2. Social has changed the landscape
Before
1 video / market
1,000,000 media budget per video
100,000 testing budget per video
Today
100s videos / markets
10,000 media budget per video
1,000 testing budget per video
Low volumes, slow, expensive High volumes, quickly, inexpensive
4. Brands need better data at every stage
Need new metrics that have proven direct relation to business outcomes, including social performance.
Content Testing Media Planning Measurement
What to invest in? Who to show it to? How does it perform?
5. Power of Emotion Data
Scale
Emotions are universal
Passive measurement
Only need a webcam
Value
Know how your audience really feels
Emotions drive human thought and decisions
Emotions correlate to social performance
10. High volumes, quickly, inexpensive
Today
100s videos / markets
10,000 media budget per video
1,000 testing budget per video
Plan Measure
Content
Create Potential
Media
Target Results
Measurement
Optimise Live
Hinweis der Redaktion
Hello everyone, glad to be here today to share some of our insights into helping brands achieve social media success.
Social media success has proven difficult to achieve for many brands - we’ve moved away
from the TV-centric approach that advertisers had been perfecting for decades.
The landscape has changed quite abruptly ,
leaving brands largely unequipped to deal with the new situation.
The same money that previously went into a single video asset
is now often spread across hundreds.
Traditional approaches are struggling to deal with the new set-up.
Traditional research methods like surveys and focus groups work well when you’re testing a couple of videos in a few key markets.
When you’ve got hundreds of videos and half the turnaround time
these methods run into their limits.
Moreover the evidence around the predictive power of self-reported data is not always convincing – it can be difficult to establish how to act on these insights.
The fact of the matter is that brands can do better at every stage of campaign,
from testing content to planning media and measuring its impact.
Specifically in the social video space, there is very little testing and intelligent targeting currently being done , which makes success hard to predict and prepare for.
We need to learn to use the many valuable and scalable data sources available to us to maximise our efficiency.
At Realeyes, we’ve focused on emotions.
Using the latest advances in computer vision and machine learning we're now able to measure how people feel through their webcams - any regular webcam - making it very easy for people to share their feedback.
Emotion data is not only scalable but extremely valuable in predicting business outcomes.
In this presentation, I'll share some insights on how emotions can be used to drive social media performance in particular.
First a quick overview of how it all actually works:
All you need is a webcam and an internet connection – so we have access to large audiences, all over the world.
People watch videos on their own device, as they normally would, all Realeyes is to clearly ask if the viewer will share access to their webcam before they view the video.
We collect and analyse the emotions of those viewers who opt-in, reporting in real time how different groups of people really feel, and translating those responses to an overall performance score that we call EmotionAll.
That EmotionAll score is a result of extensive data science work at Realeyes.
Having measured the emotional responses of over 370,000 viewers watching over 2,000 YouTube videos we have found the exact emotional response patterns that correspond to higher social performance
The graphs here show how strongly that simple 1-10 performance score aligns with how many shares, views and tweets the videos actually got.
We’re able to clearly show that when it comes to social media performance, the more people feel, the more they do.
The simplicity of the performance scoring approach has been much welcomed by our clients.
Here's one case of a Heineken campaign we helped AOL optimize for their European distribution.
The campaign had 11 videos to distribute across 3 target markets,
in just a few days, Realeyes scored each video for each market,
as you see here for a few of those touch-points on this graph,
providing concrete targeting guidance
into which emotional connection points to invest in most.
Performance wise it's all about climbing the curves I introduced earlier.
Putting your media money behind the videos and audiences that elicit the strongest response,
will yield more results per dollar spent.
As you see in this Heineken campaign, moving from the lowest 'reds' to the highest 'greens',
multiplies the performance potential of the campaign plan,
and all that for no increase in media spend.
In this Heineken case we ran part of the campaign non-optimized
and the other part in emotion-optimized mode
and observed 12x more social actions per thousand views for the emotion-optimized part of the distribution.
The example I just gave for 11 videos in 3 markets is already more than most traditional approaches can handle, at least in the time required to deploy the findings.
Well, just two weeks ago we were asked to support a campaign of 120 videos in 5 markets - all that for a total production and media budget of not much more than a million euros.
It’s the new industry challenge, how to still effect intelligent planning at such scale and with such economics.
We at Realeyes believe the solution is in the combination of new tech and a radical redesign and automation of research workflows.
Emotions from webcams are the easiest way to gather such rich feedback on such a large scale, whether as part of slick panel tests; or measured directly, live, during the actual campaign distribution - something that is now possible, as of a couple of months ago.
By automating the interpretation of results and integrating those findings directly into campaign management, you’ve covered all angles and created an approach that can handle the new requirements we now face in video marketing.
It’s very exciting to be part of the changing landscape right now, as these building blocks are all finally coming together, and share the insights we believe will truly help brands master their social media success.