3 Ways AI is changing Influencer Marketing.

Influencers, an overview.

 

Influencers have been a key part of marketing campaigns since the early 20th Century. Brands created personas to emotionally enhance purchase decisions of consumers. Santa Claus appearing for Coca-Cola, is probably the most infamous influencer followed closely by Tony the Tiger for Kelloggs. 

Santa Claus, the first and best known Influencer.

Then came a wave of celebrity influencers promoting products could be seen everywhere. Celebs would be discovered by a scout and then promote something, encouraging people that it is the best thing since sliced bread. 

Fast forward 80 years to the early 2000’s and anyone who is anyone was starting a YouTube channel or Instagram account. Small numbers of people became internet sensations, promoting products and blogging became their job.

The accessibility of social media meant that in 2016, there were 9.7 Million sponsored influencer posts. By 2019, its likely that 32.3 Million sponsored influencer posts will have been created.

Since the dawn of the influencer the messaging hasn’t changed, influencers still trigger customer decisions through an emotional connection. They create a feeling of aspriation to be like or associating with an influencer.

Today, we have the InstaFamous crew that dominate many a social feed. We know and love them, but they are slowly being replaced with Artificial Influencers.

 

Gorillaz a virtual band created in 1998 by musician Damon Albarn and artist Jamie Hewlett. 

From Gorillaz to Hatsune Miku, virtual celebrities/influencers have been around for decades. But now, they are starting to be involved in product promotion.

The main questions are  — can AI Influencers still trigger similar emotional responses? Are computer generated bots able to create the same sense of aspiration or association? 

What are Artificial Influencers?

Artificial Influencers are characters developed with psychographic (what is used to describe consumers personality, values, opinions, attitudes, interests and lifestyles) and contextual relevancy for brands to identify audience, profiles and personality.

They look a little similar to avatars in the Sims, but a little more lifelike. However, AI Influencers are not restricted to appearing inside a computer program like a Sim. They can be placed in a real life situations, with real people and with real clothes. IRL. 

The Sims, a favourite past time, but also a program shaping the future. 

They are computer generated, meaning you can create a fully customised influencer for your target audience. This includes things like hair colour, body shape, skin colour and clothing, all the way to their political views and sexual orientation.

How are Artificial Influencers created?

Humans (or their creators) will need to select the information that the system learns from in order to shape AI Influencers. A lot of data will be needed to train AI beings at the beginning. 

Later in their life cycles, bots will use machine learning meaning they will be able to adapt to settings and scenarios, expanding what they can offer to their audience. Although they will always grow in knowledge and evolve with the times, the crucial feature is that they will always controllable by software. 

Could a situation like an outspoken Kanye have been stopped if he was an AI influencer? 

How are they being used now?

Today, there are thousands of Instagram Influencers online who blur their reality by excessively editing photos and adding filters. This causes us to compare ourselves to their unrealistic existence leaving us feeling slightly deflated. = Problem for brands.

Creating an influencer that appeals to multiple audiences and finely tuned to be aspirational and associational is an art form. It’s very challenging to find a human influencer who can do this. This can, however, easily be achieved with a AI Influencer. 

Take Lil Miquela, she has collected over 1 million followers in 2 years and established a cult like following called the ‘Miquelites’.
Her average post reach is 224% higher than the influencers who have a similar number of followers, She’s is taking the internet by storm. 

But, Miquela isn’t real. 

Since her debut, she has been pictured with a number of celebrities, interviewed by a number of publications, been on the cover of a few and collaborated with well known fashion brands. She has also, allegedly, been romantically involved with a musician.

@lilmiquela and @nilerodgers on Instagram.

Controlling a bot on a computer means a brand can be much more refined with what they want to include, exclude or highlight without making the image look completely artificial. 

The non-judgemental nature of bots often makes users more honest with conversation and feedback. This gives brands far richer data than conventional exchanges.

Miquela’s identity is very real and down to earth. She is transparent with all that she posts and never poses as something she isn’t. Miquela has become a generational internet role model.

Her transparency allows a wide audience to connect with a variety of topics. From her political views, humour, makeup, fashion and music. It represents how a platform of influence is being used for positivity, not just indulgence.

 

Why are they useful?

Trust

A staggering 84% of millennials just don’t trust traditional advertising, meaning there is a big lean towards influencer content.

Advertising has always been most successful when delivered by word of mouth. Over time this has moved from being spoken word to online conversations and recommendations. People rely on people they don’t know to give them recommendations. 

The difficulty is that Millennial’s and GenZ can tell when promotions are based solely on a contract.

Brands are trying harder to keep content relevant and targeted to audiences as amounts of content posted online grows exponentially.

Having an influencer who is not obviously promoting something and completely transparent, like Miquela, will aid the brand in building this trust Win for the brand.

 

Brand Identity

Some human influencers may not match the overall brands image and miss the mark. Take a computer generated bot and you can create someone who will be the most relevant influencer for your audience. They can be kept on brand and directed as to what to say and when to say it. This limits the opportunity for an outspoken Kanye situation. Win for the brand. 

Connor Blakely CEO of YouthLogic “As an influencer making sure you believe in and can stand behind not only the product but the brand you promote is crucial to trust and credibility with an already skeptical audience.”

When influencers are showing the world something and not really engaging it becomes very one sided. Learning comes from listening, so AI Influencers are creating space where they can listen and adapt from what their fans say. Bots have the ability to adapt through machine learning and also are less precious about their ‘Brand Identity’. They are much more workable than celebs. Win for the brand.

 

Pricing

Influencers are becoming more and more expensive, meaning that to hire multiple influencers to target multiple audiences becomes an expensive task! Adding to this, with more influencers on the scene, traditional influencers are now looking for longer contracts again hiking up the cost. 

In 2017, Influencers were paid over $1 Billion to post on Instagram. In 2018 its predicted to be $1.6 Billion and $2.4 Billion by 2019. 

Miquela has never been paid to ‘wear’ pieces of clothing within sponsored posts, but has been sent free stuff from brands. Sending or creating a computer file to wear is significantly cheaper than sending a real clothing parcel. There may come a time when AI influencers, or their humans, charge per post. For now, it seems to be a significantly cheaper solution! Win for the brand.

What’s next?

AI Influencers didn’t create the conditions that allowed for them to flourish, we did. Seeing the level of influence Miquela has, proves just that. She is a creation that has been engineered to succeed. 

It’s hard not to see and refer to these bots as people as they are so lifelike, they appear to go through similar challenges and are just as real as some of the other “real” influencers online.

Below are four images of influencers online. Which ones are real and which ones are AI? 

 

Instagram is increasingly becoming a medium for heavily photoshopped, filtered and constructed identities. The line between real and not real has already been blurred.

When the “real” entertainment industry is so unrealistic, why is creating an unreal celebrity and it’s success be such a surprise. 

According to Forbes, 53% of brand marketers are planning to adopt this strategy in the coming two years.

I expect that there will be a number of AI Influencers appearing on our feed, creating emotional engagement with the ability to relate, be aspirational and associate with a much wider audience. 

It all poses a wider question: Should it matter to brands and publications if an influencer is computer-generated, if the avatar has the same influence on its following than that of a “real” influencer?