3 ms·
Great post. LinkedIn needs to stop focusing on raw numbers and work on quality engagement. As a recruiter, I have had a policy of accepting any request over th
by smedawar 13y ago
Great post.
LinkedIn needs to stop focusing on raw numbers and work on quality engagement. As a recruiter, I have had a policy of accepting any request over the years, and it has resulted in 10100 connections and almost no engagement. I almost only go to LI to accept my requests. The most value LI has provided has been as a directory and having accepted 10k+ requests, I now have a very open and thorough directory. If I was trying to use the site for anything else, it would be a mess.
I think there's a lot of opportunity for them to fix this (and they probably will), but it seems like they're focusing on growth first.
- kirillzubovsky 13y agoPost author here. Thanks for the comment. I just got a response to on Twitter from @grayj_ , saying "Making a KPI go up will often get you a bonus or a promotion; advocating good UX won't." However sad, I suspect this is the reality of LinkedIn being a large and slow corporation. Unfortunately :(
- benmanns 13y agoWhy not make % accepted requests the KPI instead of # sent requests? The former would encourage a message for every request. The latter encourages mass emails to address books, accidental requests, etc.
- notahacker 13y agoThe former is strategy is user-friendly. The latter strategy is metric-which-impresses-investors friendly. There's a reason social networks spam more than Viagra merchants.
- 001sky 13y agoThere's a reason social networks spam more than Viagra merchants. LinkedIn is pretty bad on this. Just when you think youve opted out of every mailing list, its seems like they invent or create some new 'feature' which opts you in. Again.
- kirillzubovsky 13y agoInteresting, even from the comments it seems people either blankly accept all the requests or don't accept them at all. Somewhat related - I remember reading that when Facebook engineer decided to add a message feature to a friend request, Zuck got really pissed at him. I suspect that's because "# of connections" is a really hot metric, and it grows exponentially for people who blankly accept every request. Thus, from LIN point of view, it makes sense to sacrifice UX for some people to get a sure growth for everyone else. Now, if that "growth" is just a vanity metric, now that's a whole other story.
- pscsbs 13y agoHi, engineer at LinkedIn here. As much as it might seem corporate propaganda, LinkedIn really emphasizes a "members first" attitude in engineering. Engineers frequently get commended for decreasing spam, improving UX, and securing member data. I guarantee this post will be reviewed tomorrow, and we'll discuss how to improve the flow for connecting with new LinkedIn members from the "People You May Know" feature. Please keep the feedback coming. :)
- computer 13y agoWhy do I, after several years of people complaining, still have to log in to unsubscribe from your however many different types of emails? And even logging in after clicking "Unsubscribe" doesn't unsubscribe you: you then have to go to options and find whatever caused the email yourself. For a company that puts members first, that seems very member unfriendly. But it does mean more emails to users, which no doubt increases some metrics.
- VLM 13y ago"almost no engagement." I was just recently thinking that I hadn't logged into linkedin since 2012 or 2011, and I've done some things to add to my resume. I assume a less than 1% chance what I do, would interest you. If you run the math, 10K / (2 x 365) * 0.01 * 7 means your almost no engagement should be quantified to about one interesting thing per week, plus or minus a couple orders of magnitude. So its statistically believable. "roughly one interesting thing per week" does translate to multiple billions of value in TV and music and movie biz, so its valuation is not entirely outta line.