From the Garbage Day newsletter, some reflections on the connection between virality and popularity (highlights mine):
The conclusion I’ve come to this summer — one I’m still not totally sure I fully believe yet — is that what’s really happening here is that virality is decoupling from popularity. And I think you could even argue that the very idea that mass appeal had to be accurately reflected back at us online and vice versa was an entirely millennial idea. A neurotic need to know, and quantify, exactly what everyone else was seeing and doing. A panopticon for the overachieving generation. But I also don’t think we’re returning to the 2000s decentralized content badlands either. Instead, there will be silos of popularity, online and off, global and regional, real and fake, and none of them will quite add up correctly, but all seem vaguely huge. It’ll be confusing, but, ultimately, I don’t think anyone will really care.
This resonates deeply with what I perceive out there. I remember sitting in meetings throughout the year where for election or other campaigns, where people (including, to a degree, myself) were going nuts over how well you could track interest and engagement online. Finally we truly know how many people really like this idea! Or barring that, how many people clicked on this link or that button.
By now, of course, we know that a lot of those analytics (read: behavioral tracking) turned out to be either garbage or a bad idea to begin with that is killing the web and should be killed with fire.
But then, it seemed exciting and useful. To a degree, in that particular context, it was (probably). That said, even then it never told the full story, for two straightforward reasons:
- While clicks can track certain types of transactions, none of these analytics could tell about the underlying motivations. Do people search for a product online because they are interested in buying, or learn about a scandal, or just out of curiosity? Do they click a link to a political campaign because they support the party or are they hate-watching campaign speeches? No way to know, but that often got swept under the rug, and the various stakeholders involved would simply divine meaning where there was none, invariably to support their desired narrative.
- The addiction to numbers and dashboards was so consistent that people would use whatever numbers they got, even if they were practically meaningless. For example, in the Golden Age of APIs (the early 2010s), many popular web services and social media platforms allowed gathering data through APIs. “A neurotic need to know, and quantify, exactly what everyone else was seeing and doing.” Then one by one, platforms switched off those APIs to build higher walls around their walled gardens. At the same time, more and more interactions moved to smaller, more private group chats: your Discords, Slacks, WhatsApp groups, etc. Those very few services that were still “readable” by analytics tools, that were still searchable, that still shared numbers suddenly were vastly, disproportionately relied on for those same dashboards. Only they were even less representative of what was going on than before. So more and more, decision-makers were flying blind despite having a lot of numbers and fancy graphs in front of them.
Which brings us back to the quotes at the top: “There will be silos of popularity, online and off, global and regional, real and fake, and none of them will quite add up correctly, but all seem vaguely huge.” This rings true to me. I think that despite a crushing, overwhelming, in my opinion hugely destructive online behavioral tracking machine, to a large part we won’t be getting any useful numbers to measure popularity. We’re back to using proxies. And frankly, it’s fine? It’s good enough? I have a hard time coming up with good examples of why we’d really need more. In truth, the numbers never really added up to begin with, and on the internet, everyone is famous for 15 people.
