Last month, we reported that LinkedIn has been working to make their endorsements feature more relevant, with a new back-end infrastructure that puts increased emphasis on endorsement accuracy - like if the endorser has actually worked with you, and thus, can actually validate your skillset.
This is part of LinkedIn's wider efforts to refine their data accuracy, which is important, because the accuracy of the platform's data fuels all other elements of the service - from ad targeting to search to analytics. And for those unaware, LinkedIn's data is also why Microsoft is paying $26.2 billion for the social platform - the platform's data set, which includes the career, education and job history progressions of more than 450 million people, is unmatched, and the potential of that information resource is huge.
And today, LinkedIn has revealed how their new endorsements system will look, putting a focus on 'endorsements that matter' and targeting the right people to make relevant recommendations.
Requisite Skills
As you can see in this screenshot, the new endorsements system will highlight those provided by people who've worked with that user, as well as mutual connections to give more context on that individual skill.
The system will also highlight endorsements from people who are "highly skilled" in that specific area - you can see this more prominently in this example from my profile.
How the system identifies "highly skilled" endorsers is not 100% clear, but it looks like it's based on that individuals respective endorsements in the same field. That may make those specific qualifiers less relevant - if people have gained a lot of random endorsements over time, that doesn't necessarily mean they're especially skilled in that area. Really, the endorsements that are likely to be of most value are those from a connection that knows both the person and the skill - one or the other alone seems less valuable, but still, it's an improvement on the previous system either way.
One potential flaw I can see with this process is that the value of those additional qualifiers will be less relevant to those in more public facing roles, like journalism or blogging for example, where people may benefit from their work but not have ever actually worked with them. For example, in my case, people may have read my writing, and they could justifiably endorse me for my skills based on this, but because I've never worked with them, that endorsement wouldn't be as valuable, going on LinkedIn's new process. This is likely why they added in the "highly skilled" option, so it's not just dependent on past colleagues, but also on those who are recognized as being proficient in the field.
It's hard to say whether this will significantly boost reliance on, or engagement with, LinkedIn endorsements. But at the same time, something needs to be done to make them more relevant - I mean, it's become something of a practical joke for people to endorse others for ridiculous subjects.
And the thing is, even if you did endorse somebody for all these things, it likely wouldn't matter, because no one reads the full detail of your endorsements listing. With this new system, LinkedIn's looking to make it a more relevant and accurate data point.
And as noted, that data is critical.
Professional Data
Since the announcement of Microsoft's pending acquisition of LinkedIn, many have questioned what it means for the platform, why Microsoft would pay so much for the network. Most of these queries are based around the public-facing elements of LinkedIn and how Microsoft will use the platform to promote its own products - but the real value of LinkedIn lies in the insights contained within its vast data banks.
Consider this - back in back in 2014, Re/code writer Kurt Wagner (then writing for Mashable) detailed how LinkedIn's data science team constructed a full, customized overview of how his career is projected to pan out, from promotions to job changes to where he would finally end up.
Using the various data points, based on the thousands of other people matching Wagner's educational background, interests and skill set, LinkedIn's research team was able to highlight a full, predictive professional history, mapping his life based on probabilities and numbers. Now, the results provided at that time seemed more interesting than accurate - in the end, the research team pointed Wagner to people who likely matched his career direction in order for him to reach out to them as potential mentors - but the applications for such a tool are significant.
As noted by Wagner:
"It's likely that one day, LinkedIn will offer a feature like this for all of its 300 million members. Imagine a tool that takes your profile data and returns a dozen potential mentors in seconds."
That was two years ago, so no doubt LinkedIn has advanced much further in this process by now, while they've also grown their data pool by 50%, at least in terms of member numbers.
The potential applications for such tools are significant. Imagine if, one day, you could enter your details into LinkedIn - your interests, hobbies, personal traits (based on psychological tests or similar) - and the platform could utilize its vast dataset to point you towards the career that would not only be of most interest to you, but would lead you to your most fulfilling life?
It sounds far-fetched, but it's not - in fact, tests have already shown that computer algorithms can actually make better hiring decisions than people. And that research was conducted without LinkedIn's data - with it, the system would be able to take into account a broader scope of people with similar profiles, measure how long they've stayed in each position (job satisfaction), what courses they've taken to develop their careers, etc.
In future, LinkedIn may be more than just an professional social network, it may actually become the platform that all students need to refer to in order make more accurate choices for their future studies.
And that could be hugely valuable - LinkedIn actually offered a tool along these lines once in LinkedIn University Finder, which aimed to help students connect with their ideal career path by showing them the educational providers that best matched their future intentions, based on what they wanted to do, where they wanted to work and where they wanted to live.
But the information provided by that tool was, of course, hugely valuable, so LinkedIn removed it earlier this year, likely to enhance their data value.
Though even then, in their explanation for it's removal, there's a telling statement about the future potential of such an application:
"Students are our fastest growing demographic on LinkedIn and remain a very important audience to LinkedIn."
Students are the fastest growing demographic on the platform - in fact, recent data from LinkedIn showed that around 38% of LinkedIn's entire user base are Millennials. Eighty-seven million of them.
When you see it mapped out like this, the value proposition of LinkedIn becomes increasingly clear - it's not necessarily about how the platform will change, how Microsoft will integrate LinkedIn with Word, how the immediate applications of the two entities will merge into some new, Microsoft-infused social platform. In its own right, LinkedIn's future value proposition is huge, and it's through their ever-growing dataset that Microsoft will benefit from, and capitalize on, the platform's potential, and maintain relevance with that next generation.
Endorsement Targeting
In addition to the added endorsement context, LinkedIn's also working to improve the system by calling on people who know both the relevant user and skill to provide endorsements, as opposed to just suggesting you endorse people for skills, whether you know much about them or not.
Again, it's about data accuracy, and that refined focus will have wider-ranging impacts than just endorsements alone - even if it doesn't work out and no one uses endorsements any more, LinkedIn's learning how they can better utilize their resources to improve their accuracy and focus.
LinkedIn's also reduced the amount of skills shown to people who visit your profile, with only the top three given focus. This means users will need to ensure those endorsement listings are in order to highlight the skills you most want to showcase.
They're small steps, in isolation, but all part of the platform's bigger plans.
The new LinkedIn endorsements system is being rolled out to all users of the LinkedIn mobile app from today and will be part of their re-vamped desktop experience when that's released soon.