Enterprise Data Portability needs a Reputation Standard
Today we announced that we’ve joined Data Portability. In a couple of months, we’ll ship Clearspace Community 2.1 and as far as I can tell it’ll be the most “Data Portability compliant” enterprise software on the market. But so what?
Data Portability inside of a company?
Most of the stuff in Data Portability seems counterintuitive internally. Gia covers it well on her JiveTalks post and on her blog. Basically, there are all sorts of privacy and legal issues inside a company but it’s a no-brainer to have Data Portability in our externally facing product. In the meantime, we’re interested in working with the Data Portability group to help contribute to these standards as well as new ones as well. Hopefully, the organization is now at a point in its evolution to proceed with formal and elected leadership, a standards body, voting process and the rest of the stuff that makes organizations successful.
These pieces need to add up to a Reputation Standard
The one standard yet to exist that I think would add a ton of value is a “Reputation Standard.” This would be something that would work on the social web and within your company. The simplistic way to think about this is something like eBay’s customer ratings and a sort of “technorati rating.”
How would it work?
Potentially, you’d have a top-line weighting that would show others your relative standing. So, if you’re contributing to sites that embrace “reputation standards” your contributions and interactions are earned and connected to you. The fact that you contribute, that other people respond favorably to you, that you respond to them, the size and quality of your social graph– all these things improve to your standings. Now imagine not only having these data points on the social web but if your company is using intra-enterprise collaboration software that also uses a reputation standard then you could have a holistic picture of someone’s reputation. Would people try to game it? Sure, but people game everything.
Why is Robert Scoble’s reputation good?
Is it his bajillions of followers on Twitter and Friendfeed? Is it the amount of comments on his blog? Is it the amount he contributes to Flickr? What about his connection to other people who also have the same level of interaction? Perhaps it’s the strength of his ties to others, the positive tone of his interactions, how frequently he directly responds or his overall availability.
What if you could search for reputation?
Once you have values for reputation you should be able to search for it. Perhaps I could search Google and find people with expertise and relatively high reputations. Once social networks and collaboration software are standardized on the reputation standard I could do some contextual and meaningful people-location. This could help me find answers, invite the right people to collaborate and reinforce the right behavior.
Goodbye HR reference checking
Well not totally, but could you imagine the power of having a more objective perspective of people’s reputations. If someone looked great on paper, seemed like the best fit once interviewed and had an excellent reputation score, I’d hire that person in a second.
Things people have said about this post
The basic premise here does not hold up. Making the assumption that all contributions is equal does not work well (and that is being optimistic). Actions and reputation is far more granular and subject oriented. Expertise, understanding, and valued contributions across subject matter and is an incredibly rare occurrence. Quite often people’s contributions are valuable to others on only a select set of subject, but not all people find them valuable. Mistaking many followers/listeners for value is also deeply problematic as there are many reasons people with follow/listen to others, but assuming it has value and build some sort of reputation is going to cause serious problems.
Value for reputation needs to account for the derived value of the individual’s contributions. Narrowing to subject is very helpful. But, quite often people who contribute infrequently can have higher derived value from those reading and following than those who contribute a lot. Volume does not remotely indicate worth. Actions taken around contributions are far more valuable. Who is doing something with the contributions they receive has more value than volume. These connections are more valuable than metrics around mass.
Tracking volume of followers and/or volume of contributions is a quick way to have a measurement that means very little.
I think is asking for the types of measurements that do make sense, and isn’t suggesting that volume of content or followers should be considered a major metric.
I like: “Actions taken around contributions are far more valuable. Who is doing something with the contributions they receive has more value than volume.”
So, how do you measure that? What would work?
[…] communities are for bringing customers closer. In a theoretical sense, brands could determine which customers have the best reputation, and how to keep and reward them. But perhaps, most importantly, customer experience will improve […]
For some social media active folks, GOOGLE is their resume.
I think this is an excellent move - I’m a big fan of data portability, and paranoid about locking any content into proprietary systems.
Having a consistent online image and profile saves everyone time - including HR departments vetting candidates - and hopefully will allow incresaed linkage between other social apps (Facebook etc) and a company’s outward facing community.
[…] in his most recent post, “Enterprise Data Portability needs a Reputation Standard”, he talks about reinforcing the […]
[…] Jive CMO Sam Lawrence has grand plans for where this enterprise data portability might have the chance to go: In the meantime, we’re interested in working with the Data Portability group to help contribute […]
There’s a lot of subtleties unde the skin, based on human behaviour, which lead to some interesting questions..
First, lets decide which Paula you are talking about. Inside ‘the company’, there’s Paula who is head of UAT for a given product or service, but there’s also Paula the informal - the one on the committe for the kids Christmas Party, the one who has the key to the pool table when it swallows your change and the one who knows just the right way to kick the soft drink machine. Outside work, there’s Paula the motorcyclist, Paula the WOW clan leader, Paula the daughter and sister.
Would this Paula, for example, want ALL (or even any) of those previous identities associated with her profile on a dating site ? It would be unwise to associate the dating identity with her work identity (from a physical safety perspective). She might also be nervous about SOME of her Work, WOW or biker buddies knowing as well (but there may be some she’s particularly interested in having know as well…)
Depending on whether you share the details of one identity with ALL or SOME of the contacts of another identity, you either increase the value of the second identity, OR you create a third identity.
Which of your identities has which reputation and how much ?
What, if any, crossover between these reputations do you want made public ?
What, if any, crossover between these reputations are relevant ?
What, if any, crossover between these Identities do you want made public ?
What, if any, crossover between these Identities SHOULD you make public ?
How do you trust / ensure that the owner of each of these identities keep them up to date ?
Who do you trust to manage these identities / reputations exactly as you wish (matching only as you wish, segregating as you wish) ?
I agree with the various comments that volume based metrics for measuring “reputation” are a poor idea. But that’s not the only idea Sam has floated here — voluntary, digg like ratings are another. Personally, I dislike “karma” systems like Slashdot’s, as that always seems to devolve to a popularity contest. However, it’s clearly a step in the right direction. Imagine a rating system where different raters are weighted differently — and where subjective feedback (think LinkedIns recommendation system) maps to a value in the system as well… Inadequate? Well, yes. It’s a model, and all models abstract away (sometimes important) details. Models are still damn useful, though…
@Martin: I understand what you’re saying, in my 20th century gut. But my head wonders if a citizen of the 21st might not find these worries silly. Managing all that is a lot of work, and we’ve consistently seen that our species is unwilling to invest a lot of effort in preserving their own privacy…
And Jeremiah makes an important point, which I can back up with personal experience. Being the first hit of a google for your full name has value — it’s worth real-world money. I can attest to this. My current paycheck is higher for it — I sat down in an interview, and was able to casually say that I came up first in a search for my name. That fact was worth a higher salary.
As Jeremiah also suggests, that won’t be true in every profession (yet, but give it time). But the deeper point lurking there, I think, is this: things that have monetary value tend to get exploited. This will.
To build on @jowyang’s point, I wouldn’t limit it to social media active folks. Many people have relevant info about themselves online that a simple google search would reveal. For individuals, reputation management and personal branding will become more of an issue and concern as we move forward on these issues. For companies and recruiters seeking talent, googling a potential candidate is already a de-facto standard practice, though it may not be an official policy of the company.
[…] to take this idea another step even further, read Sam’s take on data […]
I think I’m in agreement with Martin on this, at least in part. Context is important — I am many different people and your interest in me depends on how our interests converge. Semantic reconciliation can play a role here.
Beyond that I think that relying only on my participation to date is limiting. What if I have great credentials but not much of a social reputation because I’m new to the community? Taking into account only the reptutation I’ve built in this one technical context is also limiting. Certainly you will want to take these things into account but you not taking into account who I am means you’re going to be missing potentially valuable connections.
To me the ideal would be a service that’s more like a FICO score that lets you define the attributes you want to score on and the weights you want to assign to them. It would \take into account factors such as professional designations and education, or volunteer activities, or hobbies (depending on the context) and combine these with my online reputation. And it would allow me to have multiple scores depending on who I am in a given context, rather than conflating these which I think creates confusion and poor results.
[…] that even folks from my hometown can be misled by this DP media hoopla. When I read Sam Lawrence say: In the meantime, we’re interested in working with the Data Portability group to help contribute […]
[…] new operations tools could emerge, they should be part of the head. Perhaps they are things like reputation scoring, the quality of co-created deliverables, or “repair” rate. Who knows. These could have […]
Reputation is a subjective thing, especially when you remove the context of the social site it’s housed on. Reputation by whose standards and in what context?
Drawing on Martin’s hypothetical scenerio in a previous comment, Paula’s reputation as a writer may be glowing, but she may just suck as a manager. She may be a great manager of teenage boys (she volunteers at the local community center) but terrible managing a gaggle of ladies in corporate communications (they hate her military approach). She may be a great historical novelist, but terrible at research papers.
Kim’s idea for treating it like a FICO score is something worth exploring. Context would be critical. So how would it work? Would Paula say “Rate my Management abilities”? No, in Paula’s case that rating wouldn’t really mean anything. Would she be able to say “Rate me on my ability to manage teenage boys”? She would be more inclined to offer up ratings for things she knows she’s good at and avoid those things she’s bad at. Or maybe she THINKS she’s good at managing teenage boys, but when the ratings roll in, she finds out she’s wrong. Would she be able to remove the rating?
I’m not convinced yet that this would be valuable. But I’ll keep an open mind to see how the idea develops.
This is the first time I’ve visited this blog and it’s great – very thought provoking.