Constant improvements to the platform abound - with this release we follow our self-imposed RERO directive (Release Early, Release Often ...and listen to our Customers).
Given the huge release schedule we have in store for H2 2019, we need to ensure the 'deployment runway' is clear of technical debt, with all bugs squashed.
- Improvements to translation & translation management
- Bugfix: queries counter 'flicker'
- Improved parsing of Legacy modules
- Improvement: Site Group overview no longer keeps trying to load sites without any modules
- Improvement: View Mode handler improvements - increased View mode switching/handling
- Improved intermittent network issue handling - more an edge case, though it's the internet, it breaks sometimes !
- Improved Automated error reporting - ongoing efforts to ensure we can focus on the problems, not just 'noise'.
- Future proofing for code lint handling
Lastly, a note on industry language and perception - Individualization vs Personalization. Think you know the difference ? Read on to help us clarify and reclassify popular thinking!
Improved parsing of Legacy modules
We continue to add support for Content Module configuration edge cases, though with an upcoming release, we will reach parity for almost all legacy Cxense Content modules via these two powerful configuration options:
Initial Design Prototyping for Per-Position Content Collections - Build Collections of content and assign them to individual item positions.
NOTE: This is an internal prototype - production functionality will no doubt vary
Cross Site Recommendations
For those with multi-site networks, Cross Site recommendations allow you to increase revenue-generating traffic to content that may not enjoy such exposure. As with all Cxense personalized recommendations, we ensure the best of the best for each individual's experience.
Cxense's deep awareness of each and every piece of content ensures we can identify content and context relationships that other will miss due to simplistic (or non-existent) content awareness. This occurs across sites and sections, utilizing our Natural Language Processing, ML / AI-based categorization, and much more smarts built into our platform.
Per-Position Content Collections
This is the ability to define what content is delivered to each individual position or slot of a multi-item recommendation module. Perfect for dialing in a 'metered' blend of content themes, or native advertising.
Whilst the process is relatively advanced to balance revenue objectives, ensuring the management of per-position collections is super simple (think: drag & drop), Content Curators can easily balance the varied objectives of CTR and general engagement with other business intents - direct revenue from Native Advertising, increased exposure for specific sections for awareness - or revenue for that matter (e.g. boost traffic to high CPM content).
Suffice it to say, we are very excited to offer these advanced capabilities within the same simple intuitive User Interface used by hundreds daily to manage their Audience Engagement and Value generation. We look forward to general availability of these powerful value-adding features.
Individualization vs Personalization
We often hear our competitors and those offering 'recommendation solutions' referring to 'personalization'. To ensure we offer genuine evaluations and comparisons of our solutions, we also refer to our platform as having 'personalization' capability (just look to the author of these release notes!).
That said, let's be completely clear for the record:
The Cxense Content Recommendation platform utilizes real-time decisions to deliver Individualized recommendations for each individual User, not generic 'personalized' recommendations that rarely vary for each individual.
This results in superior engagement and content discovery - key to great ongoing engagement and beneficial audience habits, in turn delivering tangible value and revenue for publishers.
One could argue recommendations were personalized if you take a single context input into account prior to choosing the delivered content, such as 'User Location' or 'article category'. This is possibly just enough to be promoted/marketed as 'personalized content', though we feel this is overly simplistic.
Most everything in the Cxense platform is based on real time decisions for each individual. Far superior to 'simple personalization', though we still refer to this as personalization as we feel that was the initial intent of the term - to make it personal. We feel it pertinent to clarify and delineate to ensure we're not bundled in with the lesser capabilities of non-individualized solutions.
If a Recommendation platform refers to personalization, take a moment to validate - do they mean content delivered based on the Individual based on real time decisioning, or the lesser, generic pre-defined segmentation ?
Allen Albot from Navistone generated a nice high level graphic, and related post post on this in 2017, whilst LiftIgniter posted a more detailed article which clarifies the comparison elegantly.
Our experience with Individualization for our customers shows time and again how important it is to move beyond generating experiences for the audience of 'one' (static websites/apps), and instead engaging the all important Individual.
Happy Individualized engagement !
Simon & the Personalization Team
Comments
0 comments
Please sign in to leave a comment.