Evaluating LinkedIn Recommendations

July 17, 2009


An interchange on Twitter with @jowyang and @carissao got me thinking about LinkedIn recommendations. The question was this: do LinkedIn recommendations mean anything anymore? Seems to me that some do while others don’t. It depends on three people: the recommender, the recommendee, and the reviewer of the recommendation.

  • Recommender: fairly or not, a recommendation from a higher-ranking person may mean more as their time is often scarcer. More importantly, does the recommendation show unique detail about the person in question? Is there a bit of thought behind it, or does it look like a stock template? If in doubt look at other recommendations written by the same person.
  • Recommendee: A few high-quality recommendations beat a plethora of boilerplate ones. Some folks blast out a cattle call for recommendations to all their connections. Do that repeatedly and it’ll be evident that you value quantity over quality.
  • Reviewer: I’ve had two hiring managers tell me in the past year that “LinkedIn recommendations are worthless.” In Geoffrey Moore terms, one was a laggard (no LI profile) and one was an early adopter (heavy user of social media). Both thought of it as a system that could be easily gamed. Most people in Moore’s model are early or late majority. In my experience both as a hiring manager and as a job seeker, those early & late majority folks may still value LinkedIn recommendations. They read them judiciously just as they’d consider the source in evaluating any other kind of reference.

What are your experiences? Do you take LinkedIn recommendations seriously, and how do you evaluate them? Would they be improved if, as @jowyang and I discussed, there were an element of scarcity – if you had to choose wisely, knowing that you could only recommend a limited number of people? Where else do you look for recommendations and insight on candidates and potential business partners?

Update: @jowyang blogged on the same topic after our interaction – you can read his thoughts here. In the spirit of this discussion let me say that I heartily recommend it!

Update #2: It’s a rare occasion when this blog intersects with my other, much snarkier, blog. So it’s a special day when this happens.

Photo by otherthings


25 Responses to “Evaluating LinkedIn Recommendations”

  1. Russ,

    thanks for raising the issue. I pay attention to recommendations, though I approach them with skepticism. I scan them and if I see a genuine, detailed one, I note it. If I see what looks like a lazy, boilerplate recommendation that doesn’t tell me anything other than platitudes, I note it too.

    I also look at the approximate ratio of connections and recommendations. If you have hundreds of connections and few recommendations, it testifies to your appreciation for quantity and weak relationships. If you have fewer connections and relatively many recommendations, I will pay more attention to you.

    I only give recommendations for people I can write in detail about. I’ve received requests from people I met once or twice. ‘X is a nice guy and I like the way he thinks’ is not a recommendation, so I’m not going to write it.

    Beyond LinkedIn, a Google search tells me more about a person. Their online/offline activity, what they say, what others say about them on other sites, customer reviews, etc. I once didn’t pursue a deal because of what people said about the person in their book reviews on Amazon (yes, the person, not the book).

    I’d say, recommendations matter to a degree and if you take them in the context of other information about the person.


  2. Arun Vemuri Says:

    Interesting take. Reminds me of the dilemma I was in when seeking my first reco on linkedin. Have captured the same on my blog http://ivak99.wordpress.com/2009/03/20/rubicon-of-requesting-recommendations/

  3. elzmcc Says:

    Social network recommendations — whether it’s LinkedIn or Naymz — are a reality and have importance. They require and deserve planning, similar to developing a resume or cover letter.

    _Seeking recommendations? Map the context and content of the conversation you’re hoping to develop.

    _Offering recommendations? Provide unique sets of information — the situation, solutions, achievements.

    Great eye-opener. Look forward to addt’l replies.

  4. Bill Fitzpatrick Says:

    Well, since you asked my opinion I’ll say I treat them as any information garnered on the internet. Question it’s validity as evidenced by the source, the context and the importance of any actions taken based on said information. Or in short, question everything and don’t trust anything with some form of verification or authentication. :^)

  5. rsomers Says:

    Peter, I love the idea of the ‘meaningful recommendations / total connections’ as a metric. There’s much common sense in your approach of evaluating in context. The more folks who approach it the way you do – write only recommendations when you can do so on a meaningful, informed basis – the higher the value of the system.

    Arun, your post raises a great question: if you don’t have any (perhaps because you’re in early career stages or new to LI), how do you get started? It is a Rubicon to cross. I’ll freely admit that my first reco was a tit-for-tat exchange with a peer. But the ones that have meant to most (to hiring managers, not just to me) have been the unsolicited ones.

  6. strom Says:

    I disagree and think that LinkedIn recommendations have value. Yes, there is the log-rolled mutual back-scratching component, but honesty trumps that. If I am honest, I won’t
    recommend someone that I don’t feel has done a superlative job for me or with me in a past engagement, simple as that. There is some negotiation that happens out of band
    before the rec is posted, and you can develop a sophisticated series of controls or point systems or whatnot, but the fact remains that you want some form of mutual trust and communication for all this to work.
    Yes, people have gotten jobs via LinkedIn. There are forums that I subscribe to where local (to St Louis) jobs are posted first before Craigslist or anywhere else — the smart folks scan these forums frequently and many people have gotten work as a result.

    Yes, I have gotten gigs from my LinkedIn activities, and I have also gotten sources for my freelanced stories from LinkedIn contacts. It is a very useful tool for me. And one of the other things that I like about LinkedIn is that it offers in a single place what people are saying about my work history, which helps when I am pitching a new client and they can see that I am someone who has been around this industry and the Internet for many years.

    I will continue to use the recommendation engine at LinkedIn. While imperfect, it is the best way that potential clients can get a picture of the kind of work and relationships that I have over the years.

  7. […] Russ Somers has extended the converation on his blog: Evaluating LinkedIn Reccomendations Share This […]

  8. Coromel Says:

    I think you have to look at them with a grain of salt. Obviously the person asking for the reference is not going to allow a bad one to show on his or her Linkedin. It’s just a tool I use. I don’t weight it very much unless I know the person personally who gave the reference.

  9. John Feeney Says:

    Recommendations very between profiles. As mentioned I do approach this “referals” with caution. The one problem most are “cookie cut”, “Great person” plug. Not to often do we come across some real “FACTs” about the recommendation.

    What I don’t see is recommendations to connect? Like this individual is an expert within his field, not just a Social Marketer creating a profile to pretend…

  10. […] Update: Russ Somers has extended the converation on his blog: Evaluating LinkedIn Reccomendations […]

  11. rsomers Says:

    Wow, many great comments – I will write you all LinkedIn recommendations saying that you are thoughtful people 😉

    More seriously, ‘valuable, but take with a grain of salt’ seems to be the prevailing vibe, and that makes a good deal of sense to me.

  12. Peter W Says:

    I only partially agree with this article. First – As an open networker on LinkedIn I get a lot of requests from people I don’t know saying “recommend me and I’ll recommend you.” I don’t play that game. Ask me twice and I remove you from my contacts.
    Second – if someone asks me to forward a connection request I’ll gladly do so – with a note that states whether or not I actually know that person and, if so, what the connection is. (Friend, Co-worker, Client, etc)
    Third – I agree with the comments about boilerplate recommendations. It’s fairly easy to see through that.
    Fourth – Yes a lot of people do what I call the “twitter-shuffle” (you follow me and I’ll follow you…) So they will write a recommendation after you write one for them. There’s nothing wrong with that as long as both recommendations are honest. I’ve had that happen and have not posted some that were sent to me.
    Fifth – Obviously nothing beats an in-person (phone call) recommendation in this sue-crazy society we live in. Libel laws can make putting something in writing that is negative very risky. Most people are not willing to stand by what they said if confronted and it is negative. That is why a lot of businesses now utilize a probation period to see how new employees work out.
    OK – I could go on but I’m not feeling well and need to get off the soapbox 🙂

  13. rsomers Says:

    Thanks for the thoughts Peter (and I hope you’re feeling better!) It’s good to hear from the perspective of an open networker – I like the ‘two strikes and you’re out’ rule.

    I particularly like your observations on the ‘twitter-shuffle’ or tit-for-tat recommendations. People I would honestly recommend may often be the same folks who would honestly recommend me. That doesn’t make either recommendation invalid, if it passes other smell tests.

  14. ashishtewari Says:

    I found there’s another dimension to LinkedIn reco’s – is there some thread of commonality running through all reco’s received by someone? Or are all generic, “he’s a nice guy to know,” type? If everyone is talking about good leadership, or analytic abilities, etc – chances are that’s what you’ll get. If everyone is talking generalities – pinch of salt needed.

    • rsomers Says:

      Good point, Ashish – seeing the same skills/value repeated from multiple perspectives adds confidence that the skill is really there. “Nice guy” recos, on the other hand, might be the equivalent of “great personality” in the dating pool.

      “Great personality” is all I ever had to go on there, but we work with the tools we have 😉

  15. Jim Says:

    I don’t make my Linkedin recommendations visible, just as I don’t have my most important relationships listed as connections. Real relationships are too scarce, precious and fragile to expose to the world. Want my references? I have many great ones. When we get to the point that you need to contact my references, out of respect I’ll ask their permission and get back to you. Want to see if I have connections? You should already know that, because if you really have a relationship with me, you know a lot about me.

    Linkedin is all about manufactured relationships and creating a “personal brand” that may not even be anything about who you really are. Corporate marketing spin at a personal level.

  16. rsomers Says:

    Interesting perspective, Jim. Most of my LinkedIn connections are people I worked with and would be comfortable reaching out to for advice or an introduction. Many of them are “real” relationships to the point of being close friends (we went to one of their kids’ birthday parties yesterday). Am I creating some kind of risk to that relationship by allowing that person to post an unsolicited recommendation they wrote about me, based upon several years of work together in two different companies? Relationships are indeed precious and fragile, but I don’t see any risk in documenting them online.

    But the fact that you and I view and use the tool differently drives home how creators of such tools need to allow for different users with different preferences. If I don’t want to display connections or recommendations, I need the freedom to make that choice. And I agree that, whether LinkedIn recommendations are puffery or potentially good information to be consumed with a grain of salt, they don’t replace the permission-based reference check you mention.

  17. […] – Russ Somers (Egghead Marketing) posts an extended the conversation on his blog about the  Evaluating LinkedIn Reccomendations that references Jeremiah’s article and the resulting Twitter […]

  18. […] Update: Russ Somers has extended the converation on his blog: Evaluating LinkedIn Reccomendations […]

  19. […] Somers made an interesting point in a blog post entitled “Evaluating LinkedIn Recommendations” that he did the same day as Jeremiah’s: wouldn’t the recommendations mean more […]

  20. C.H. Low Says:

    I view recommendations on Linkedin as just another piece of information.

    If a recommender is somebody I know or is a second degree of connection, I reach out to get more honest feedback.

    What is said in the recommendation becomes another topic to have a conversation with interviewee.

    I agree that Linkedin recommendations by itself has little value of its own standing.

  21. Peter Says:


    I googled “LinkedIn References” after seeing a former project manager get a rush of glowing recommendations shortly after being asked to leave his current role. I was amazed to see the comments, which were clearly an example of back scratching as the person in question is lacking in most of the skills required for the job, hence the reason for him being removed and the position advertised. A number of people had cited him to HR as a partial reason for them leaving the company but if you would take his recommendations at face value, you would assume that you were dealing with a hard working, intelligent star – which would be far from the case. I really think examples like this show that the LinkedIn recommendations can be seriously flawed.

  22. rsomers Says:

    Thanks Peter – certainly a good real-life example.

    This problem existed prior the creation of LinkedIn, or the Web for that matter. As I’ve said before, even the Unabomber can get three people to say good things about him. The typical reference check process (allowing the applicant to cherry-pick three references) isn’t any more reliable than reading LinkedIn recommendations and taking them as gospel.

    So I certainly wouldn’t take LinkedIn recommendations – or recommendations from the old-school reference check process, for that matter – at face value, for exactly the reason you cite.

  23. […] read my post How To Write An Excellent LinkedIn Recommendation at the link.  And, here’s a followup to Jeremiah’s post by Russ Somers, and here’s a response to Jeremiah’s post by LinkedIn on their own […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: