Personalized Content and Content Management Systems

I’ve been meaning to write this post for a while. Couple of months ago, Bill Maher‘s ending monologue was about personalized content. All of us who have been working with Sitecore and other content management systems knows what this is, of course. A very big majority of the users of Sitecore are marketers, and their goal is to captivate as much of their audience as possible, part of their strategy includes tailoring the content to the reader. Sitecore offers sophisticated tools to be able to track what the user prefers and then serve up specific content that interests the reader. This is great for the reader, who gets to see more content about topics that they like.

So – Bill Maher’s premise is that personalized content is making us dumber. According to him, more and more Americans are only getting “news” from places that only reinforce what they think. This leads to unchallenged thoughts and opinions, which can hinder an expanded thinking process. You can view the entire transcript here: http://bit.ly/1hXHpzE

You can also watch the video:

I get where this is coming from. I do believe that it is necessary to get a variety of information from all different sources in order to form a solid opinion about something. But as usual, there are exceptions to rules for everything. News is supposed to be about facts. It is factual information being told back to us. This IS supposed to come from multiple sources because one party may not always have all the information. There is a certain bit of corporate responsibility that news organizations should have. Its one thing to only get news from one source, and its another thing to get news about one thing. Organizations can control the latter, but they cannot control the former.

However, I think the level of personalization comes in at a much lower level, way after one has already chosen what source they want to get their information, and so it does not affect pursuing peoples focus. And this is why it does not really matter if the content at this level is personalized. If I like landscape photography, then I do want to read about landscape photography at my favorite photography website. But the website does not have any choice about the fact that I chose THAT website to go to. Nor does it have any say about the fact that I am not reading about any other sort of art.

This specifically (I believe) does not apply to branding and product sites. The idea of these sites is to make the audience interested in their product, and they have a very specific goal in mind – conversions. The site that Bill spoke about like facebook, yahoo, etc have a much more global reach, and a much wider variety of audience. Their only goal is to have people staying on their site – a very different goal than corporate and branding websites (not to say they don’t want that, too) – in the end I think people have certain responsibility as well, to seek out information and perspectives from a variety of sources.

If you do want to personalize content, take a look at Sitecore’s Personalization Editor

An edited version of this post also appears on The Runtime Report.

Dynamic Placeholder Keys

This post is just a bookmark for ourselves to this very interesting solution by Nick Wesselman to resolve the problem we have in Sitecore when adding multiple sublayouts of the same type into another sublayout.
Here is his solution: http://www.techphoria414.com/Blog/2011/August/Dynamic_Placeholder_Keys_Prototype

Even more goodies about Page Editor from Nick in this must-see post: http://www.techphoria414.com/Blog/2012/May/Sitecore_Page_Editor_Unleashed

Sitecore query for fields and custom TreeListEx control to support queries

One of the things that I’ve always had to deal with multisite solutions is handle the METADATA (or a set of assets) folder for each site – that is, to make sure that each site correctly uses its own METADATA folder in all the list fields in the content editor. In order to do this, you’ll need to put queries in the Source field for each list field. If you have a METADATA folder for each site, you’ll need the path you put in the the Source field needs to be relative to the item that is being edited. The way to do this is to put sitecore queries instead of absolute/hard-coded paths. There are a two different types of queries that Sitecore supports:

1. An enhanced syntax using parameterized datasource

This is basically a querystring style format with parameters you can pass into the source, along with the path and it uses the parameters to filter the resulting list. Mark Ursino has a really good blog post on its usage.

2. query: or fast: syntax

Some of the fields also support the more versatile query: or fast: syntax, but unfortunately, not all of them do. THe fields that support the query syntax are:

• Droplist
• Grouped Droplist
• DropLink
• Grouped Droplink
• Checklist
• Multilist

The ones that do not support query: or fast: queries are:

• TreeList
• TreeListEx
• DropTree

So – what to do? Well, Sitecore being as extensible as it is, it is pretty easy to make a custom control that can support this. John West has an approach to make relative fast queries work, by adding a custom pipeline processor.

An article in the Sitecore scrapbook gives a way to override the ‘Source‘ property of the TreeList control – whenever there is a query in the source field, the override property executes the query and then sets the source parameter for the control to use. If you inspect the code, notice that in the ‘get‘ method of the new ‘Source‘ property, we use the base property the control provides called ‘ItemID‘, which basically points to the context item.

The TreeListEx control works a little differently, because the control doesn’t actually load the tree until the user clicks the ‘edit‘ button, so this property doesn’t exist. I was going crazy trying to find out how to get the context item, and eventually found (with the help of John West) that majority of the context fields are available in Sitecore in the viewstate – all you have to do is expose it as such:

        public string ItemID
        {
            get
            {
                return StringUtil.GetString(this.ViewState["ItemID"]);
            }
            set
            {
                Sitecore.Diagnostics.Assert.ArgumentNotNull(value, "value");
                this.ViewState["ItemID"] = value;
            }
        }

Now that there is a handle on the ItemID, you can just run the same statement for the TreeList control.

There is a another point to consider, which is the links database. Now, instead of creating a new control like the article states, you can just override the existing TreeList and TreeListEx control and don’t have to worry about anything else. However, if you want to add your control, then you will need to consider the links database. The links database tracks links for all sorts of items Sitecore, one of the things it tracks is field usage. So when you go to delete it, it will pop a message saying that the item you are deleting is being used by other items. If you add your own control, then Sitecore doesn’t know about it anymore, and doesn’t track it. This can be easily solved as well, by adding some field types definitions in the /App_config/FieldTypes.config:

<fieldType name="Treelist with Source Query" type="Sitecore.Data.Fields.MultilistField,Sitecore.Kernel" />
<fieldType name="TreelistEx with Source Query" type="Sitecore.Data.Fields.MultilistField,Sitecore.Kernel" />

Alternatively, you can add your own config file in the /App_config/Include folder:

<configuration>
  <sitecore>
    <fieldTypes>
      <fieldType name="Treelist with Source Query" type="Sitecore.Data.Fields.MultilistField,Sitecore.Kernel" />
      <fieldType name="TreelistEx with Source Query" type="Sitecore.Data.Fields.MultilistField,Sitecore.Kernel" />
    </fieldTypes>
  </sitecore>
</configuration>

Note: Don’t put the namespace of your custom class in the FieldTypes.config file – you need to put the base class that it inherited from.

An edited version of this post also appears on The Runtime Report.