The 265 DriverTags (psychological content codes) placed on television shows were distilled by machine learning and other data science techniques from all the words in the English language. The recommender using the RMT system achieved 18% average conversion to loyal viewership of a series never watched before, and users gave it over a 90% satisfaction score. Netflix website indicates a 70% satisfaction score for their method, which we feel is the second-best method the industry has.
Here is a one-minute video demonstration of one form of the system we call dlitr (pronounced “delighter”).
Here are further details:
Win/Win for Viewer and TV Industry through “the Pandora of TV”
dlitr
Executive Summary
- A disruptive, proven and scalable solution for a $16B US market (network TV tune-in advertising)
- dlitr solves problem of gaining rapid awareness of new programs now causing unaffordable “program infant mortality”
- Solves a troublesome viewer situation facing viewing choices
- High credibility and trust Team
- dlitr is the means of attracting large sample of people willing to contribute their audience data to get benefit of one stop shop “knows you” program recommender
Prologue
Consumers need a personalized one-stop-shop way to decide what to watch on TV/video. Networks need an affordable way to gain new viewers for their programs. This free-to-consumer app solves both problems. It’s also the first form of “pull” tune-in advertising, and the first pay-for- performance tune-in.
The RMT Recommender was successfully pretested by Liberty Media in 1997 as part of “the first Hulu” called Your Choice TV which was too far ahead of its time and did not survive. The time is now right for this program recommender, the only one based on science, not just common statistical tools and publicly available data.
It Takes 23 Minutes a Day to Consult Program Sources
This is 1.3 years out of a person’s life who dies at 80.
Source: Ericsson’s 24-country study.
Choose it or lose it: Media choice abounds, but many Americans stay with what they know.
—Nielsen Newswire, July 1, 2019
⇒ “58% give up and go back to their favorite channels”
⇒ “21% decide not to watch and do another activity”
⇒ “Only 26% check today’s recommenders”
“Most Recommenders are mediocre.”
—Michael Collette, when running data strategy for Vizio
Rearview Mirror-based Program Recommendations
dlitr
IT GETS ME!
Clear-cut Advantages
- Other recommenders take about 8 weeks to build up the duplication data they need to recommend a new show
- Many new shows are unofficially cancelled by then
- RMT has predictive data before the premiere
- Collaborative Filtering/Duplication ignores the average 70% non-awareness of new shows
- It also ignores the lack of trial of new shows among the 30% aware
- Duplication based recommendation reflects less than ten percent of the potential audience
Recommender Field Trial Results
- Field trial in cable homes
- Recommended programs never watched before according to set top box data
- > 90% user satisfaction
- Most common user verbatim: “IT GETS ME!”
- 18% conversion norm 3% for tune-in ads
Why Better Than Today’s Recommenders
- DriverTags developed scientifically
- >13,000 psychological words found in English
- 1562 clusters based on 10,000 US sample, rating how well each psychological word describes the respondent
- 1562 key words used to metatag TV shows
- Set top box used to measure programs already viewed, created psychological profile
- Recommendations that caused viewer acquisition reflected 265 of these metatags, dubbed DriverTags
DriverTags Proven in BTB
- Sales effects of TV ads increased +36%
- Purchase Intent increased +37%
- Unaided Awareness increased +62%
- A single Resonant exposure lifts branding effects 5X among low frequency group
- Among all 3830 brands in Simmons study, DriverTags increase ability to predict brand adoption +83%
- +52% increase in ability to predict Nielsen ratings
—Sources: Nielsen Catalina, 605, Simmons, David Poltrack
<h2Who Uses RMT Technology?
RMT Solutions – dlitr (“Delighter”)
“One Stop Shop All Televideo Choices”
AI Consumer Viewing Recommender
- Mobile app supporting consumer “TV Everywhere” choices
- ALL TV
- ALL OTT
- ALL Digital Video
- Features RMT’s proprietary DriverTags technology + ACR
- Predicts consumer acceptance of specific video content
- Supports situation specific content selections – Cross-Platform.
dlitr (“Delighter”) vs. Reelgood
| dlitr | Reelgood | |
|
|
Tune-In Advertising in U.S.
- $14B/year worth of airtime used for program audience promotion
- $2B/year paid media for tune-in
How to disrupt and improve this $16B industry?
Viewer Requested Tune-in: dlitr
“Ridiculously high cost per conversion (CPC) when networks buy off-channel TV” (dlitr far lower)
—Howard Shimmel, former CRO, Turner