I’ve been on a mission to substantiate my loss in Klout score since the company announced changes to their algorithm in late October. At the time of the change I saw my Klout score of nearly 70 drop to the mid 50’s. Very shortly after the changes to the service and their PeopleRank algorithm I embarked on a test to see how the network connections affected my score.
During the process I’ve come across some incredible privacy concerns and have data that conflicts with what Klout has said about disconnected networks not affecting your score.
The week following the changes I decided to unlink Google+ and Facebook from my Klout profile. At the time I also unlinked Instagram, YouTube, Flickr, Tumblr, and Last.fm. I retained my primary connections to Twitter, LinkedIn, and Foursquare so that the service was still receiving data from my active accounts.
Klout Privacy Concerns
To unlink accounts in Klout you go to Settings -> Connected Networks -> Unlink on any of the accounts you chose to disconnect.
After unlinking the networks I had expected to see an immediate drop in my Klout score, which did occur – but not until the week of Thanksgiving. During the testing period I was surprised to see that Klout appeared to still be accessing my unlinked networks. How did I exactly determine this?
During the test period I would log into my Klout account on a a daily basis. Users of Klout will recognize that the service will occasionally prompt you about recent notifications as well as ask you to invite users from your network to being using Klout.
In what appears to be a lapse in privacy, nearly every day when logging into my account Klout would prompt me to invite users that were on Twitter, but I was surprised to see that the service was also still prompting me to invite users that were only part of my Facebook network.
How is it that Klout was still recommending that I invite users to their service although I disconnected Facebook nearly three weeks ago? Of everything I learned during the experiment, this appears to be a privacy issue that Klout has overlooked, or ignored.
There have been some previous privacy concerns with Klout measuring public social data across the web. In a blog post titled “We Value Your Privacy“, Klout CEO Joe Fernandez confirmed that the service is indexing this public information and you could opt-out of the process on their privacy page. Personally I think Klout should be an opt-in service, and does the general public know that they are using their social engagement for the platform? Likely not.
Also in that post Fernandez states that “Klout has no interest in understanding the influence of minors.” But it is obvious they are still tracking Justin Bieber. Afterall, isn’t he a minor?
If the service is still collecting information about you after you’ve unlinked a network in your profile then Klout should be honest about what they are still tracking. I honestly expected more from the service, and am concerned that it was still monitoring Facebook in this case and mining my contacts when I opted-out by unlinking my account.
What Networks Influence Your Klout Score?
So what networks truly influence your personal Klout score?
The first week after unlinking many networks I didn’t see any change in my score. The following week I reconnected with YouTube and Flickr, two sites where I have a significant amount of traffic. A week later I didn’t see any change in my score. At that point I reconnected my Instagram account becauseI was fairly certain it wouldn’t affect my rank if the other two services didn’t. And as expected, it didn’t.
But on November 23rd I saw a significant drop in my score – nearly 10 points. I initially suspected a glitch with the system. I checked with trusted colleagues and they hadn’t seen a similar drop. I waited 48 hours to see if Klout would announce any problems with their service, as they occasionally have in the past, and there was no such announcement. At this point it looked as if my Klout score was permanently lower.
What did occur during that week was Klout’s announcement on November 22 that Google+ was now part of their algorithm.
Wait, didn’t Klout say that Google+ was now part of the service in September? They did, but apparently it didn’t begin to officially affect your score until November. I’d prefer more transparency here to understand exactly what data is being collected and what services are affecting my score. I understand that Klout likely needs data before they can build a new network into it’s PeopleRank algorithm, but just be transparent about the process. Is that too much to ask?
In Klout’s announcement on November 22nd, their blog post stated that “if you are active on Google+ and have connected your Google+ account to Klout, you will see a Score increase. Regardless of activity level, no user with Google+ connected will see a Score drop. Similarly, we do not penalize users who do not connect Google+ accounts. We measure influence on Google+ by analyzing public posts.”
That is a very clear statement. My score should not have been affected my disconnecting Google+, or Facebook for that matter, after the Klout algorithm changes in October. So why did I see such a huge drop? The only logical explanation is that Klout penalized me for not having Facebook and Google+ linked to my profile.
As soon as I relinked both Google+ and Facebook to my Klout profile I saw my score increase back to my previous score range. After about eight days of hovering below 50, my score jumped back up to 56 on December 1st.
Klout Only Measures Five Networks
Wait, I thought Klout actively measures 12 networks as well as multiple Facebook pages? According to Klout’s most recent blog post, Fernandez confirms this isn’t true. He states that “currently, we actively measure five networks: Twitter, Facebook, LinkedIn, Foursquare and Google+.”
Specifically, on each network this is what Klout analyzes to build your score.
- Twitter: Retweets and Mentions
- Facebook: Comments, Wall-Posts, Likes
- Google+: Comments, Reshares, +1
- LinkedIn: Comments, Likes
- Foursquare: Tips – Todo’s and Tips – Done
If I am truly an online influencer, and have incredible reach online with a very engaged audience does it matter that I am on one network vs. ten? Will the service continue to normalize their rankings as new services launch? Their blog posts indicate that the amount of networks don’t necessarily matter, but my experiment proves otherwise.
Lack of True Transparency Means Concerns Still Exist
In closing, I have a lot of questions as to how Klout values privacy and what networks they are actually tracking, even if you have unlinked them. While Klout CEO Joe Fernandez preaches privacy, my experiment does show that there is still some questionable practices happening here.
In changing their algorithm it does appear to be more accurate. But I am still as a loss as to how my influence dropped so drastically after their October announcement.
So will Klout respond to these issues? I hope they do. As each day passes I have less and less confidence in their scores and am left with serious concerns about privacy and their algorithm.
What are your thoughts on Klout’s privacy issues? Have you seen similar issues with your score when unlinking networks? Share your thoughts below or find my on Twitter. (@djenders)