Influencer Marketing Blog

[Infographic + Video] The Future of Influencer Marketing: Predictive Modeling | CreatorIQ

Written by Brooke Hennon | Jun 16, 2020 2:42:00 AM

CreatorIQ hosted several thought leadership presentations during Social Media Week’s #SMWONE virtual conference, including this one covering how to scale influencer marketing using data science & predictive modeling.

Here are the key takeaways, and if you would like to view the full 40-minute session, the video is available at the bottom of the page.

#SMWONE Full Video Recording

 

If you would like to read or share the key takeaways, here is the plain text:

1. Scaling Influencer, Advocate, & Consumer Partner Identification via Lookalike Models

  • The Challenge: Influencer marketers spend far too much time searching for additional influencers that match the aspirational collaborations or existing relationships.
  • The Solution: To assist with the search & discovery of influencers at scale, and in a fraction of the time, CreatorIQ is leveraging millions of data points across creator performance, creator approvals/rejections, brand affinity, and industry alignment to build a creator recommendation engine that is trained to identify the best-fit creators for any campaign. Additionally, the creator recommendation engine is constantly evaluating available performance data and can be leveraged to quickly identify additional creators on-demand, based on current, high-performing creators.

2. Increasing Content Performance Confidence via Content Attribute Prediction

  • The Challenge: When thinking of content direction to use to brief influencers, influencer marketers sometimes use little science and mostly rely on subjectivity to think about what briefing criteria might make sense (within brand and campaign guidelines).
  • The Solution: CreatorIQ is leveraging visual insight engines to build data science models focused on the identification and recommendation of high-performing content attributes. We are leveraging best-in-class visual recognition engines such as Google Vision, to analyze tens of millions of pieces of content to build our custom models. These models can correlate specific visual attributes that are detected within the content, with the performance of the content, and recommend content that is likely to perform well with paid spend, for example.
  •  

3. Scaling the Targeted Reach & Frequency of Influencer Programs via Lookalike Audience-Powered Paid Media

  • The Challenge: Marketers want to maximize the reach of influencers’ best performing content - beyond the fans & followers of participating influencers.
  • The Solution: CreatorIQ is solving this problem with a model that leverages influencer data such as demographics, organic, and paid performance to drive the creation of lookalike audience seed segments that are used as inputs to social platform audience targeting. In our experience with these models, we have seen significant improvements in conversion data when compared to using standalone audience targeting.

4. Scale the ROI & Business Outcomes of Multivariate Influencer Programs via Predictive Campaign Guidance

  • The Challenge: Marketers want to fully optimize the many campaign tactics that combined will result in the best possible performance
  • The Solution: CreatorIQ is combining the multiple data science models covered above - targeted at influencer identification, content recommendation, and audience targeting - to deliver an active, outcomes-based recommendation engine that seamlessly allows our customers to optimize their influencer marketing programs and deliver significant increases in Return on Ad Spend.
  1.  

If you would like access to the featured slides in PowerPoint, Google Slide, or PDF format, please email sales@creatoriq.com