 seconds to load a page which is not great but also not alarmingly bad.

The speed of a website is one of the factors that influence whether a website is ranked high on search engines. Therefore the website speed is not only an usability metric, considering users converts to a higher degree on fast websites, but also since it’s one of the ranking factors of Google since a couple of years.

If the website has poor performance it is commonly one of the reasons below:

• Slow web host. Is the site on the website host’s budget class? Then you cannot  expect particularly good performance – at least not for a website based on a content management system like WordPress, etc.
• Poorly optimized website. WordPress is among the most common systems nowadays. The problem with WordPress among others is the ease of adding plugins that negatively affects performance. Often the website get an additional stylesheet and an extra Javascript file. Also, the plugin is of varying quality. Evaluate whether there are better versions of the plugins you need.
• Too many, or too big, files. To upload lots of pictures of high quality and have lots of Javascript features affect performance significantly. A user usually cannot download more than 3 files at once from your website. If you have more than three files in total, a queue is there for the remaining files. That queue is fairly long on most websites, unfortunately.

A lot of tricks and my own experiences regarding performance optimization is available in Web Strategy for Everyone.

Common solutions to performance problems are:

• Optimize images so they do not take as long to download. An image file size can be optimized in Photoshop, for example using its feature ‘Save for the web’, using applications such as Imageoptim for macOs, FileOptimizer for Windows or web services as Smush.it.
• Ask the web developer to combine stylesheets and Javascript into as few files as possible. The Javascript library Jquery is often used on websites. Instead of having the file on your own website, you could use Google’s so-called CDN (Content Delivery Network). If you have many Javascript files you can look for CDNs offering to serve your unique files as well. This is often free, perhaps you’ll like Cloudflare (as much as I do).
• Change web hosting. Often small hosting companies are faster than the big ones for about the same cost. How come is the topic for another blog post, though. If you have plenty of money for your web project, maybe consider a Virtual Private Server (VPS), or even a dedicated server. If so, this is the time to talk to web developers or involve an experienced web consultant. If you’d like my advice just post a comment to this post.
• Evaluate the code behind the website and review any plugins. Not infrequently, there is code that can be improved and other plugins that provides better performance.

Meta description of pages

The description, or ‘meta description‘ for the technically minded, is a, to the browsing user, hidden text that briefly summarize what the webpage is about. The meta description should always be unique to each page of the website, otherwise there is no point in having them.

The picture above shows one of the few cases when a user can view the meta description text. Next to the page title when showcased on a search engine’s results page, its SERP (Search Engine Results Page). The meta description is supposed to summarize the content of a webpage and need to be well-written to attract the users to your website.

The meta description is placed in the source code between HTML tags <head> and </head>. It may look like this:

<meta name="description" content="List for those who want 
learn the latest on substance X. Here you will find advice and
tips from the professionals X. "/>

If you’d like to see what your competitors are using for their meta description texts, choose to view the source code for each webpage, and early in the code, you will find something like the above.

Keywords (is really not that important in external SEO – the last 15 years)

Keywords was once important to reach out in the search engines. They are still used today in some search engines. But because of the proportion who use Google as their search engine, and since Google does not care about keywords, it is not always worth the effort to work with keywords on your website.

The reason that Google, and probably other search engines, do not care about keywords is since its ease of manipulating the keywords in a massive scale. Spamming search engines with keywords have simply led to them not considering its use any longer.

If you still choose to work with keywords (mainly for your own internal search engine) you are recommended not to put all general words on all pages. Try to keep it unique for what each page contains.

For large organizations that have their own search engine, or a internal site search, keywords can definitely be worthwhile to be used on their own search engine.

If you’re looking for the keywords in the HTML code, look between <head> and </head> like in the code below:

<meta name="keywords" content="keyword1, keyword2, etc." />

Validation

The website’s underlying code is supposed to validate the code standard it is claiming to adhere to. Anything else is sloppy coding. However, all mishaps are not critical to how well your site performs in the search engines. Still, everything that is easy to fix (or deemed serious) is good practice to fix. If your website’s code is not even close to validate it can prove difficult to interpret for search engines and other machines you rely upon.

To find out if a website validates according to the web standard, you can go to the web standards organization W3C and enter the address you want to check – validator.w3.org

Useful tools and methods

Create a list of priorities words and search terms

Open a spreadsheet and enter the keywords you think are relevant to your website. Also, rank them in their relative importance and how descriptive they are.

Using Google Keyword Planner, see below, you can get help and insight regarding accompanying words used and see how much volume the search terms have.

Google Keyword Planner

With this tool you can enter the keywords you want to find accompanying words and see the search volume for the search term. Really good when you need guidance on the wordings and expressions to choose when writing web copy.

Keyword Planner is part of the Google Adwords platform, but it does not require an account or active keyword campaign to dig into the keywords. Try Google Keyword Planner ›

Plugins for your web browser

There are lots of browser plugins, at least for Firefox, that help you out with search engine optimization. Two of the most appreciated for novices is SEO Doctor and SenSEO.

With SEO Doctor, Firefox can review every webpage you browse, even your competitor’s webpages. Clicking to open the SEO Doctor plugin shows a dialogue with details explaining the score. You also get hands-on guidance on what can be improved with the individual page, such as including a sub-heading or meta description for instance.

SenSEO is advantageously used to evaluate how a particular word is optimized on an individual webpage.

In other words, when you have found a word you want to rank better in the search engines is then SEO tool to see what can be changed.

Performance Measurement

There are many online services that assist in analyzing how quickly a website can present itself to the users. If you are not inclined in web tech, it is worth talking to a web developer to get some perspective on the reports.

One of the services is WebPagetest that looks like the picture above. Another is Google Pagespeed Insights, another is Sitespeed.io for those feeling a little more technically minded.

Validation

W3C is the organization that writes the recommendations for many web technologies. They offer a service to see how well a website adheres to the standards they’re claiming to follow.

Above is how well GP.se adheres to the code standard they chose. Unfortunately, not that flattering. Check out your own website with the W3C ›

Checklist for editorial SEO

If your should you make a checklist about SEO for your web editors, or when you write copy for your website, then check out the below suggestions.

Below is a checklist in order of priority, for which effect can be achieved through SEO.

1. The page title contains relevant keywords and is shorter than 70 characters.
2. The main header (H1) contain important (and possibly unique / new) keywords, or supplemental synonyms the website needs.
3. Subheaders (H2) contains relevant, or use supplemental, keywords.
4. Thoughtful keywords are included in links’ anchor text on other websites, those who link to your website.
5. If there are images, they have suitable alternative texts.
6. Content that is suitable for lists are placed in lists (UL / OL).
7. Important keywords are one or more times in the body, preferably also early in the text. Ideally some synonyms as well later on in the text.

More on Search Engine Optimization

• Search Engine Land on Google RankBrain
• Google webmaster guidelines
• Yahoo! Content quality guidelines
• Bing webmaster guidelines
• Search Engine Optimization according to Wikipedia

Also, check out the book Web Strategy for Everyone ›

The post SEO – an introduction to Search Engine Optimization appeared first on Web Strategy for Everyone.

You can lay you hands on the e-book today, it is sent to you shortly after you check out your order. The printed book, though, will take a few weeks before it is sent. The e-book’s formats are ePub, PDF and Mobi, they cover virtually all mobile phones and tablets, and Mobi is specifically for you with an Amazon Kindle. If you’re looking for a different format I can recommend that you download the program Calibre, then you can convert to many more formats yourself.

The best offer is, I think, to purchase both the e-book and the printed book. It costs about 25 $ + VAT. Then you get 90% discount on the e-book.

Order the book at Intranätverk – from 10 $ + VAT ›

What is Web Strategy for Everyone covering?

It is 60,000 words and 80 pictures spread over 212 pages. The amount of images and the layout makes it not so burdensome to read, like many other similar books. At the same time it is written in a very ”condensed” manner and to the point, so it’s like most Swedish literature – rarely repeating itself over and over again. Or, as one reader put it:

”Recently read your book and it is an achievement how much value you have managed to put between the covers. Very good!”

– Håkan Liljeqvist, founder at Kreejt

The subject of the book may seem obvious given the title, but at the same time there is no real definition for what is meant by ’web strategy’. So I chose to make the book to cover the fundamentals in many areas a web strategist must know about. Partly, the Web’s history, a lot on information architecture, different approaches to web design, some about the increasingly hyped topic of web performance and last a do-it-yourself with hygiene factors to check on your own website.

Not just a translation from the Swedish original

The book is not really just a translation of my Swedish book. The English book has more international image examples and it turned out to be unexpectedly many cultural references that non-Swedes would hardly understand. So pretty much of the book is rewritten for an international audience.

Interestingly, several Swedish friends and acquaintances are waiting to buy the English book, despite the fact that the second edition of the Swedish book have been out for several months. Many Swedes, myself included, are probably more used to read in English and may have more benefit of an English version when in a multilingual business environment.

Better English in the book compared to what I write in the blog

A first turn of the translation work was done by a Englishman who is very knowledgeable in web development and intranets. After that, a language agency also had their way with the book to raise the quality even further.

Needless to say, my publisher has really invested in the book and I think it’s going to be great even for you guys having English as you with English or American as your native language. For obvious reasons we can not afford to make the same effort with every blog post I write in this blog, so do not be too quick to review its language if you find peculiarities in my blog posts.

Table of Contents

Since we’re not going to judge a book by it’s cover over the Internet I’ll share the ToC. Perhaps this content gives you an idea if you’d make use of reading the book.

• Before we begin
  • Why you should read this book
  • About me
• The Web’s history and future
  • Web 1.0 – a network of documents
  • Characteristics of Web 1.0
  • Web design 1.0
  • Web 2.0 – the engaging web
  • Characteristics of Web 2.0
  • Web design 2.0
  • Web 3.0 – a network of data (also known as the semantic web)
  • Characteristics of Web 3.0
  • Web design 3.0
• Information architecture
  • Content choreography
  • Examples of poor content choreography
  • Master Data Management prevents unnecessary duplication
  • The importance of marking up information with metadata
  • Metadata specification makes your data more standardized and interchangeable
  • Controlled vocabulary
  • Folksonomy
  • Architecture using APIs and open data
  • Public APIs, open data and the PSI Act
  • Background to the European Union’s PSI Act
  • Some take issue with the PSI Act – cumbersome access to data
  • What then is open data?
  • The benefits of an API for a startup business or when building anew
  • Design a public API with the developers’ experience in mind
  • Friendly terms and a free license
  • No surprising the developers with unforeseen breaking changes
  • Provide data in the expected format and in suitable bundles
  • Error handling and dimensioning of the service
  • Provide code samples and showcase success stories
  • Promote via data markets and API directories
  • What is the quality of data needed?
  • Microdata – semantically defined content
  • So, what is the problem?
  • The potential of semantic information
  • Microdata standards such as Schema.org and Microformats
  • Digital Asset Management (and Adaptive Content)
  • Adaptive Content
  • Image and media banks in your publishing system
  • Personalization of information
  • URL strategy for dummies
  • Common excuses for breaking established URLs
  • Ok, how to then?
• Web design
  • Gov.uk design principles
  • Start with needs
  • Do less
  • Design with data
  • Do the hard work to make it simple
  • Iterate. Then iterate again.
  • Build for inclusion
  • Understand context
  • Build digital services, not websites
  • Be consistent, not uniform
  • Make things open: it makes things better
  • Keep it simple, stupid – KISS
  • Do not break the web
  • Persuasive web designs (PWD) – design that convinces
  • Be clear in everything
  • Be very careful of what is the default setting
  • Visual hierarchy is important
  • Focus on the common goal you and your visitor have
  • Try not to overexert your users’ attention
  • Responsive web design
  • The mobile moment
  • The elements of responsive web design
  • Arguments for responsive web design
  • Notes on responsive construction
  • Responsive typography
  • RESS – Responsive Server Side
  • Adaptive web design
  • Design with data – a data first-approach
  • Get started with design with data
  • What you know about your visitors
  • Continuous A / B testing
  • Examples of A / B tests for monitoring the website, and other communications
  • Mobile first
  • Mobile first vs. responsive web
  • The mobile opportunity
  • Mobile restrictions
  • The mobile moment – when mobile users are in the majority
  • SPA – Single Page Application
  • Design of SPA websites
  • Challenges of SPA
  • Web standards, and usability
  • Progressive enhancement and graceful degradation
  • Usability vs. accessibility
  • Gamified design
  • Design and plan for errors that will occur
  • Your website is a magazine, not a book!
• Web performance
  • Planning for the unplanned
  • Performance optimization of databases, web servers and content management systems
  • General troubleshooting
  • Planning for high load – use cache!
  • Content Networks (CDN – Content Delivery Network)
  • Databases
  • Web servers, content management, own source code and external dependencies
  • Measuring and improving interface performance from the user’s perspective
  • Helpful tools
  • Editorial performance impact
  • Technical settings for performance
  • Recoup an investment in web performance – is it possible?
• Test your own website
  • How to document your test
  • SEO
  • Indexable for search engines
  • Duplicate content
  • Page title’s length is under 60 characters
  • Page title is readable and understandable in the search engine results page
  • Page title contains relevant keywords that describe the page
  • Correct headings are used
  • Search engine friendly URLs
  • Descriptive text on all important pages
  • Reasonable number of links
  • Pictures have alternative texts
  • Structured description of the information
  • Web analytics
  • Current visitor tracking scripts
  • Tracks the use of website search
  • Performance
  • Reasonable time for loading the page
  • Compression of text files
  • Usage of the browser cache
  • Scripts and style sheets are sent in a compact format
  • Images are optimized for fast transfer
  • Reasonable number of background images, scripts and stylesheets
  • Requesting files and pages that do not exist
  • Minimal amount of scripts and CSS in page code
  • Images are not scaled down using CSS or HTML
  • Identical files are not referenced
  • Reasonable amount of scripts in the page head
  • Content networks are used when necessary
  • Accessibility and Usability
  • Website validates the chosen code standard
  • Using correct header structure
  • Anchor-texts are descriptive
  • Link titles not used for non-essential information
  • Favorite icon is present
  • Possible to navigate with keyboard
  • Texts are written to be read by a human – not with exaggerated SEO
  • Language set in the source code
  • Not depending on browser features
  • Specifies image sizes in HTML
  • Works with and without the www prefix
  • Only one domain is used for the website
  • RSS subscriptions can be detected
  • Useful error pages
  • No surprises when scrolling
  • Enough distance between links, buttons, etc.
  • Acceptable text size
  • Zoomable, also on mobile
  • Icons for the website
  • Useable printouts
  • Others
  • Forms and other sensitive information is sent through a secure channel
• Tips on in-depth reading
• Sources & references
• Thanks goes out to…

Check out the Web Strategy for Everyone by the publisher Intranätverk. It costs from 10 $ + VAT ›

The post Web Strategy for Everyone released today appeared first on Web Strategy for Everyone.