stekker :: repository

The new social economic rules for the web or the differences between consumer-to-consumer, business to consumer and business-to-business applications.

In the growing pains of the web we are getting to the point most people actually have figured out what you can do for and with the Web. Technology is weird. It enables us to do things we never imagined doing but it also limits us to what a technology is capable of. Although you can go on and on about technology and how it works it is not where I want to go in this article. It is not really relevant how the some technologies appear or not, it is important how they are used and how they evolve in comparison with economic and social developments.

Suddenly it appears, a phenomenon that you never imagined possible. It is there because new technologies make it happen. One of the last I read about is the tail effect of markets. I could have thought of that myself but I didn’t. (That’s why I’m not a industry leader). What it makes clear that the Internet works on many levels and while it is called the “world” wide web, it becomes less and less global and more and more individual. (Ok, boundaries of national states disappear on the Internet). Say the mass-market goodbye it will all become one big niche.

So what implications does this have usage of the web in general? Where and how can you make money on the Internet? Where will it al end up? Well here some thoughts on that divided by business models of consumer-to-consumer, business to consumer, business to business and the missing one mentioned … Welcome to consumer to business!

The old stuff

In the relationship between entities and markets onl y a few where worth looking at for marketeers. Business to Business (B2B) and Business to Consumer. Consumer to Consumer is not really relevant if you want to do something to make money with. So B2B and B2C are pretty standard and well described in all kinds of books and publications. While I was surfing on the internet I noticed that most cool sites do something that is hardly described Consumer to Business. (I could not find it so if anybody knows some articles about it let me know.)

Businesses are always trying to find out what their customers want. Mainly because they are trying to sell stuff, and you want to do that it is quite to know. One of the ways to do this is market research. Ongoing research determines what you have to, at least for the bigger companies. This is hard to do and thus expensive. Next to that it is not a exact science. People tell some way or an other what they want. While I was studying Industrial Design Engineering, I got an reasonable amout of marketing and consumer behaviour studies, I always wondered why some products did better that others, and to make a longer summation short… one of my conclusions is:

“In most cases people do not know what they want or that they want it all.”

This goes for every type op product - simple or complex - and once in a while you see products where somebody held to this principle and just did her or his own thing and made it good enough for the rest of the world. When you have to develope products this is important to know. This also made me decide why I did not want to in the marketing business, next to the fact that i hate money off course. Design is much cooler but it isn’t my thing.

Why guess…

So if it is hard to figure out what people want why don’t use companies the internet to figure out what people want, or even be designers. What makes the internet different from normal business customer relations, well I think two things: Interaction / exchange of ideas. Eureka!!! Let your customers deside what do and make. Ok this is not suitable for all products. But good ideas will pay off and why hope that you hired these people to think of these ideas.

Off course you have to do something in return. Depending on the idea you will have to ensure both parties that they don’t misuse the concept. This can besolved by a control mechanism (like ebays rating systems) or just try to be honest and trustworthy by allways keeping up the communication and have dialoges with customers and idea people. (so this is not content push but content pull).

CMS systems nowadays

As I have tried to figure out in my first real post, CMS systems come in different forms and shapes. In the four groups which I defined in that post there is in my opinion one that could use a overhaul so it would be more user friendly. The simpler sites and CMS systems. This is probaly no place to find any money (i don’t care) but the place where the profit for the user is greatest. For simple sites it is a pain in the ass to explain how a CMS works to a normal user. Especially when CMS systems cover all possible usage of the cms. This will only end up in an elaborate CMS of which users only will use a precentage. Leursism gave a link of Jeffrey Veen in which he describes the same probem.
Ok in the end everything can be learnt but as I stated in the post about menu’s hierarchies in these sites people have a hard time understanding hierarchies. Next to that with the CMS systems usally have a back end (the difference between the front and the back is miles appart) which makes it hard to imagine what happens in the front of the site. Some CMS systems have inline editing but it is only a part of the CMS and not their main edit area.

An Example:

the front…

the usual login

the backend dashboard WP

or an other example: Mambo

mambo dashboard

mambo page management.

See the resamblence between the front and back? I don’t! Don’t get me wrong for specific purposes these ways of managing sites are good. Probably Word press is an bad example, because the usage is some what other than normal sites.

Because it’s open source it is good, but applying mambo on simple sites is overkill. With really simple site I would rather use a html editor instead. Most of the sites I know have static pages in them. Pages like mail forms. These are usally not editable for the user, this because the actions they imply are most of the time server driven, so you need to know server side scripting to make them work. I will discuss a possible solution later.

What normal CMS’s need to have…

I’m talking about CMS so you need to manage your pages. Otherwise it is hard to speak about a managment system. So what does a user want form a CMS in terms of functionality? I think this:

• page editor
• page meta discription editor
• page manager
• file manager (images and other stuff)
• menu/category manager/editor (just to place pages in a hierarchie of some sort)
• user manager (sometimes)

In most cases all these functions are what the back office is about. And if you put this in a information architecture flow you will get something like the image below. For this example I simplefied it somewhat all the real managements tasks like: edit, new and delete are assumed within the pages.

What you see is that the “link” in between the front of the site is cut loose from the back office. The back office consists of pages that have no relation what so ever with the sites stucture. The back office starts with a dashboard like page where user can choose a managements task. The managements tasks can be generaly be divided by:

1. page management
2. category management (the hierarchy of the site > menu)
3. image or file management
4. user management

After the the page management tool there is something to give the page an abstract, title in other words meta data. After this has been saved a category is chosen to place the page in the menu. (choosing categories gives problems on their own) Next the page can be edited itself. In the page editor you can edit the contents like; text, images and links with a RTE. Links can be made via a page choser or a user can make external links.

A combination of new technologies: Inline Editable CMS

Problem is that there is some parts in the cms where the user should follow a path and sometimes edit content with popups. The CMS does not give any feedback where the user is in process while using the path. (for example pick a page form a page list).

Some of these functions are accually in the site already. Navigation (management) through pages are there (at least I hope so) and placing pages in categories so they will fit in the menu is also there. For linking pages internally you can use a popup where also the pages are displayed like the normal menu or in collapest form. For normal editing you can use the page ifself. This can be done with a RTE for text that should be formated. Depending on the design every editable text could have a chooser for styles (this is also possible in the RTE but should be set properly). The rest of the functionality can be placed in pallets. Shown in the figure left.

AJAX

editing the title of a picture in flickr

To make something completly inline editable you need something like AJAX. This way you can directely edit every field, the way flickr makes it possible. Within flickr there are some features that make the use very easy. Left the is an example of editing the title of an image. when you rollover the title it lits up (1, 2). Then by clicking it the litup area changes to an editable field (3), which you, when done changing the title (4), can save. It even says “saving” (5) while it is being saved in de mean while. Afterwards the new text is visible. (6)

With all this you end up with a CMS that feels more like a normal application like WORD or something like this.

Templates (XHTML) and CSS

What you need to implement this is well structured XHTML and CSS that can be easily extended to the cms. The CSS od the frontend should be actually allmost the same as the “backend”, but with an extra layer on top of it. This way you have the feeling that noting much changes while you login as a site admin. With the use of templates this is easy to reach. Problem offcourse is that not every design is suitble for this approach. I think you can sum the two most important guidelines for implementing this kind of CMS:

1. The interaction design (stucture, navigation) should not be too ragid.
2. The userinterface / graphical design should be not to strickt.

Let me explain somemore about this.

Ad. 1: When implementing a site like this, the interaction design should be open. So everything you can do with the site should be easilly accessed or navigatable. Sites that hide information and content that are accessible via seaches or forms are hard to implement this way. Each object in the site sould be edited like in a wiki, but more direct. Other wise you building a meta CMS instead of an inline site CMS.

Ad. 2: Appearence designs should leave room for editing tools. Otherwise the editing part will like jelly: each time you want to edit a block or object all other objects in the site will be jumping around which gives the user a itchy feeling. At least make it so that the editing object only ‘falles’ a bit and let all other objects be in place as much as posible. Also give some space to the design that this feasable.

Comments closed (due to slpog) if you want to leave a comment send me an email to mic at this domain…

Making Categories

We are continiously categorizing the world around us. Just because it is the most efficient way for our brain to work I guess. So why is it that categories in site menu’s do not work so well? One problem is that everybody works in different ways. Especially when it comes to making categories. If it comes the point of hierarchies, people are really bad. We do not work so systematically. People rather make connections (just as our brain does with different neurons). And if something is really important the connections are stronger and easier to be remembered.

Hierarchy and structure

Developers like hierarchal menus. If it comes to programming or scripting it is the easy solution. In my years up to date I have not seen one good menu. (Ok the simpler the better). While instructing people to maintain a site, you always have to explain them how the menu structure (hence hierarchies) work. Menus with pages are ok: it is a way to make pages in your site accessible. Menus with sub menus and sub sub are awful for finding pages and hard to control (these cascading ones). You cannot see what is content is about. You as a user have to guess what category you page can be in.

Making good structures is hard

If you have to make a interaction design for a site you almost always end up with pages that are really hard to categorize. For the same reasons mentioned above. Which will end up that some categories will only end up with one page (links category is a good example). Some categories have so many pages in them that you really need sub categories to avoid menus with more than 10 pages beneath them. If you have more than a hundred pages, search functionality is an option. But if that is not the case you have to find another solution.

What to do?

Users are generally scanning pages for the right information. Menus in that sense are generally avoided. People do not read the categories as a text but they scan them. Then most of the time you guess what category your page is. Just clicking the menu and hoping it is correct. Because we are working with hierarchal structures here, one page can only be inside one category. Why is this? Probably it is because for developers it is more difficult to link a page to two categories or more or that’s not the way they think.

• Solution 1: Not making categories.
  So what would be the problem if pages are placed under more menu items, as long as people know that menu items are them to guide them. If you visually make clear that the pages displayed in the menu are somewhere else in the menu I don’t see the problem. Actually if the menu is not so dynamic that it changes every time you use it you know how you got there. The advantage is actually that the system is multi interpretable.
• Solution 2: Keeping track of the user and gathering his most visited pages.
  So make a user customized menu item. Ok, you need a cookie for this, but if you explain why you want to set it in the first place, probably it will be no problem.
• Solution 3: Leting other people categorize (or tag) your pages.
  A solution i found on the web (and which I though of myself) was by chris j davis who lets users tag posts he put on his blog(?). More than one person is deciding what a page is about.

What to do at the back end?

In order to avoid categorizing every page, a page should be judged on content. In short descriptions a user maintaining the site can ‘tell’ a cms what the page is about. Maybe an extra can be categorizing the page multiple times so a user can see what the maintainer had in mind while putting the content online.

Order within the short description is valuable information but not the main part. If a more than one page fits a description it can be grouped into a menu item. Pages in their term can be part of more groups. The order can be used to order the menu item pages. For example, say you have four pages with the following short description:

• Page 1: Content and ‘information architecture’.
• Page 2: Maintenance of site relating to content through CMS.
• Page 3: Different kind of CMS related to the content and maintenance.
• Page 4: How to make better menu items through ‘information architecture’.

The menu could look something like this:

All the similar words have been picked out and made into menu items. Pages appear in different menu items. If you say limit the the number of menu items and increase the number of pages the systems should figure out what the best combinations are. This way probably the pages that fit best together will be grouped into one menu item.

In the next coming weeks I’m going to try to make this work.

Comments closed (due to slpog) if you want to leave a comment send me an email to mic at this domain…

STICKY: if you any suggestions please comment…

Preface

There are number of site types. Ok thats true you say but why do you want to do this? Well because up till now I have never made an overview for one. the other is I want think about different ways to approach these sites from the point of view of users, maintainers and developers. For me internet is a medium for exploration.

Not that I “surf” the old way but most sites I frequently visit have some way of interaction. Do not read interaction in the way of playing and clicking buttons and so on, but more interaction with animation, information, the way sites are structured etc etc.

Because the internet is growing to a medium that can be mantained via the medium itself (maintaining a site via a site instead of knowing HTML) it can be helpfull to see what kind of content needs or uses what kind of management systems.

So in this article I want to dig into the content part of sites and how a user can maintain the different kinds of sites. With maintenance i mean the management of the organisation or person who is responsible for putting the site there.
Starting the list: (this can always be more specific so tell me if you know more….)

1. presence
2. gaming
3. experiance
4. portals searchengines.
5. blogs
6. informative
7. news
8. web shops
9. communty driven
10. applications

Exploring the differences content wise

1 PRESENCE Sites

Allthough not so common any more, what I mean with this is that these kind of site do not really have content. Sometimes companies have them to be present on the web. Portfolio sites with only images an no real conten fall for me in the same category. Once you have visited them there is no reason to return. “I have seen it and ok it was fun but i have no clue why I should return here and waste my time again”.

I think there should always be a reason to return (I was guilty on that part as well). They are easy to make and maintain. A static html site would be in most cases the best option.
Examples: piping care, any portfolio site (architects etc) this is nice portal… dutchdesigners

2 GAMING Sites

These sites have special services. They are not my thing but they exist. The games are the content and viceversa. Content wize there not really any thing to get. The only thing why you would return is if you are playing a game. This offcourse can be addictive. Maintenance wise there is not much to do, but making the games itself is off course more time consuming.
Example: shockwave

3 PORTALS / SEARCHENGINES

Portals were something really popular for the internet bubble. A giant collection of links to other sites. So no real content. I know lots of people that used them. because they were quite handy. Problems with these kind of sites were that as a user you still had to look for the right link. Sometimes is is just easier to type the URI you think it should be and just open the domain. Content and maintance wise they could easilly be static HTML. But for category reasons they are probably managed. Bulding and maintaining does not have to take much time.

Nowadays searchengines have taken over because they are easy to use and gvie good results fast. Technology prevailed here. I am not going deeper into searchengines and the technology behind them. (I know enough of them to make them work for me but how they accually work is not my cup of tea). Problem with searchengines is that they do not look for information the way we do. (It is improving though). Now there is a new branche that acually do SEO or Search Engine Optimalisation. So we are adjusting to the machine instead of vise versa. But momentarilly there is no real other solution.
Example - Portal: lawyers, startpagina.nl - Searchengines: google, msn, yahoo…

4 ENTERTAINING / EXPERIANCE Sites

Two types I’m trying to describe here are mostly used for experimental purposes (Entertaining) or branding reasons (Experiance). The first of the two were kind of popular in the time of the internet bubble. Artists and programmers were experimenting with dynamic interface and animation. Although fun to visit. Most of the time there was no propose to them.

Recently some of these experiments are made of use. The dynamic relational news finder on cnet, I can no longer find it though. Especially interaction experiments were interesting but sometimes not usefull at all. Depending on the pluginn used they were made with, Java, Director (Shockwave) or Flash. (Same as the Gaming sites). Maintenance is done thought the pluginn itself (Flash sites) or via HTML. Making the experiments costs propably more.
Examples: yugop, amaztype

Brands sometimes use sites to express themselves in a way that is impossible through other media. Most the times these site contain no real data or content. These are specific sites probably because a brand is not in to makin shops. Distribution is skipped this way, but as an user you should be able to find the thing you want to. The content is an experiance, in some way comparable with TV or broadcast. See it as advertizing in a different way.

All the other sites here mentioned can be a part of this experiance. Because the site is linked to brand or a special product the sites do no require maintenance. They are just build for people to use for a certain amount of time. When pluginns are used to make the site function and they only are ‘live’ this is no real problem, else maintenance can be real issue if that is required. Building them are relatively time consuming.
Examples: nike, sony, coca cola

5 BLOGS

Blogs of course is new variant of a personal homepage. The difference however is that blogs are more in a dairy form. These sites are hard to classify in the sense of what the content is about. Sometimes blogs are collections of different things which are of interest of the maker. So for instance nice news items or just pictures made during the week. Others can do some serious journalism on their blogs. The key thing of blogs off course is that, this is one of the examples of not knowing how the internet works specific and adding content to it.

Development is mostly done in a open source setup and you may donate, but if you don’t want to you can just use them for free. The time developing a blog is relatively short compared to amount of time maintaining one. Maintaining a site like a blog is the time killer of course. The content however is most of the time ‘low’ tech: Text, images and links and sometimes video or audio.
Examples: regnyouth, this one, and some friends: maan, remko (see in menu left)

6 INFORMATIVE Sites

These can be very extensive and big. Usaually made for distribution of data which is complex. Governmental sites of knowledge databases. They come in two types hierarcal and relational. The first is only usefull went the information in the site can categrorized easy and good and the quantity of the information is not to large. When it starts getting bigger and bigger.

The problems for the user arise (and the designer and developer): How do you pin point the right information in these quantities? Searching by keywords or phrases can be a solution here. An other way is by making the use of linked relations. Where you see relational information in site is tutorials. Normally when searching a function or any thing the page displayed is also showing related functions.

Developing and managing these kind of sites takes time. the more complex the releation between the data the more time it takes. If you want to build these sites in a good way you should figure out what and how to display when: in otherwords interaction design and information architecture. Nowadays these sites are maintained via CMS’s. This is kind of logical for a delevopers point of view.

Especially with complex information this a solution. I don’t think this is the most user friendly way for a user to maintain a site. True, if a huge site can be build from other information managements systems, then, for a developer it is easier to build a shell where a user can manipulate the information shown on the site. But it is a bit like telling your maid to put the landry in the machine, because the manual was writen in a language you do not speak. I will come back on this issue later.
Examples: apple kbase, bosch, php, overheid.nl (dutch)

7 NEWS Sites

the big difference between news sites and informative ones is that the time span of the content is shorter with news sites. When the content is outdated, the relevance on the site less and so it is harder to track back. Allthough for professional news sites it is a good idea to store all the news ever writen (just for research puporses). Some news is out dated faster than others. Probably not accessible for every user they have large ammouts of information that can be searched on relevance etc.

Other main difference of course is the commerical aspect. Many sites of commercial news corporations have large numbers of journalists and editors. These people have to be paid. In a news corporation editors need to be in control of what is published. (professional responsibillity) so managing the streams of information is primairy goal of the back end system.

No simple CMS is able to provide something like that. Developing systems like that are specialized for that purpose and that purpose only. The size requires maintenance from system maintainers. Developing take large amouts of time. Especially for news papers who have multiple output media it is key to have the information stored in one place, and later on port it to the different media like print and web. More and more, news sites are accessible for everybody, but subscribers have more access to certain areas of the site.
Example: cnn, cnet, the times, bbc, nrc (dutch).

8 SHOPS

Allthough shop are in alltogether different category I still want to discus them briefly. Shops - of course - try to sell stuff. (Being digital things or physical it doen not really matter). The content in informational sense is rather poor. There are product discriptions and product codes. But finding shops via google or any other search engine is hard. If I search for a mac book pro i probably end up by apple or any other review site of this machine, instead of the shop around the corner. Content wise there is not much to do, but developing or maintaining a site like this is more complex.

Many factors determine the process of selling and buying. For a user, what the steps are to buy anything, can be the reason for succes or failure. So the front is of site is more complex to build. For the backend all the different aspects of selling anything managable. Say for instance stock, pricing, special offers, customer information etc etc. Sometimes shops have be connected to other systems so it is all the way maintained through a different program.
Examples: applestore, amazon, bluelink, KLM

9 COMMUNITY Sites

What all good community have is that all users have something in common what these users want to share. The first communties where probably about the web and what you still see is that communties for posting and responding about questions like programming issues are really working. Accually is this what communty sites are a frame work for sharingideas, questions etc. Developing something like this takes time but if you have one working, managing the content is rather easy.

Content is generated by the users of the site. So mantaining communities is about managing people rather than content. You need moderators. What you see is that in these kind of systems, the specific role for different users is determining, what a user can or cannot do. Userlevels and their specific functionality.
Examples: cybernetic broadcasting, phpfreakz

10 APPLICATION Sites

Applications are no new thing to the web, but starting to get more common. The first build applications were off course CMS systems and shops. Applications are tools for getting things done and not much information. Applications work well in complex problems environments. The use of web-based applications can help sometimes because they are centralized, up to date systems and give the possibility to offer a solution to the apparent problems that need to be solved. A good example in this case is funda. Funda is a house search site in Holland with many possibilities. How complex the application can be on first sight, if the effort is worth taking they will succeed. If the interaction improves the application will be success. New developments like AJAX are making that happen.

What is appearing now are applications which are not commercially based (what their first appearance is anyway). You can call it web 2.0 but instead of trying to label it in a technical sense (like AJAX etc) I like see it as an usability and social development on the web. More and more people are starting to what this medium delivers best. Sharing information and that is what these applications like flickr do really well, hence their success.

There are two types within applications:

1. Self-maintaining services like ebay or flickr.
2. Services need maintenance from the people who supply the service like funda.

These are the two extremes but there are more possibilities within the spectrum. So how does a good application compare to content and maintenance?

First the first: self-maintaining services

Depending on the type of application the type of content differs. For the first in the spectrum above I think that applications are there to manipulate content. For user content, in the strict sense and in that case, is object to change. These changes can sometimes be done by the application. Communities and applications like flickr do that, although they both do it differently. Communities or forums work with posts and threads. (threads are nice when the are not so long).

Applications like flickr tend to do it more in a way of direct manipulation. So adding content and edit this directly in almost the same spot. The maintenance question here, is mostly one of keeping the application working. Servicing 8 million users a month is demanding on your servers and every possible problem should be tackled fast. A good interaction design for an application like this key.

Testing is the second step, maintenance the third. The does not necessarily have to be a web tool for keeping these kind of sites but probably they are there. You could say this is an application for the application for the applications sake.
Example: flickr, ebay

Then the other: not self-maintaining services

Process control is one of the main purposes of applications like this. A lot of these services are not public but closed for whatever reason. If a service is public the data has to be provided by the party offering that service. Funda makes the process of finding a house easier, but before you can find houses they have to be put into this system. The basic information has to be put into the system but usage can off course also offer information, the more ‘intelligent’, the more the options.

Logging usage can be really valuable, because the application is offers ease of use by remembering what you did last time. The other force is combining services. While searching for something with multiple parameters support for these parameters can be helpful. For instance searching a fish restaurant in a specific neighbourhood it can be helpful to see on a map where it can be found.

What this means is that content can also be provided by a third party. Maintaining a site like this can be rather extensive. A whole system has to be build with multiple user types and their login so the content ends up right.
Examples: funda

Conclusion

If you set the 4 parameters here discussed in a graph it will lool somewhat like this (for the rest of the conclusion i’m skipping the first three / four types):

Note here the bars have no real dimensions - they are ment to give a comparison between the different kind of sites. What you see is that the more complex a site is the more time it takes to maintain it. So simple sites can do with simple cms systems and when larger companies are involved the larger the site hence a more complex cms. Simple cms systems do not really need user right systems (they are not nessesary because few people work on these systems). More complex organisations need user right systems because different kind of people can edit the site.

Applications have other issues for maintenance. Moderartion is one. Here probably you need user rights management as well. So content only need to be managed. Thats why in the graph you see management complexity is quite high (time needed off course) but contributing content doesn’t take much time.

What I notice are two things:

1. People / Organisations who need simple cms systems use too complex ones. They are better off with a site that they could edit directly.
2. We are going more and more towards direct manipulation (applications, simple cms systems).

For the last point: Simple cms systems should be more like editing a texteditor on the web. It is on sense of managing lots of pages when you only have about 50. Especially when only one or two persons maintain a site like this.

Comments closed (due to slpog) if you want to leave a comment send me an email to mic at this domain…

Allthough I’m no authority on the internet I want to use this site from now on to explore all kinds of things about it. Hence the name “Repository”. A repository of thoughts about this medium in general and sometimes specific. Recent developments on internet and the state it is now in leads me to thing it is not yet fully developed (when it will be remains the question).

The point of view of this site will be mainly webdevelopment from a usercentered of design background. This because it is the field I’m working in. Most of the things i encounter are developed from a technicians or developers point of view. This tends to result in products are only usuable for these kinds of people. (No offence but, that’s most of the time not the best approach).

That’s it for now…..