Identify at-risk publishers with the predictive power of AI to reduce delivery issues and improve outcomes.
Since generative AI burst onto the scene in 2022, software companies have raced to build it into their offerings. These implementations run the gamut from smallscale productivity boosts (like LinkedIn’s ad copy drafting) to total workflow replacements (like Zapier’s natural language automation creation). With such a broad spectrum of possibilities, we took a long, hard look at how we could bring AI into the Convertr platform.
First and foremost, we’re pragmatists here at Convertr. Every decision we make is driven by what delivers the most day-to-day value for our users with no thought of what’s flashy, sleek, and trending (and let’s be honest – those latter three could be the tagline for AI right now). We want to add AI features because they’re useful, not just for the sake of touting our software as ‘AI-Powered.’ This led us to a goal – leverage AI to make an existing feature smarter, more proactive, and ultimately more valuable.
At its core, Convertr turns your campaign data into clear insights via the campaign dashboard and business insights tools. While effective, these tools are outcome-oriented, showing you what’s already happened. This became a clear opportunity to introduce AI – we want to empower you to leverage your campaign data predictively and proactively, enabling you to pivot and optimize your campaigns in real time.
With that in mind, we turned to a core use case for Convertr – publisher management. Marketers managing multiple third party lead providers know it can be challenging to stay on top of them. Publishers deliver leads erratically – large volumes one week and barely a few the next, making it hard to know who’s on track and who’s at risk.
The current solution is somewhat rigid – take the percentage of leads delivered and compare it to the percentage of campaign time passed. If the delivery percentage is greater than time passed, you’re on track. If not, you’re at risk. We’re applying linear expectations to a non-linear activity, making the insights less accurate than they could be. Enter – Fulfilment Risk Insights.
Our first foray into AI turns historical publisher performance data into proactive notifications, highlighting which partners are at risk for underdelivery before it’s too late to pivot.
The Fulfilment Risk feature will assign your campaigns to one of three categories:
A campaign’s risk category is determined based on a dozen data points (lead volume requirements, campaign progress, changes in campaign progress over time, just to name a few) compared over the course of 20 intervals across the campaign timeline. When we feed these data points across all of these intervals across all of your campaigns into the model, it starts to paint a clear picture of which campaigns are likely to succeed and which need your attention.
Let’s walk through what we’ve developed.
When you log in, you’ll notice a new section on the dashboard designed to alert you to any potential campaign delivery risks. This analysis is specifically focused on publisher performance and The model is trained exclusively on your campaigns, not across our client base, aiming to maximize accuracy and relevance by tailoring these insights to your unique efforts. These insights are also governed by user permissions, ensuring that individuals only see the data relevant to them.
From the dashboard, the ‘View Campaigns’ button will take you to a dedicated window highlighting which campaigns have medium and high risk publishers.
From this view, you can navigate directly to a specific campaign. The updated Publishers tab now includes a Fulfilment Risk overview to surface these insights.
Similar to the dashboard, clicking through to ‘View Publishers’ allows users to see a detailed breakdown of at-risk publishers and links to quick actions to review allocations or contract terms.
For more information on using Fulfilment Risk Insights and details on how it works, click here to view the full support guide.
For more content like this delivered directly to your inbox, subscribe here. We typically send emails twice per month and you can unsubscribe with a single click if you aren’t finding the information useful.
Product