A New, Free Tool to Determine the True Influence of a Twitter Account

Rand Fishkin
Jul 11, 2018 · 3 min read

Number of followers is a crap metric.

And yet, as I’ve been watching and learning from folks who do content marketing, PR-focused outreach, social media marketing, and influencer campaigns, I’ve found that more than 90% of the time… that’s *exactly* what people use. They’ll nod their heads and say “yeah, I know it can be gamed,” or “yeah, some people still buy followers,” or “yeah, but what else is out there?”

So… darn… frustrating.

Here’s an example: my wife Geraldine runs the Twitter account @everywhereist. She has a respectable 72,517 followers. I run the Twitter account @randfish, which has 413,696 followers. From those numbers, you might assume that an average tweet sent from @randfish reaches ~5.17X as many people as a tweet sent from @everywhereist. That assumption is dead wrong.

In fact, oftentimes it’s the case that @everywhereist’s tweets reach more people, get more engagement, and have a much larger number of impressions than a tweet sent from @randfish. If you were, in fact, trying to determine which account you wanted to send a broadly-targeted message from in order to maximize the impact, choosing @randfish, simply because that account has 5.17X as many followers, would be a bad decision.

That frustration led my cofounder, Casey, and I to build a new, free tool: SparkScore

SparkScore is entirely free because our goal with it isn’t to make money directly, but rather to hopefully build some credibility for SparkToro as a company that produces high quality, more transparent, more reliable scoring systems for measuring influence. Longer term, as we build out our paid product (which is still 9–12 months away), we’ll be expanding and improving upon SparkScore, supporting more networks, and (hopefully) becoming a standard that many folks who need to compare the *true* influence of Twitter accounts can use.

Using our previous example of @everywhereist and @randfish, you can see that SparkScore correctly identifies the @everywhereist account as having more reach and influence:

We’ve also tried to make the scoring system as transparent as possible, so you can see how we calculate the scores (for example, we do use follower count as ~1/3rd of the overall score, because we’ve found it does correlate, albeit imperfectly, to total reach+influence):

Our final innovation is the one I’m most excited about — Engagement Score. This metric shows how relatively engaging an account’s tweets are, given its followers. No surprise that @everywhereist is incredibly high on this scale; that’s a big part of what makes that account’s SparkScore bigger than @randfish. And it shows that we’ve found a way to help address the catalyzing issue for the tool’s creation.

The tool isn’t perfect — this is a first version. But it’s already giving us useful results, and we hope it can be useful for you, too. Give it a spin, let us know what you think, and stay tuned for improvements. Oh, and if you find it truly valuable, it would mean a lot to us if you shared it with your Twitter friends 🙂


Originally published at sparktoro.com on July 11, 2018.

Rand Fishkin

Written by

@SparkToro founder, author, speaker, blogger, husband to @everywhereist, tiny Backstage Capital/Techstars investor, & feminist. Formerly @Moz.