Invitation to join the game
Personalized games by country:
Argentina, Mexico, Colombia, Chile
Integrated "subscription" CTA
across experience
Re-targeting for users
who dropped
LIFT STUDY RESULTS
JTBD: “Drive Linear viewership and tune-in”
​
We run a multi-cell experiment in 4 countries simultaneously: Mexico, Chile, Colombia and Argentina.
​
As Amazon does not share their data with us, the main metric we used to measure against the JTBD was Reported Action (RA) with a question “Have you started watching El Presidente?”
​
With only a maximum of 28% of the total budget the cell with the bot campaign drove significant lift against the BAU:
RA
CHILE
1.5PTS
ARGENTINA1.1PTS
MEXICO
1.1PTS
ARGENTINA
0.9PTS
VS 1.1 LATAM/1.1 E&M
600k
People responded that started watching the show
4 OF 4
Countries, the bot strategy had statistical significant results on reported actions
2 OF 4
countries, the cell with the Bot strategy over performed de BAU Strategy.
ADITIONAL RESULTS
​
In most countries the campaign generated lift in Standard Ad Recall, in some cases well above the LATAM norm and in one case 2.0 PTS above the M&E norm.
​
Of the 121K+ messages the BOT sent, only 10% went unanswered.
​
45% of users click on the subscription button inside the bot.
​
The BOT had an 89.84% of responses, well above the minimum standard of 85% to be considered successful.
KEY LERNING
​
An immersive experience and ads specially crafts for our platform will perform better than Amazon BAU (trailers)
​
Also thanks to this project MarSci will push to do significant changes in the way on how to measure Amazon results in the future.