We predicted the Winner using AI not Facebook likes

It was 8.15pm Sunday 7 July 2019 and The Voice Grand Finale was broadcasting live on TV.

I had my favourite finalist and submitted my 6 votes online. I was one of hundreds of thousands of consumers of live music entertainment who participated in this ‘survey’. It could have been considered a massive real time online survey / poll.

They said it was “the tightest voting result in the history of The Voice competition”. We can’t prove or disprove this statement but I would hazard a guess that it was definitely a close race. Not sure if it was really the tightest in the history of the show, now in its 8th year.

It doesn’t matter. That’s hype. That’s entertainment. That’s marketing.

This article demonstrates how 

Social Media does not provide REAL ACTIONABLE INSIGHTS into how consumers engage with a brand, a product, a topic or issue.

Facebook metrics (#likes, #followers,#shares, #comments) do not measure deep engagement where it matters.

Does 1 million facebook followers mean you have deep engagement ie. where consumers will actually buy your product? No.

Does increasing your facebook likes by 50% mean an effective marketing campaign, which should ultimately translate to increasing your sales (short term) and your brand (long term)? No.

Does listening to a few heated comments on facebook about your product warrant you to invest in refocusing your product or service delivery? No.

Allow me to explain my negative nancy-ness for Facebook. I’m not against Facebook at all. I’m against using the wrong data to make important decisions. There is better data and advanced analytics tools to measure and predict consumer sentiment and engagement.

It’s like drinking a soup with a fork – why would you do that? There are better ways. It was like referring to newspoll and roymorgan surveys to tell us that the ALP were going to comfortably win the recent federal election. Epic fail wasn’t it.

How many times do we need proof that surveys do not predict consumer sentiment or measure engagement.

Let’s come back to the recent example that was played out tonight with The Voice Australia Grand Final 2019. This was a nationwide survey of over 100K consumers of entertainment to vote for their favourite.

The Setting

4 finalists – 3 males, 1 female. 3 males were aged in their 20s. 1 female aged over 30.

Viewers were given 6 votes in total.

The Current Way – Facebook

The traditional methods to measure engagement use social media – likes, followers, shares, comments. These are metrics that marketers rely on to measure engagement and marketing effectiveness. So let’s take a look at the Facebook profiles of the 4 Finalists:

No alt text provided for this image

According to #Likes on Facebook, the winner should have been Daniel Shaw.

It also measured Zeek Power having more engagement than Diana Rouvas. He was actually in the bottom 2.

The Better Way – AI

At 8.15pm, an 1 hour before Channel 9 announced the Winner for 2019, I ran analysis on the 4 finalists and measured the level of engagement using AI. As proof I did this pre-announcement, I posted it on linkedin.

It was very clear from the results, who was predicted to win and it wasn’t Daniel Shaw.

The 5 points about the methodology and why AI was superior in measuring engagement than Facebook:

Data source: All open source, unstructured online content (includes social, video, mass media)

Scope: Everyone who uses the internet and engages with online content.

Analysis method: Artificial Intelligence including Natural Language Processing and other advanced analytics methods.

Predictive Insights: Measure Engagement through Emotional Depth (how people feel), Content Power (what is being talked about), Media Power (which channels have influence).

Speed: Analysed within the hour.

We predicted the winner was Diana Rouvas (Facebook said she ranked 3rd). She displayed, by far, the strongest and deepest engagement with the market. She was categorised as Transformational, which is the highest level of engagement we can measure.

Transformational engagement means intensely-engaged experiences with strong timeless themes that are dynamic, on the move, fast-changing, and potentially result in lasting transformation of the world. Diana Rouvas has Transformation Engagement.

This indicates that she has the power to change the world. She will not fizzle out as a fad. I would expect to see her brand around Australia in the future. Her name will be remembered. Having deep engagement means people will buy. People will buy her records and tickets to her shows. She is a brand that will grow and is on the rise. Watch this star shine.

AI for Marketing Insights is a powerful method to strategically assess brands, products and issues that matter to the market. More importantly, through measuring emotions, we can measure engagement, which predicts likelihood of behaviour change.

So next time you consider predicting human behaviour, please consider the data and tools available that are far superior than a fork.

Register for our launch event

Tuesday 15 December 2020, 6.30pm - 7.30pm AEDT


Maven is officially in flight and ready to take off - join us for:
• Dazzling data and AI
• Prosecco shipped to your desk
(exclusively for the first 50 registrations ONLY!)
• A chance to win a $10,000 Maven Data Research pack

• Plus much more...

Are you ready?