Using behavioral targeting to customize content
- Started
- Last post
- 14 Responses
- poomoo
At Scottish Television, we are actively exploring ways to serve up a degree of custom content to every user the 'deeper' they travel into our sites.
Taking a typical example of a user interested in football, we would serve up more football related articles to that user. We would be able to listen to the meta data on each article and page viewed in the user’s journey, apply weightings and serve up more customised pages.
As the user goes deeper into the site, we would narrow the focus of articles and categories. As the user goes ‘back up’, we would widen the focus of articles and categories.
The primary aim of "silent intelligence" is to create a continually evolving and richer user experience through increasingly contextual content.
We are making the assumption that a user wants more content on a particular topic. As we learn more from what they are reading, we want to provide ever more related content.
I am concerned that by continually 'narrowing the focus' of content we could lock users into a dead end. However, it may not be any more of a dead end than unrelated, uninteresting content. Although, ultimately the user is the judge of this, regardless of the method we use.
It's obviously important that we offer ways to 'widen the focus' quickly. It's also important we balance the amount of content that 'narrows'.
I'd be really interested in hearing the thoughts of UXD/UI/IA peeps here, positive or otherwise.
Thanks!
- detritus0
No UI expert here - but how fundamentally do your actions change the topography of the site? I mean - is the form and layout substantially changed through the user's travels, and if so, do you risk getting them lost in shifting sands, before they even hit the dead end you fear?
- graham0
I think something like this works well, and you could avoid the dead end scenario by keeping something which would promote the site as a whole - something like the BBC's "Most Popular Stories" widget.
- poomoo0
There may be subtle layout changes which we would iterate when prototyping. However, we certainly wont want to confuse or stress the user.
What would change, is the type of content we present to users based on their journey to the point they currently find themselves.
I think it makes sense. We would obviously track and AB test the final liveries to determine if it's successful, understandable and usable.
It would be good to hear differing opinions.
- Autokern0
Seems a good idea to me.
But "the assumption that a user wants more content on a particular topic" is quite wide. Say your user is a fan of the Rangers, you can probably determine it (but after how many clicks?) but would you be serving him news about other teams, just because is football? Or what if the the news about that team are quite scarce/old?And the problem of determining which is the topic he is intertested after few clicks may be a huge challenge in other sections, like news, economics etc. You may end up determining the topic of interest to late (after too many clicks) or just serving him "popular stories" which are quite random, at least for his/her interests.
But anyway, sounds good.
- poomoo0
We would be able to determine a range of topics the user is reading about. We then would be able to apply a weighting and some other intelligence to determine what similar articles should be served.
In your example, we would be able to draw that the user might interested in Rangers, the SPL and the Champions League. The more he reads, the more we learn.
However, at one point, when the focus is so narrow, there might not be any more scope to serve new topics and content. That's the dead end I am worried about.
Like I said earlier, we would need to investigate ways that widen the focus, quickly and easily. How far we open the lens, so to speak, needs discussed.
- mikotondria30
Surely the site would be designed so that the user is in charge of the verticality of the content they arrive at... Sport -> Football -> Rangers -> Upcoming Fixtures -> B-Team...
If I click these in this order its pretty much certain I want info on Rangers' B-Team upcoming Matches - give me that info..
How is what you're proposing different from this ?
- poomoo0
What you describe is a user browsing a section top to bottom. So yeah, the content for each of those pages will be Football/Rangers/Upcoming Fixtures/B-Team.
What I want to do is serve up data, at a category and article level, based upon what the user has actually read in the sections you described.
So at a Football level, this user might see more Rangers news than what other users see. They might also see more Rangers B-Team news if they have read a lot of articles about that.
A better example might be someone who reads many articles about Scottish Politics, Alex Salmond, the NHS and Edinburgh local news.
We can guess a fair amount from that. So, instead of seeing a default Politics Page, we might present a section or a component or some sort of UI element that serves up content we know the user likes reading.
- ninjasavant0
Its a good idea but its not a new one. Its an essential part of the user experience at Amazon.com. What you're describing though has little to do with UXD and more to do with Data Analytics. There are companies and services out there that will help track and mine data from users and based on their data you can serve up content. The breadth of the information would be dependent on your algorithm. What standard deviation do you want to use from their core data? What percentage of non-related content do you want to serve as a function of spontaneous discovery? That kind of stuff.
- poomoo0
Yeah It's based upon the Amazon idea. We have already talked to several companies that provide these services.
Why do you think it has little to do with UXD? I think it's fundamental to the way a user, experiences and consumes our services. That's why it's part of the UXD.
Does anyone think it wont work and be problematic from a user perspective in the context of a news site? I'd love to hear your thoughts.
- ninjasavant0
Not at all, so far as I know most corporate news sites do exactly that. The reason I say it doesn't have much to do with UX is because this process defines the WHAT of the site. The HOW of the site should be developed at a certain level independently. Figure out the data flow of the site in broad strokes first. What types of information should be going where. Article here, ad there, related stories here, etc. Then use the data to populate those things. If you design for the specific content first you're going to limit the usefulness of your design over time.
- ukit0
This is nothing new...
Although I think personalized content should be indicated that way. ie, "Stories You Might Be Interested In" as a sidebar on the page.
Otherwise, it gets confusing when someone tries to link you to a page with some content on it, and it isn't there.
- poomoo0
Ninjasavant, can you give me some examples of news sites that use behavioral targeting to serve unique/custom pages and content? The BBC, The Guardian, The NYTimes, Google and Yahoo don't use it. It would be great to see it in action.
ukit, every piece of data would have a permanent and findable URI.
- ninjasavant0
Thats kind of what I'm getting at. CNN, MSNBC, all the news sites do it based on your surfing habits on the site but you'd never know it because its just served up content in related links and advertising. I think the best example of creating a custom news feed would be Google's news site. You can tailor it to whatever you want. Or Amazon. I look at razor blades, all of a sudden I see front page ads for shaving cream and personal hair maintenance. You can see it even more clearly in the "Name's Amazon.com" page.
- poomoo0
Cheers. I think we're kind of misunderstanding each other. It's my fault for not articulating things properly.
- Its possible, I do that a lot. Or at least thats what my girlfriend tells me.ninjasavant