Our Chief Product Officer Matt Simmonds has published the second of a two-part series in PerformanceIn detailing our new launch of Intelligent Partner Discovery. The first article focused on our approach to finding great new partners that will drive strong conversions and revenue for brand partnership programs. This time he shares his thoughts on how AI and machine learning will transform partner marketing over the next eighteen months. Here is an excerpt from Matt’s second article:
In my first post, I detailed how Intelligent Partner Discovery represents a shift in how brands can expand their stable of partnerships with maximum benefit and efficiency. By applying machine learning to historic and real-time partner program data, we’re helping brands identify their “best next” partners based on their propensity to drive strong incremental conversions and revenue.
This feature represents an evolution for partner marketing which, for the most part, still operates using very manual, human resource intensive processes that are left over from traditional affiliate networks. Just as with the other channels of digital marketing, over the next twelve to eighteen months, partner marketing will transform to fully leverage artificial intelligence (AI), machine learning, and data to provide automation and optimization – freeing up those human resources to be reallocated to where they can add more value. Here are some ways this will happen:
Finding the best partnerships for your business
Brands will start to leverage machine learning-based tools like our Intelligent Partner Discovery to automate the process of finding the best partners to work with. This once manual process will expand to cover new partners such as influencers, who are now beginning to be tracked based on conversions rather than or in addition to engagement.
We will also see more automation in the tenancy buying process where brands have the ability to offer a tenancy budget to their partner base at scale. Partners can then bid on the budget and offer up placements. I wouldn’t expect this to end up as advanced as programmatic ad buying – but there are certainly options to use data science to automate what is still a manual process.
Going beyond the traditional partner types, I’d expect to see a data-driven approach used to connect brands with other brands. This will enable marketers and business development teams to be more certain of the potentially huge benefits of partnering with another brand.
Making sure brands are not paying out too much or too little
Partner commission strategies are often complex, especially with the amount of data available now. At Partnerize, there are brands that commission their partners based on multiple metadata parameters such as length of stay or loyalty tier in the travel vertical. But as complexity increases, creating the optimal commissioning strategy can become time-consuming.
Data science will help solve this problem by providing automated optimization techniques similar to those used in A/B or multivariate testing. Partner marketers will be able to use these auto-optimizing algorithms in conjunction with manual rules to make sure they are both optimising revenue and incentivizing their partners. This will also apply to optimising commissions throughout the funnel, so that partners that contribute to the sale, but not necessarily the last click, can be rewarded fairly.
Beyond commissions, AI and machine learning will be more widely employed to help identify and block potentially fraudulent transactions, particularly in emerging markets or with influencer partnerships where fraud can be more prevalent. This task is perfectly suited to a machine – by analyzing millions of transactions in real time, and applying the right machine learning, partner marketers will no longer have to worry that they are paying out for fake or inflated purchases.
Read the rest of the article in PerformanceIn.