A series of features and products that uses signals from readers to create dynamic experiences that improve user engagement.
Key features:
Behavioral content recommendations: A block that shows recommended content based on reader interests.
Desktop view of the Card: Grid Behavioral block with all default custom properties.
Dynamic targeting: Ability to dynamically target specific users based on specific signals (such as, their reading habits and when they sign up for mailing lists/newsletters, existing subscriptions for users at login). For example, high-frequency readers may get more content and less ads.
Custom messaging: The ability to send specific messages, offers, etc. to a group of users based on their reading habits.
How does reader targeting work?
Our program has four main parts, which work together to create a unique experience for the reader.
Signals: As a reader browses through the site, we pick up on signals that help gauge their interests. BLOX CMS has dozens of built-in signals that are part of the software, such as answering a poll question, clicking on keywords, saving an article, browsing sections, signing up for mail lists, existing subscriptions of the user at login time, etc.
User profiles: The signals for each reader is collected and stored per reader, whether or not that reader is actually logged into the site. Profiles are then de-duplicated across all devices - mobile, tablet and desktop (or multiple desktops!).
Content:
With Behavioral Content Recommendations, the content is served based on a complex algorithm involving reader interests, recency and popularity.
With Audience Targeting, the media organization decides at setup time which audiences are targeted based on their goals.
How can I import data from an external source?
The TownNews.com iQ program has the ability to import data from external sources, as long as each user in the data set has an email address. This is a service that will cost thousands of dollars with other vendors.
If a customer has such data, then can create a CRM ticket and send the request plus an example of the data (or, the entire data set) to the Lotame DMP Tier 3 ticket category for review.
Some example sources that would be useful for ad and content targeting:
Current subscriber database
Contest database from a source like Second Street
Purchased data from external vendors (such as voting data)
How do I sign up for iQ Engage?
This is a new product that will have a small usage fee. In order to participate in the iQ Engage program, however, it is necessary that the customer also is part of:
These are required because we need both the SmartTagging feature that comes from participating in Content Exchange and the iQ DMP Program, as well as the Data Collection from the iQ DMP Program.
How can I troubleshoot audience setups that I don’t have?
All audience targeting and behavioral content recommendations operates on local storage, data which you can change for testing purposes. To see your local storage data, use Chrome and open the Chrome DevTools - (See Chromes Tutorial Website). Then go to Application / Storage / Local Storage and click on the Name of the website.
You’ll see key called tncms.dmp.audiences. This is where the audience data for content is stored. For troubleshooting, you can edit the value to temporarily place yourself into an audience. Simply click on Edit Value.
If you have no audiences at all, you can create a new key for:
tncms.dmp.audiences
And enter your own Audiences.
You can also use this as a starting point:
{"audiences":["keyword:art-subscriber","keyword:art","keyword:art","keyword:arts and entertainment","keyword:baseball","keyword:coach","keyword:donald j. trump","keyword:donald trump","keyword:economics","keyword:entertainment","keyword:institutes","keyword:social services","keyword:sports","keyword:tournament","keyword:trump"],"expired":1521746680000}
Behavioral Content Recommendations
Can I target local audiences in Behavioral Content Recommendations?
Behavioral Content Recommendations uses a different data warehouse of topics and keywords that is more specific to content, and all media sites in the iQ Engage platform use the same data. There are a few reasons for this:
We wanted to ensure that the audiences available match with the semantic keyword topics being added with SmartTagging.
Audiences added to a reader on one site will be usable on another site, providing more opportunities for targeted content.
We wanted to increase the differentiation between 'advertising-related' keywords and 'content-related' keywords.
Having a Data Set of content-related topics and keywords helps increase chances of content matching.
That being said, it is possible to create new audiences if it is a viable use case. Please contact us and let us know about your ideas!
How do I add Behavioral Content Recommendations to my site?
This is done by placing a configurable block onto your site, in the location where you want the stories to show.
This new block is called Card: Grid Behavioral and is available under the Static block type in the BLOX Block Editor. It is available for Flex templates only. This block offers lots of configuration options, like most blocks on the Flex Templates. You can change the background color, number of stories displayed, or block title.
In addition, this block allows you to choose the type of algorithm used to display content. There are three choices:
Behavioral Only: With this option, it will show only “behavioral” content - that is, content based on the user’s preferences. If no preferences are discovered, it will not show anything at all.
Behavior with Fallback: Or, you can choose to have a “contextual” fallback so that stories related to the current article are shown if there are no user preferences.
Contextual Only: And lastly, you can choose to show contextual content only.
NOTE: When someone comes from a social media site looking for a specific article, they will be shown contextual first, then behavioral based on an algorithm by TownNews.
It is possible to have multiple blocks on the page showing a mix of the above options.
How does iQ Engage know which content to match to readers?
The Behavioral Content Recommendations use user signals (as described above in the iQ Engage section) along with the following:
Semantic Data: We also use machine learning to extract specific topics and keywords from articles (even if those topics are not specifically mentioned in the article). For these topics, we use a lexicon of subject-matter tags so the data will be consistent everywhere it is used.
Recommendation Algorithms: The TownNews.com iQ Engage system uses the user signals on a profile together with the semantic data to surface content that this specific user may find interesting. It uses predictive analytics to see what the user has viewed in the past in order to surface items the user will be more likely to engage with in the future.
There are several factors involved in this algorithm, including:
Signal Strength: There are varying levels of “strength” of different signals. A certain signal (like signing up for a “Sports” newsletter) may indicate a very strong interest in a topic, while other signals (just reading a single Sports article) may be weaker hints. The TownNews.com iQ Engage system recognizes these varying strengths and accounts for them appropriately.
Popularity: Popular topics that have a history of performing well are be boosted to improve user engagement rates.
Recency: Articles that are more recent get a boost so that the newest content is more likely to be shown, if possible.
Context: Articles with contextual relevancy may be surfaced if the algorithm determines it will be better than the signal data for this specific user.
Are there best practices for behavioral content recommendations block placement?
If you want to target local, frequent users with behavioral recommendations, the best place to add this block is to the bottom of the article. This way, the user will have more options as soon as he or she is done reading.
Other possible recommended options:
Right rail on article pages
In-article Utility Regions
But, it is always the case that, depending on your specific users and site design, that you may have better results adding the block somewhere else on your article pages. Feel free to experiment with the various Utility Regions for assets to try new things.
Dynamic Targeting & Custom Messaging
What is Dynamic Targeting & Custom Messaging?
This concept describes the process of a media organization segmenting their users based on audience, and targeting specific experiences (blocks of content, ads, etc.) to those specific groups. The experience is dynamic because it is different for each user, depending on their audiences.
For example, a media site may want to show more ads to “drive-by” users (users who have rarely visited the site), and more content to loyal users (users who visit the site frequently). They can do this simply by creating the relevant audiences, and then targeting blocks of content or ads to those specific audiences, and hiding blocks from other audiences.
What are some use cases for Dynamic Targeting & Custom Messaging?
There are many things that can be done with Dynamic Targeting. For example:
Audience Segmentation: This involves splitting the audience into four groups:
In-market, High frequency: Local, loyal users. Target them with more behavioral content blocks because they are very likely to re-engage with content they like. Show them local ads because they are more likely to be interested in local advertising offers.
In-market, Low frequency: Local users who don’t visit often. Target them with promotions to get the to come back again - sign up for newsletters, sign up to subscribe, more behavioral content they they may like. Show them local ads because they are more likely to be interested in local advertising offers.
Out-of-market, High frequency: Out of market users who have an affinity for the site for some reason. They may be former residents, or they may have a particular interest in the region (like coverage of a sports team, or agriculture coverage, etc.) Target them with more content that they may like, to get them to re-engage.
Out-of-market, Low frequency: Out of market users who are unlikely to visit the site again. These will be the users that have no reader signals, or audience data. These are likely to be users that are coming in from social media or another website to see viral content or a single article. They are unlikely to come back anytime soon. Target them with more programmatic ads.
These “segments” can be any audience available to the site. Such as men versus women, young versus old, in-market versus out-of-market, etc.
Custom messaging: You can hide subscription messages or offers from users that have already subscribed. You can import your subscription information to create audiences of users who are known subscribers. Then, place subscription offers on the site, but hide them from known subscribers.
Show local news blocks to local users: If you know the user is out-of-market, replace Business Directory promos with programmatic ads.
Promote special email newsletters to users in that interest group: If you have a newsletter specific to entertainment, art, sports, etc., you can target users who are interested in that topic.
Can I target local audiences in Dynamic Targeting & Custom Messaging?
Yes! These audiences pull from your local data set, and can include any audiences you need for your specific site.
How can I create new audiences?
If the user is part of the TownNews.com iQ Enterprise Program (where they have full access to their DMP data), they will have access to their DMP admin and will be able to create audiences themselves. If they still need help, a ticket can be sent to the DMP Program Manager.
If the user is part of the TownNews.com iQ Standard Program (where they use our DMP to serve programmatic ads on their site), they can still use the iQ Engage features. To create a new audience, they must ask us to create it for them. A ticket can be send to the DMP Program Manager who will assist.
How do I set up blocks to use Dynamic Targeting & Custom Messaging?
This is the current list of blocks which support Dynamic Targeting:
Business card: Business
Business card: Grid
Business card: Offers
Utility: Email signup
Utility: Text promo
Card: Grid
These blocks all have a new block property group called “Audience targeting.” That group has two properties, one to add audiences to whom this block 'will be shown' and one to add audiences from whom this block 'will be hidden.'
A few best practices for using Audience Targeting on blocks:
If there are multiple audiences, simply separate them with a comma. For example: "trump, politics, art."
These blocks, regardless of the Audience Targeting setup or audiences on the admin users, will always show in the block editor to make it easier to troubleshoot.
If there are conflicts (i.e. the same audience in both “show” and “hide” targets, the block will be shown.
How do I set up ads to target to specific audiences (as described in the audience segmentation example)?
This would use the standard process for the TownNews.com iQ Enterprise or Advanced Program using Google GAM.
