Questions on redesign and upgrading of projects always arise at web sites owners and are the spirit of the time, visitors’ requirements are going up. Much was said about difficulties that web site builders face when a qualitative (or more qualitative current) project should be made and what control system of a web site to chose – free or paid or maybe develop your own. One of the most popular questions on the web building forums is “Could you please advise CMS to me…!” but how can one advise something if it is quite difficult to understand from a single phrase what requirements he makes to the system. Today I would like to devote this article to navigation – basic and secondary.

Global navigation

An average site feels fine in the context of a single domain name but the situation changes significantly if the web site grows up, the content size increases, new services are added and new partners appear. And the site owner understands that any perfect rubricator isn’t enough in order to assemble the information architecture to a single whole. A model example: when the site owner runs some business or provides some information not only for visitors of his region but for partners from all over the world. Then the single solution is to release website in several languages and allow site visitors choosing the language he prefers to obtain the information in. Last days there is a much popular move which works like this: you don’t develop a global navigation within the limits of one system of site navigation and the single template of visual interface, but to divide the templates completely for the language versions, and further still to divide not only the templates but the websites also and to give them different names.

This comes from my personal experience: I took part in the development of multilingual websites and I can say that there is a real problem with the templates for different languages. For example, one website has the English, Spanish, Italian languages and the Greek language should be added. The site was send to us for updating with a ready design which was impossible to change, and I’d say that to insert the complex navigation on Greek, the whole content under the adjusted illustrations, etc, it was something. As a result we have changed the engine imperceptibly. And we had to implement every template with design for every language separately; the simple copying was enough for the first three but an entire rework was done for the fourth one – re-makeup. As well as imperceptible changes in design were done. Web developers face such problems who write for the multilingual projects including for example Arabian or Hebrew and I am not talking about double Dutch. In these cases not only individual templates ate needed but the individual structure also as well as distinct content meaning not just the translation but modified. In such situations we may need the control management system not of the site but of the sites; and such systems are presented on the market and have been successfully developing.

Rubricator

The main element of navigation is the navigation menu of the website is developed on the basis of the document rubricator. When the document is added to the site it has a correspondent heading. Names of the main headings correspond to the names of the elements of the main menu. The document which was defined a heading to will appear on the site at the correspondent category, if a user opens this category via the menu he sees his document as a last added. At the same time classification of all documents comes at the reverse order, i.e. the last added document becomes the first on the list.

Basic definition of the headings is the necessary service which almost systems, free and paid, provide. The task becomes more complicated if one should have more than two layers of rubricator. I.e. main heading – subheading – document. Two or three nesting are enough for the average corporate or informational website. In the most complicated cases the number of nesting in the rubricator may be optional, large as you wish actually. Besides, there is a problem as a result of deep nesting of the rubricator to make navigation convenient and plain for the visitor according to categories of the site; the visitor won’t be much interested to click the site categories one after another in order to find the required document reaching the 50th step of the depth being unsure that he has selected the right chain.

The second drawback of the systems which provide the capability of the similar complex informational architecture with a numerous amount of category nesting lies in the fact that if the system is set for the project which doesn’t require such a service, the website obtains the redundant engine which is too bulky and difficult in the setting as well as in using. Even if the system uses the modular structure allowing to disconnect the unused services and to pinch the capabilities, the matter doesn’t already concern the fact that the system is optimal and compact.

The problems with categories won’t touch those projects where the number of documents is moderate and the order of visitor’s actions is much predictable: select the category, see the final amount of subcategories, look through the documents in the subcategory, and obtain the required information. But what if there are dozens of thousands of documents on the website? Or maybe hundreds of thousands? It is quite easy to orient in the small amount of categories or subcategories but one has to pay for it by the too long paging.

Paging

A traditional kind of a secondary navigation. It brings out lists blocks of the links on the given number of positions; it presents a characteristic example of the linear navigation. One can meet it elsewhere. A typical example is the search system which gives the search results as a rule of 10-20 positions on the display, the next page contains the next block and one can look through all the search results page by page one after another. It is much easier to sort data by some leading features and give them page by page at any sampling where is much more data than adjusted. In the popular forums, for instance, sorting is carried out through the date of record creating: new records appear at the top of first page, later ones are at the farthest pages. Usually at search systems the sorting is carried out through the weight and through the measure of significance of the document, then when turning over the pages one can look through the documents from more significant to less.

At the library the documents from the one separate category are brought out according to the time of adding. Although if it is not difficult to look through the samplings and to browse some pages at some categories, then in others, much filled, to browse all the pages seems to be an incredible deed: dozens of pages with dozens of links to the documents, articles, news, etc. This is the reason of such a trouble – some quite interesting articles and reviews are simply lost in the stream. Nevertheless that such an article doesn’t lose its urgency, it becomes invisible for the visitors of the library. This is the main drawback of this type of navigation: it is the right solution for the sampling with a small amount of documents but doesn’t solve the given task along with a large number of pages. And the user has to give out either a simple navigation, three layers maximum, create a simple rubricator and long paging or to structure information more complicated and to increase the number of nesting of the subcategories at the navigation, thus reducing the number of documents at every youngest subcategory. It won’t be convenient for the user in both ways. And that is why in order to find the required information he would rather use the search system.

Search system

Every website should have a search system. Even if the site contains ten pages, search service is the typical example of the secondary navigation allowing the visitor to get the required sampling and maybe the necessary document only through one step. The visitor is unaware that your site contains just 10 pages, or 100 or even 1 000. The analysis of libraries shows that every 35 persons out of 200 use the search box in order to find the necessary article. At the same time there is no such a question on how to make a search over the site till your site is of a moderate capacity; the same goes in the case of rubricator, paging and other types of navigation, problems occur with the growth of the website and they are still the same. The search result may be a list containing the hundreds of positions which the user would have to browse 10-20 items; this is already wrong and one should already look for some alternative solution.

Such an alternative solution may be deployment of the advanced search to the site and search by additional features. The most popular and already usual features are the limitation by:

1. category (category - subcategory). Here one can add the global navigation not only by the categories of the site but also by the network of sites;
2. date of publishing (by a fixed date or to set a period – for some specific year or from one date till the other);
3. logical conditions (sampling contains word1 and contains word2; contains word1 or contains word2);
4. rigidity of search (consider/or not word register or the rigid correspondence of word or word1+word2 or allows non-rigid correspondence);
5. etc. (what else the developer’s or customer’s imagination is enough for: by the author of the document, by the logical anchor – commercial/non-commercial or what not).

It is clear that no one would create such a complicated configuration at the main page or at the internal pages. There is an already usual and nice solution to place on every webpage a simple search box (search by a keyword or a key phrase) and a link to the page with an advanced search.
Another option is to use syntax for building more complicated inquires in the context of a simple search form – statement brackets, symbols for words combining at the query or some exceptions, labels and instruction words. One can place a link to the help document with a description of this syntax near the block of simple or advanced search. But who will use this syntax? And if the website audience consists of web developers or advanced users it is possible then to offer such a service to visitors. But what if the website sells some everyday equipment where the good visual search system is a must?  What is the possibility that the visitor would start learning this syntax in order to make a single purchase? Or not to make a purchase….

Besides it is wise to install the capability of syntax using on the website only when the language of inquiries will be standardized thoroughly, when w3c.org will publish the approved standards and when one single standard except the instinctively understandable operators will be known for building any complicated inquiries. And when in schools at the first classes on informatics students will memorize the main rules and mechanisms of searching the same as multiplication table in order that in future not in the context of web development or rather not just in this context but for a common everyday surfing over the net and every housewife will enter a complex inquiry in the single text field automatically just as quickly as counting the change at the supermarket.

Anyway, sooner or later an informational website will grow till the stage when traditional, typical methods of navigation would not be sufficient and the CMS developers will have to invent new solutions to help visitors obtaining the necessary information.

Reference source

New and recently added materials will be always more available to the visitor rather than those published months or years ago, usually they are promoted on the main page and are situated on the first pages of the categories. How to make the recently added materials and not always out-of-date texts not to be lost in the stream of time and in the paging bar? How to relate the document published today with the subject data which were published previously though related to the new ones? Informational websites use the list of articles on the subject which is a type of secondary navigation for the above purposes. For instance, the visitor came to the website (is it his first time?) and saw a brief article, news release, about some conference. It is obvious that the news release format doesn’t allow publishing full data about an event at the every informational message and the previous news releases already contained some information about how it was held, who took part in it, what subjects were touched and what it began with, etc. A common sense in this very example suggests organizing links block on the previous materials under the laconic news release and the visitor will be able to get the whole evolution of events without browsing through all the news lists and without scanning all the titles where only the every 30 record contains the information on the required subject.

Usually the editors of news informational portals understand the necessity of such a list containing links on the subject, although if the system doesn’t provide for the mechanism of forming such a list, the editor has to compose it manually. Firstly the experienced editor recalls all the previous texts and knows where to find them in order to copy the title of the text with a link to the correspondent document. Sometimes he can use the search system and will also compose the list manually out from the ready sampling. This work is quite routine and boring, it takes much time and complicates publishing of articles and news releases. How to automate the process of such a block formatting? This block should be formatted while adding the text to the site. Is the material adding via admin interface? Ok, then in the same window we do the following:

1. set the search by several key phrases from the article which is publishing right now;
2. select the appropriate materials with check-boxes from the articles list obtained from the search;
3. and just with a single move of a hand generate the unnumbered, for example, articles list from the selected positions by the check-boxes with a subtitle “articles on the subject”.

Navigation via address line

So as we are already talking about documents addresses … It is surprising how the web developers resist inserting to the control systems the capability of setting the right URLs which all the experts on search engine optimization and usability of web-sites prefer extremely. And here what is strange – this service is not that complicated in the general volume of system developing, i.e. lots of documentation and the code doesn’t become heavier. And still the programmers discover some ways and reasons explaining why it should not be done. If previously the main reason according to which the right URLs forming was simply a necessity, the fact that the search engines either indexed not properly or didn’t index at all the addresses containing dynamically synthetic parameters, at least convinced the programmers to make some efforts but today when all the respectable search engines index documents with any text in the address line, the right URLs lovers have lost their main trump card.

Then the disputes go like this: programmers say that there is no necessity in such a service in SEO if the search engines index any URLs. However the analysis shows that the presence of keywords in the URL influences the density of the keywords in the document and is considered while relevance estimation of the document according to the inquiry in Google for instance.  Programmers say that users do not use the navigation from the address line and give an example of some research results. But as for me, personally, I use. Many web developers familiar to me use. When opening a great amount of bookmarks in the browser on the set subject of searching, one orients towards the document address significantly – what is the site, what is the chapter, what is the page name.   

What else should be said regarding this not much popular secondary navigation through the document address? In some forum they have explained to the web developer that the SEO experts ask to write the code not via links in the form of GET queries but to make them in the form of friendly URLs and then to parse it wit the help of a script or with the help of mod_rewrite explaining that Google, for instance, indexes these sites better and gives a better rank to them, which doesn’t correspond to reality. However there is a slight difference between indexing and ranking. Another matter that too many parameters are considered while ranking. And there are many examples when the document having a right text in the address but containing an unclaimed content will be situated lower in the search result rather than a useful document with address containing some text with parameters. You can do everything properly but it won’t give the expected results. Is it right to eat one apple every day? Doctors recommend it and it seems to be right. However if you do so, nothing will change in your life. Ask your wife what is better: a man in clean shining shoes or in dirty ones. Every woman will say that a man in clean shoes is much better. Then clean your shoes and go out from your house and come to any woman you meet… and nothing.

The same goes here. One may not do beautiful URLs and still raise the project till the worthy position. And one may worry with a valid make-up and with right URLs and still nothing.

Tags

One more example of secondary navigation is to set tags to the document. Usually 1-3 tags are used at the one document. On the website each tag is a link, by clicking it the visitor gets the sampling of all texts with a tag with the same name. Such a sampling will differ from all navigation systems discussed above. Navigation via tags is more flexible developed in comparison to rubricator, i.e. there is no need in altering a tree of categories (it is much better for a user if the system of categories will be quite steady and not much deep) besides each document out of the list received via tags may belong to various categories, i.e. the sampling will be crossed, carried out through the all categories. Such a sampling will differ from the search results over the site, according to correspondent word, as there will be links to all documents containing this word in the search results; as for the sampling by tags there will be only those documents where this tag was set up. And of course articles over the website will differ from the sampling as for the big projects anyway you will include to this list the final number of documents, as a rule not more than 10. Besides, the editor moderates this list manually, shortening it according to his idea of the correspondence to his article. Setting tag/tags to the document makes a possibility to get another sampling which may be more complete rather than a list on the subject, as far as more relevant to some specific word rather than a sampling from one category (besides it is possible that the documents from such a sampling may be from various categories) and much convenient rather than search results.
 
 Besides, setting tags to the website may lighten lives of editors and moderators. A typical example is creating FAQ on the basis of debates at the forum: just imagine that as a result of heated arguments and experience exchange, opinions of experts as well as newbies; they have found a nice solution on some question. Time goes by. And there is another newbie asking the same question. Oldbies’ reaction is “Learn how to use forum!” and you won’t get any other answer. This problem is partially solved in the live forums with the help of creating FAQ manually; still if it is possible to set some specific tags and weight to some messages in the forum and the new post is generated automatically by this tag in the subject of discovered solution, then it will be much more useful information in those forums and its discovering will be much quicker and easier.

And the last one is the possibility to generate URL according to tags which may be indexed by the search engine. This possibility is quite pleasant for the site owner but there is a chance that mean black SEO specialists will simply add this method to their arsenal and this leads to the fact that users of search systems will be disturbed with the search results where they will have links lists by the keyword from the other website instead of required document. And this can lead to the banned website.