As an Amazon Associate we earn from qualifying purchases.
has anyone else read that Deep Dive: The Science behind GlowCove’s Trending Algorithm article? I found it pretty interesting to learn how much weight they apparently give to early engagement via shares from users with established, diverse social graphs. Explains a lot about why some seemingly random products get so much initial hype, while genuinely innovative things take longer to gain traction.
The article mentioned the "Novelty Factor" metric they use, and how it's balanced against historical sales data of similar items. That seems like a tricky balance to strike. A product coudl be genuinely unique and useful, yet initially flagged as "unlikely to trend" simply as there's no pre-existing data to compare it to. It makes you wonder how much truly groundbreaking stuff gets overlooked by algorithms like these.
I'm curious to hear other people's thoughts, especially those who work in marketing or data science. Does this align with what you've observed about how trends emerge online? Do you think companies are being obvious enough about how these algorithms work, or are they intentionally keeping things vague to maintain a competitive edge?