Harnessing AI, community, and authenticity to unlock influence

Question & Answer

What AI or machine learning models and platforms can we use to identify new influencer trends, find suitable influencers or niche communities in Singapore? Can the platform also determine which influencers should be blacklisted?

[Evangeline] 
To identify influencer trends, find niche communities, and blacklist problematic influencers, use SaaS tools with NLP for trend detection, computer vision for visual analysis, graph analysis for community mapping, and fraud detection for spotting fake followers or engagement manipulation.

[Chris]
There’s no one-size-fits-all tool or platform, what we have done is built our own toolkit and workflow tailored to our needs. This includes social listening tools (for video, influencer, audience insights), as well as our own data (SocialHub Performance Score, social-listening) and feed into various LLM models to inform strategy. A few key things to remember:
1) There's no perfect one-size-fits-all platform or tool. Your brand strategy will likely have different requirements. 
2) Every platform is different, and they will always change. Your frameworks, strategies, will continue to evolve. And platforms will always play catchup.
3) Be a practitioner (even if it's on your own) and iterate. Create accounts just to understand your industry/consumers. The learning flywheel is far more important, than getting the trend right. 

What data points or metrics should brands focus on to measure the effectiveness of their influencer partnerships?

[Alex]
In terms of metrics, focus on video completion rates, shares and one-to-one messaging (if relevant) over likes and comments. Do also track growing communities on dark social platforms, such as WhatsApp or Telegram. And finally, regularly review your benchmarks because the platforms are constantly changing how they define their metrics.

Are we starting to see signs of influencers fatigue? Keeps getting bombarded by selling content these days.

[Evangeline]
It depends on how we execute the campaign. When you tie up with a famous or overpopularised influencer with a lot of sponsorships, it could sometimes deliver a bit of fatigue; while uncovering different niche communities (for example someone travelling from JB to Singapore or a business person in Singapore) can be a lot more interesting.

What trends do you foresee in influencer marketing over the next 5 years?

[Alex]
In terms of influencer trends, I predict more long term relationships with a smaller group of influencers who truly know, understand and love the product/service being promoted. AI influencers – both virtual and avatar – are here to stay, but they won't replace real people in our lifetimes. Most real people want to connect with other real people (for now, at least).

[Evangeline]
Totally agree with Alex, plus a few points: 
1) Influencer marketing will spread to more platforms beyond Instagram and TikTok. We’ll see more activity on platforms like LinkedIn, Spotify, Lemon8, and Red Note. 
2) Niche community leaders will also gain traction, building loyal audiences in specific areas. 
3) At the same time, we'll increasingly see the already-famous influencers start creating businesses and pockets of monetisation by engaging their community. 

[Chris]
With AI, we can only expect more content more creators across more platforms; even digital doubles where it's hard to distinguish. The line between consumer / creator will continue to grey out as we are able to find ways to work with more different niche influencer/creator & communities. Ultimately, regardless of scale, every consumer and creator/influencer, is a human relationship to build.