 |
| Falls and rises—Klout scores have more elevation shifts than heliskiing in Alaska. |
This is a break in normal outdoor programming to describe a behind-the-scenes look at some of the first-world problems of working as a social media journalist.
Much of my livelihood depends on my social-ability. Klout is a company that is widely known in the industry as tracking that ability and measuring it. Many times, it is what is used to determine someone's talents and skills in engaging and interacting with people online and thus companies will use it to offer benefits and jobs to individuals.
My agitation continues to increase over this because Klout's scoring system appears to be completely arbitrary and unreliable. Although my score is quite good, it can't be trusted. Not because I feel I'm not rockin' at connecting with folks through social media, but because Klout doesn't practice what it preaches, nor does its score shifts accurately measure what is happening in real time. Or at all. So, although it may make opportunities for my livelihood a bit more difficult, it has come time to shut down my account. For those of you in my field, you'll understand that this is a hard decision.
The following is an open letter to Klout (and to all sponsorship & hiring personnel that use a faulty Klout score to determine what a "good" social media player is):
With a score of 65.7 dropping to 63.7 in a matter of days, I've decided to disconnect my Klout account once and for all. Faulty measurements, wild shifts, mysterious metrics & no interaction from a company which measures interaction? Trust lost.
I rely on my social scores because I am a freelance photojournalist with national magazines & major companies. I get hired for assignments and opportunities based, in no small part, on my social skills. Klout claims to accurately reflect those scores, but that claim goes beyond false advertisement; it's an outright lie.
My score drops the more I interact. To make sure of this claim, I conducted several experiments, the most recent being this week. I ramped up interaction, receiving increased likes on Instagram (some days even hundreds), a massive increase in likes and comments on Facebook and held Twitter conversations among people with Klout scores in upper 60, 70s and lower 80s. I returned to Klout each day to check my score hoping, beyond recent experience, that you would pull through and score my diligence accordingly. But no; my score dropped each day, losing over two full points in less than a week.
However, when I do my normal amount of tweeting/Facebooking and Instagramming, my score rises. So is your lesson: do mediocre work on social? Maybe I should just RT whatever you are doing; seems whenever I mention @klout in my tweet, my score mysteriously rises.
I have tried interacting a ridiculous number of times with you through your Twitter account with many people joining in the conversation. Still, you remained silent. Unacceptable. You are an interaction-measurement company. INTERACT.
Those of us who rely on companies digging our social-ability are slaves to a broken system that is Klout. You guys were at your best when you still had your interaction style graph (two years ago). You were consistent then. You were friendly to use. You were young and open.
Maybe you've gotten too big for your britches now, but many of us are getting turned off. You're mysterious and frustrating. Fix your stuff. In the mean time, see ya! You're a weight off my shoulders.
With no regrets,
Gina Bégin
Hallelujah. I'm no longer a number.
This is really interesting to me and reflects the problem with any non-transparent product. If you can't see under the hood you need to trust that the algorithm they're using is working. The only way you can test that is by doing something you think would be good for your score and seeing if it rises.
ReplyDeleteIt sounds like you're doing the right thing - though it's also totally possible that your recent efforts have been in some way counter productive even though intuitively they should be increasing your presence.
Either way it does not matter, because Klout won't let you under the hood even if it just appears to failing to capture good behavior in your Klout score its failing.
I gave up in Klout a long time ago because it just did not make sense!
ReplyDeleteSocial scoring platforms like Klout offer nothing more than a service that sells social impressions. They're essentially advertisers - they sell numbers to clients, who buy their services and then wonder why they get very little return from that investment. They don't measure influence - they measure noise and social signals. Scoring is to influence what Nickelback is to the music industry... ;-)
ReplyDeleteI found that it was bringing out the worst in me: so I stopped using it last year. Do I miss it? Not really, I just have more time to make cups of tea.
ReplyDeleteGina,
ReplyDeleteGood article, and great insight.
With any social measurement tool, one must always ask what THEY are getting out of it. There's no money in offering such a service to social media users. But there IS money in "Perks," which is the advertising units that Klout is selling to their real customers.
Just as with Facebook, where users are tricked into believing that they are the customer and not the product; Klout's users are not the customer, Perk-buying brands are the customer.
The trick Klout has very effectively pulled off, is convincing social media professionals, and the companies who hire them (hire us) that this bullshit ranking system is somehow important, not to mention accurate and meaningful.
I've been grouchy about Klout for some time, but last week I met my own nonsensical score fluctuation.
After about three months of vacillating, I finally killed off my Facebook profile, though I had to leave one of my profiles open to admin the 20 or so pages I have under management. I also killed off my Tumblr account.
I was well-connected and well-engaged on Facebook, and it was with elation that I killed that account off, and removed it from my Klout profile, along with my Tumblr.
Just as expected, within 24 hours my Klout score dropped from 66, where it had been hanging for the better part of a year, to 63. No worries, and who cares, right?
Well, the very next day, my Klout score inexplicably jumped up to 70, an all time record for me. Wait. What? How does THAT work? One more day, and I'm at 71.
Klout has been characteristically protective of their algorithm, and rightly so, as it represents the sum total of their IP, but it's also worthless, from a social measurement standpoint.
Gina,
ReplyDeleteI'm trying to do a better job of commenting on others' blogs (not just replying to comments on my own), so consider this a first, feeble attempt.
You're not alone in your sentiment, but I think some of your frustrations may be misguided. Here are a few things I noticed.
"my livelihood depends on my social-ability. Klout is a company that is widely known in the industry as tracking that ability."
Actually, Klout doesn't track social ability, it tracks the effectiveness of your reach or, as they say, your influence. Being good at social media is only a small part of being influential on social media.
"Faulty measurements, wild shifts, mysterious metrics & no interaction from a company which measures interaction?" "You are an interaction-measurement company. INTERACT."
Again, Klout isn't trying to measure interaction, they try to measure influence. Interaction is a piece they look at, but it does not mean influence. Spam bots on Twitter interact more than anyone. If Klout measured interaction, they'd have the highest scores of all.
"However, when I do my normal amount of tweeting/Facebooking and Instagramming, my score rises. So is your lesson: do mediocre work on social?"
Not sure I follow you here. Are you saying that your normal amount of tweeting is mediocre? If mediocre means a more natural volume of posts, I'd agree with Klout on this one.
I think it's important to remember that Klout scores aren't perfect. It's like saying you're "10 spaghetti noodles tall" and your friend is "12". I mat not know exactly how tall you are, but I have a pretty good idea.
It's very similar with Klout. While It's not exact, it's gives a pretty good, side-by-side idea of the influence/reach of a person's online presence. And keep in mind that they're not just measuring you and everything they measure is taken in context. If your metrics that would trigger a high score go up, your score might not actually increase if everyone else's metrics went up even more than yours.
Are there mistakes? Of course. Are there mistakes in Google, Facebook, Gmail, Amazon, etc. Yes. It's part of doing something automated at scale. Because Klout is a bit more personal, I think that's why people have trouble giving them any room for error.
It may sound strange, but your experience actually increases my trust in Klout scores. If I build an algorithm, I'm going to want to guard against people gaming the system to keep scores accurate reflections of who they are in normal situations. By your score dropping from trying to game the system and posting an unnatural volume of content to get interaction, I'm a bit more impressed with their ability to score and track changes.
Anyway, just two-cents from a fellow Aggie (if I remember right that you went to USU). Cheers,
Gregg
Glad I found your article because lately I've been asking some of the same questions. I was busy for a time earlier this summer finishing a new book. Once done I got to work marketing the book and working on my 2 blogs which ramped up my social media activity. At that time my Klout score was 62. A few weeks later I decided to check it out anticipating at least a slight increase and was stunned to see it had dropped to 58. The problem is that walking away sometimes isn't that easy. For example there are important annual awards for my niche that - for better or worse - rely on Klout scores as one of the criteria for determining the "winners," so not paying attention to Klout is like thumbing a nose at those opportunities. That said, I can't help but wonder if no one is really clear about how they come up with their scores how one is supposed to intentionally improve. Thanks for your insight.
ReplyDeleteSo when you act fake, your Klout score goes down. Seems like it's working to me. If you care about stupid shit like meaningless numbers on which you base your worth, that is. Klout score. Of all problems in the world, this is yours. Well, congratulations, I guess.
ReplyDeleteWow, angry Tyler. Simmer down.
DeleteNever said I acted fake. In fact, if I don't feel like posting, I don't. If I am not happy, I don't pretend to be. If you had done any reading at all before posting a very judgmental comment, you'd have seen that none of my posts are fake. None of my social channels have fake content. And again, read. I don't care about numbers. My future employers do.
This is not my problem. This is a problem of those of us who are employed by organizations who rely too heavily on numbers to determine worth of a potential employee. Earning a wage is my problem. That is a need for most everyone on this planet. Try reading the article again—and try to write with a little more courtesy; I don't come to your site (which I notice you didn't link to or provide an email; a common thing amongst those who aren't brave enough to back up what they come on to bash an author about) and glaze over the details of something you posted for your employers to read and then bash you with crude language because I didn't agree (or didn't really read it).
Respectful disagreements are fine. Try again.
Your Klout will drop the more you check it. Saw a Social Media Examiner (or similar blog) piece on it. Essentially, Klout knows you are checking out your score and they will lower it to try to make you do more. If you check repeatedly your score drops. If you ignore it for weeks it will start rising so they can send you a "New High Score" email and try to get you to start Klouting gaming again.
ReplyDeleteYou're 100% right about how stupid it is.
But, it is nice from my end to be able to see a Kout 20 vs a Klout 65.
Fascinating, Keese! I had no idea. No more checking in every 10 minutes. ;) (jk, I only used to do that.)
DeleteHi Keese, I'm afraid that's just not true. Your Klout score is completely independent of the frequency with which you check it, any behavior on klout.com or in the Klout mobile app. What impacts your score is simply how much reaction you elicit in the social venues you've connected to your Klout account or on your public Twitter account. The only exception to that is for celebrities who have verified Wikipedia pages but that's rarely applicable for people who have difficult to explain score fluctuations. So let's talk about those.
DeleteThe reactions you elicit on social media are tabulated for their frequencies and unique sources across different trailing time frames. The score is also transitive; a reaction from someone with a very high score will impact your score more than a reaction from someone with a very low one. So if you had some big post or tweet that elicited a lot of reactions (or simply reactions from someone with a very high score) and it boosted up your score, that boost has a half life and eventually goes away. It happens to me, it happens to everybody. This is not completely opaque, it is translucent though. Klout moments show your activities that elicited reactions and how impactful those reactions were.
I'll get into some of the technical specifics here. If you're getting a lot of reaction that should be boosting your score but isn't visible in your stream of Klout moments, that could mean one of a few things:
1) those reactions aren't available to Klout except through queries that require your authorization. Authorizations become invalidated if you change your password. (IIRC, that's in the OAuth spec: if you change your password, you invalidate any tokens for third parties such as Klout to access the service on your behalf.
2) authorizations also expire - easy fix is login to Klout or otherwise using that service to get a fresh authorization
3) permissions are set restrictively - e.g. you have a post on Facebook that's elicited a lot of reactions from your friends but permissions are set that prevent that from being visible
... there's probably some more edge cases but those are the common ones
Full disclosure: I'm an engineer at Klout, I don't work on the score algorithms or anything but I do enable the infrastructure that they require. I'm not authorized to post on Klout's behalf or cleared to write this with PR people or anything. But I wanted to share with you what I know. I'm sure there will always be people who find reason to hate Klout or complain about their score or whatever, I'm personally sorry to read of people getting so upset. I just wanted dispel some of the misunderstandings of how Klout works. I hope that helps.
Spidaman, the Klout algorithm is still vey spotty. I had a public FB post five days ago with 92 like and many comments from high-profile people. It didn't even register in Klout.
DeleteThe exact same thing happened to me a couple days ago with a post that gathered 74 likes, but got conveniently skipped over. WTF??
DeleteOMG...I totally agree. Even in the world of pet blogging, companies have drank the Klout Kool-Aid. The thing is that it is not accurate nor an even playing field. For example, my Blogger friends can attach their blog to be counted in the scoring while I can't connect my Wordpress blog. As a social media strategist, I am very intrigued by your little experiments and glad that you shared the results. I am more convinced now more than every that the score doesn't mean much. I just wish companies would stop thinking that it does.
ReplyDeleteit sounds like FoleyPod has similar frustrations so I know there are others out there that are too. I do like Gregg's point about Klout being about influence, not just sheer interactions, though. While I appreciate spidaman's points, I think it strengthens my argument that the Klout world is not a even playing field (or rather, not an accurate measure because someone's Facebook might not be counting while another person's is).
BTW...honored to be on #TeamSierra with you!
I just wish they'd quit sending "perks" like free magazines for 3 months then 50% off after that. Wow sign me up! Not! Glorified spam disguised as an incentive is disgusting.
ReplyDeleteAgreed, Darrell. I would rather not get perks; just show me how I can improve, and give me tools to learn how to do it. I like that they are now showing content that may be of interest to you and your audience (which is helpful if you're in the social marketing field as a career). But the perks thing is "meh!" for me.
Delete