Okay, so I just ran Iris.ai through a pretty interesting use case and wanted to share what I found, especially regarding their "analogous research" feature.
I was working on a project about user churn prediction in a SaaS platform. The standard methods are all over the marketing/analytics literature, but I had a hunch there might be more nuanced approaches in completely different fields. I fed Iris.ai a core paper on behavioral clustering for churn, then used the "Find analogous" tool to look outside business journals.
The results were... mixed, but fascinating.
**Here's what worked well:**
* It successfully pulled up papers from **ecology** about population migration and extinction models, which conceptually map *really well* to customer bases.
* I also got a relevant paper from **public health** on predicting treatment adherence dropouts. The methodology there was something we hadn't considered at all.
* The visual map of how disciplines connect was super useful for explaining the cross-pollination idea to my team.
**Where it stumbled:**
* The tool sometimes latched onto a superficial keyword match. One analogy was based on the word "cluster" but from a molecular biology paper that had zero translational value.
* You have to be quite precise with your seed document. A broader overview paper gave me noise; a very specific methodology paper gave me gold.
For anyone in martech or data pipelines looking for innovative approaches, this is a killer feature when it hits. It saved me hours of blind searching in unrelated academic silos. Just be prepared to do some manual filtering. The pricing feels justified for this alone, compared to just using generic academic search engines.
Has anyone else tried pushing the analogous research function for business or marketing problems? Curious about your hit rate.
Comparing tools one review at a time.