TRAACKR

By Laurent Liscia and Pierre-Loic Assayag

SUMMARY
What is Traackr’s goal?
Our goal is to collect data from a variety of sites in order to calculate a user’s buzz, popularity and reach (which are proprietary definitions). Today this data is not readily available and requires ad hoc extraction processes that are not convenient to the user or require considerable logic resources. Making it available, however will raise a host of issues, and possibly user revolt, unless the users are put in charge of their own data.

Types of user data

What kind of data are we looking at?
- Data posted by the user (such as audio, video, and text content)
- Data posted by other users about the user
- Additional data resulting from a remote site computation: for instance, an Amazon reviewer ranking, or a LinkedIn number of connections

Our argument is that this data should be readily available to third parties, not just Traackr. The data should be interoperable (ie understandable, parsable, usable) from one social site to the next and to other sites that rely on it for computational purposes.

This poses three very clear societal issues: privacy, validity, and ownership. Do I hear you groaning? No, really, you need to read this.

Privacy

First, let us point you to a resource page overflowing with brilliant and generally pessimistic thoughts about the whittling away of our privacy by the Internet:

http://digitalenterprise.org/privacy/privacy.html

There are two schools of thought on this: the industry should regulate itself, based on user feedback or revolt, as the case may be; conversely, a governing body should regulate the industry. The first camp is winning out: in these Homeland Security- flavored times, there seems to be no desire on anyone’s part to genuinely protect individual privacy.

Sure, there are industry guidelines defined by the Federal Trade Commission, that stress the need to notify the user, give them a choice (opt out), ensure their participation in the privacy process, secure their data, and instate an enforcement mechanism in case a violation occurs. Other feeble attempts include the W3C’s P3P platform, a standard for privacy policies and the OPA’s privacy guidelines.

In our view, the lack of a more stringent framework is a class-action suit waiting to happen, possibly one brought by the deep and wonderful minds at the Electronic Frontier Foundation.

Validity

While privacy is a societal issue, validity is a social one. (It’s nice to split hairs from time to time). What we mean: how do I know you are who you say you are online? With word of mouth accounting for a sizable amount of purchases, and online WOM accounting for an increasing portion of total WOM, how can I trust another user’s product review to be genuine – and not, say, a post from the manufacturer? (See our blog post on user reviews).

There are interesting but isolated initiatives to answer the question. One is Amazon’s “Real Name”, by which users agree to authenticate themselves, and take responsibility for everything they post on Amazon. The upside is that they also create a trusted micro-brand in doing so. Another initiative under way is OASIS’ recently announced online reputation standards.

This feels right: Internet users should have responsibilities corresponding to their rights. In Traackr’s influence paradigm, your incentive for posting responsibly is that you increase the value of your micro-brand.

Of course, there will be cases where the value of your brand grows with each new inflammatory statement; but let’s not go down that path today.

Ownership and control

Data ownership is the economic crux of the data conundrum, and in our opinion, the way it will be resolved. If our prophecy proves to be wrong, you can always nail us to the wall; or ask us to buy you a beer.

If you ask Facebook who owns the data that appears on your Facebook profile, they will say they do. If you demand that they remove it, they will simply tell you to delete your account; but they will hold on to the data. If we ask you who owns this data, you will say you. Who’s right? Both parties. In Traackr’s view, you have the right, and in fact, the responsibility to manage your public data as a brand. If this sounds crazy, just ask yourself why you dress up to go to work every morning: because you have an image that you must uphold with your co-workers. You watch what you say, you measure how you respond to people, you evaluate what you did and sometimes circle back: every day, you are managing your micro-brand.

You need to do this online because increasingly employers and marketers are trying to find out who and what you are. What would you not want to control this data? Control, in our opinion, is more important than ownership. It’s Facebook’s business to own some of your data. They need it to make a buck. But it’s your job to control it. That’s why we created a tool that lets you aggregate all your data feeds and assess how your online brand is doing. The good news is that there’s tremendous value for you in this; you can constantly monitor, manage and improve your image via Traackr.

TECHNICAL ISSUES
How are people getting to your data today?

What we and other people are doing right now:
- Requiring passwords to access user accounts
- Screen-scraping/parsing of result pages (such as LinkedIn or Facebook profile pages for instance, based on the user’s first and last name or nickname)
- Aggregating search results and using logic to weed out irrelevant content (ZoomInfo, Buzz)

What could be done
There are roughly four approaches to collecting user data; these are not new, and have been around since the beginnings of the Internet in various guises.

A. A browser-based approach
This would implement a sort of super-cookie; where something like AutoFill would allow the user to control exactly what amount of information is dispensed on any given site; and a recap of the default profile information the user is willing to surrender. This could be changed on the fly. This approach is probably the most straight forward but the most limited as well as it only pertains to the data that the user him or herself controls and wouldn’t include, for example, what others say about me.

B. APIs for data extraction
This is what most sites are doing. LinkedIn will be next, under pressure from Facebook.

C. A new XML standard for social networking
This feels like a logical extension of the previous point. Major players would get together and agree that content would be formatted according to this schema. There would be some room for variability in the schema, but anything specific to the site would be implemented in the schema, and divulged to the standards partnership.

D. A de facto standard that will emerge due to market leadership.
For instance, a data format Google might impose.

The Traackr project is a recognition that there needs to be such a standard, in whichever way and shape it emerges. Our contention is that this standard will empower users to share only what they want to share, when they want to do it, and in a way that is much more effective than it is today. Of course, there will be pluses on the marketing side as well; but one might argue that if an ad is truly well targeted (ie nano-targeted), then it is no longer an ad, but valuable information to the purchaser. Additionally, the ability for a user to parlay their online data into an influence score, or a sphere of influence will create huge opportunities for marketers and consumers both.

Authors: Laurent Liscia & Pierre-Loic Assayag

Summary: While it may not be possible to buy user reviews lock stock and barrel, the industry hasn’t done enough to turn them into the powerful advisory process they ought to become.

Is it possible to trust user-generated reviews?

Generally, the review process is a diptych:

a) Part 1 is the Web site’s official review. On most sites, it’s a marketing blurb provided by the manufacturer. On original content sites such as CNET, it’s an editor’s review. (You ought to know because I asked a CNET editor and was surprised, that this process is impartial and not paid for).

b) Part 2, below the official review, or on another page, is a bunch of concatenated user reviews, usually summarized as approval or disapproval percentages or ratings

While user reviews are a great concept, there is no implementation to date that delivers consistent quality. Some of this has to do with the obvious limitations of the format: people who mistype, don’t have experience with reporting likes and dislikes, or more gravely, people on a mission either for or against the product. Sometimes there’s no review at all, or just one suspiciously glowing ode to the product.

In other words, it’s very difficult to create trust around reviews.

Reviewers are usually aware of these shortcomings. When the product is popular enough, they read other users’ reviews, and may respond to rants and raves in ways that return the content to a more impartial center. Of course this forces the potential buyer to read most (or much) of what has been said; but it seems to me that consumers don’t have a problem meticulously researching their purchases. The real issue is that even extensive content may not be helpful.

Should review sites do away with user reviews then? Absolutely not. User reviews provide a service that the editor’s review cannot: a general feeling about the product AND the company, via support and customer service, and best of all, over time. Typically consumers are not interested in returning their product or fiddling with it: they want it to work out of the box, and user review bring them the long-term perspective they are looking for.

Working to make the review process better

Most sites have included behavior modification routines such as: “Was this review helpful”, and the ability to post comments to someone’s review, much like a blog; or even ploys to eliminate conflicts of interest such as profile tabs. This type of information is strictly volunteered, so the results have been mixed at best.

While it’s probably not possible to “professionalize” user-generated content, especially in a society where grammar and writing are considered luxuries, it is possible to reward exceptional reviewers and value their contribution to the consumer welfare.

Amazon understands this very well; it has created tools to single out the top-performing reviewers, and give them added credibility based on the content they produce across Amazon categories. Great idea … which deserves to be expanded.

A special group of reviewers

There is in fact a special group of reviewers who, out of the kindness of their own heart, out of passion for the product or line of products, or simply because they are aspiring writers or done baking apple pie for their Senior Center, write oodles of mind-boggling content. We believe these reviewers are influencers, not perhaps in Gladwell’s “connector” meaning, but in a nano-marketing sense. They write enjoyable content that I may agree with, and may publish on a variety of sites. Ideally, I’d like to find out where, because once reviewer “Z” has me hooked (usually because they seem similar to me), I wanted to see what else they might have purchased or tested.

They have effectively become my nano-expert, and we need to promote the emergence of this class of users – which is what Traackr is about. Traackr allows these users to aggregate their work in one place and increase their Internet reach.

The other benefit is that these special users bring a novel and compelling value proposition to viral marketers. That’s what we mean when we say that Traackr’s goal is to become the marketplace for influence.

Authors: Laurent Liscia & Pierre-Loic Assayag

Summary: The Traackr team capitalizes on Gladwell’s Tipping Point theory by creating a marketplace for influence; but discovers along the way that there can be nanospheres of influence; and that mass-marketing may well be dead. Long live nanomarketing.

There are three major ways in which marketers have leveraged the Internet: one is a transposition in email of direct marketing, and has had spam as a side-effect; another is a transposition on the Web of media buying via mechanisms like Adsense or Doubleclick, but with a level of automation, precision and tracking that are not possible offline; a third is what we would generally call micro-marketing: the ability to respond to an individual user’s expressed query with a targeted offer. While today, this is still limited to keyword placement, we can count on Google to offer much sharper vehicles as its slew of free apps helps build more and more accurate profiles of individual users.

With Malcolm Gladwell, however, micro-marketing has taken on a whole new meaning. Marketers as a group have realized that the Web is the perfect medium for “tipping point” phenomena. It now seems possible to target a select group of influential people online, and to let them do your selling job for you. This select group, because of its importance, would become the object of an even more refined kind of marketing: nano-marketing.

It’s already an industry, and in case you think we’re fantasizing, here’s a snippet from Fast Company’s Clive Thompson: “According to MarketingVOX, an online marketing news journal, more than $1 billion is spent a year on word-of-mouth campaigns targeting Influentials, an amount growing at 36% a year, faster than any other part of marketing and advertising.”

The Tipping Point

First, let’s review the principles expounded in the Tipping Point.

To summarize them [for details please refer to the Tipping Point article on Wikipedia]:

The tipping point is the level at which the momentum for change becomes unstoppable.

1. There are laws of social epidemics, the first being the “Law of the Few”, stated as: “The success of any kind of social epidemic is heavily dependent on the involvement of people with a particular and rare set of social skills.” These are the “connectors” who bring people and ideas together; the “mavens” or information specialists who serve as experts on certain subject matters for others; and the “salesmen”, who, much like your star insurance or car salesperson have charismatic powers of persuasion (You have to wonder how often Gladwell has been on a car lot lately).
2. The second Law is the Stickiness Factor: what makes your content stand out and stick. To borrow a concept from the evolutionary theorist Richard Dawkins, you want your ideas to become “memes”, concepts that spread like viruses. Otherwise there’s no social epidemic.
3. The third Law is the The Power of Context: epidemics don’t happen in a vacuum. The way society is set up predetermines how the epidemic will spread. The existence of the Internet, for instance, is an accelerator of change and new ideas.

To marketers and particularly online marketers, the Tipping Point has become something of a Bible. Traditional direct marketing experts are no longer dismissive of these concepts. Symbolically, Karl Rove’s departure from the White House may signal the end of the Direct Marketing paradigm, and the advent of the Tipping Point paradigm.

Strange side effects

Here are some of the ways in which Gladwell’s points have come spectacularly true, but in peculiarly useless ways.

The strangest epidemic on the Web is the proliferation of tools that are supposed to MAKE tipping points happen: social networking sites.

The snake bites its own tail: there’s been a tipping point of tipping point tools, with thousands of startups attempting to capitalize on what they perceived as a trend, and which can be summarized to the meteoric rise of MySpace, LinkedIn and Facebook.

This shortcoming may have something to do with a fatal flaw in Gladwell’s theory. If you believe Yahoo’s social network theorist Duncan Watts, Gladwell’s “Influentials” theory is all rubbish. In wild computer simulations of “meme” spread, and real-world social experiments, Duncan Watts has found time and again that there are no laws of propagation; that in fact, the average Joe is just as likely to spread the message as a so-called “influential”.

In other words, anyone can be influential.

Traackr: a new theory of influentials

First off we would reformulate the Second Law as: content is king. TV knows a little something about that. It’s anyone’s guess what makes content sticky, so to try and define stickiness has more to do with magic than marketing. Anyone have Harry Potter in their address book?

The basic corollary of this law would simply be to say that a site that plans to engage users on a regular basis has a better “capture” potential than a mere tool. In that sense, Facebook’s “social utility” is a misnomer. It’s not a utility, it’s a hangout, in ways that MySpace is not. MySpace went overboard on the personalized homepages. Users are less interested in self-expression than in communication. Duh! We came from the savannah, where knowing where the food, the water, the fun and the danger were ranked much higher than showing off the latest bird feather.

Also, the Third Law is a tautology, a.k.a rubbish. Of course context is everything, but you need almost 100 years to figure out what the context was at any given time. That’s the amount of time it takes for the noise in the system to die down and the actual signal to appear. 100 years from now we will all see what a great president George Bush actually was. Just kidding.

Finally, anyone can be a connector, a maven or even a salesperson in turn. It all depends on what the topic is. And that’s the key to successful nano-marketing: the topic, not the influential!

Traackr is a platform that shows users that what they do on the Web creates a sphere of influence. It teaches them to focus their efforts and leverage their presence into greater influence. It delineates areas of expertise and connection. And it recognizes the power of every individual to impact several decisions, whether political or economic. Emarketers need to recognize this too, and give up on the pipe dream of influential elites. Endorsements will continue to make sense; but viral campaigns are not about buying influence; they’re about understanding how influence emerges and operates.

Companies like Buzz Agent have been surfing Gladwell’s wake – but rather than finding the tipping point, they seem to have crested and crashed. It seems that promises to manufacture influence, the way American Idol manufactures musical talent, are destined to remain an Orwellian concept. (In some ways that’s a relief). Why is this kind of approach bound to fail? If we accept Traackr’s premise that content is king, then only content will gage the credibility of a specific influencer with his or her nano-audience. Manufactured influencers can’t pass the community test. Real community influencers are all too aware that their credibility is their main currency and will be hard-pressed to compromise it. Does this mean that influencers and marketers can’t work together? Not at all, but the rules for such a relationship have yet to be set. Nano-marketing is the first and only step marketers can take to reach out to influencers.

Authors: Laurent Liscia & Pierre-Loic Assayag

Summary: Every social networking Web site claims to have a “viral accelerator”, a mechanism by which joining the site causes other people to join, in a never-ending beneficial feedback loop. Alas, the loop has only benefitted those sites and not the marketers they ought to be selling to. Why? Two reasons. First, these sites are missing a key ingredient: a good targeting platform, such as Traackr. Second, because playing the “viral game” means relinquishing a degree of control over your brand, which is anathema for marketers. Will there be a choice?

The only beneficiaries of the social networking craze so far have been the infra-structure players, whose “viral accelerators” have exemplified ways in which you might get new users on board.

Viral accelerators only work when there is a reward for the user willing to spread the “virus”. Some examples:

• Users benefit directly when they get other people to sign on. The old-style version of this was the pyramid scheme, and the problem with that is that with each passing iteration, the number of “winners” diminishes. The new style version is to just get someone to join with you and you both benefit: by becoming my friend you immediately gain a friend. On paper, LinkedIn has the most compelling viral accelerator out there: it’s a business site with a clear business purpose, a focused message, and a seemingly compelling business reward – although what you end up doing with all those hard-won connections is a mystery…
• Offering tools that require multiple user participation: when you join Airset you usually need other people to join with you for it to work. This is a good marketing model, because chances are people are going to be spending some time on Airset.
• Offering to import address books for viral purposes: less work for the user
• Opening up your API to let others figure out what might stick – think of Facebook apps. This is an indirect viral accelerator, and perhaps the most powerful one.
• Offering a shared space for content: chances are, friends of the creator will come check out what she or he has posted.

Hold on: if LinkedIn has the “best” viral accelerator, why is Facebook more popular, and even replacing it in some respects?

Just think about what goes at your office. Do your co-workers prefer those long-winded vendor meetings or gossiping about Bob and Geena at the water-cooler? In other words professional networking is boooooring, and impromptu banter much more titillating to the primates that we are. To use Tipping Point terminology: Facebook’s content is “stickier” than LinkedIn’s. Of course, that’s a glib statement, and we’d be more interested in finding out why titillation is critical to us – but we’ll leave that to Robert Sapolsky.

This means LinkedIn will always remain a niche player (albeit in a vast niche), and Facebook will take over the social networking world. Here’s an enlightening statistic on this topic: 20 million registered users on LinkedIn and 3.2m active last month vs. 70 million users on Facebook and 30 million active in the last 24h! Who would you rather reach: occasional or captive users?

Wait, Facebook already has already taken over the world. So much for prophesying.

The vacuum of social networking

However, if you scratch the shiny surface of social networking with even a short unfortified nail, what do you find as a marketer? A gaping abyss, colder than Melville’s vasty deep. Look at the ads on Facebook: are you having late-night informercial déjà vu?

“Get better at dating”, “Work on your smile”, “The thrill of gambling”.

That’s because marketers don’t trust Facebook. There’s no easy way to understand the data Facebook collects, and my guess is that Facebook does a terrible job of explaining it. In fact their advertisers section sucks.

What about LinkedIn, Hi5, FastPitch and other such, then? They’re business-focused, not tainted by the “catch-all” aspects of Facebook, and have a wealth of info to mine.

Several issues: how much do people actually use them and why? And what does it mean for a user to have 500 friends or connections? What does that say about that user other than they should probably be reported to their employer for spending most of their waking hours promoting themselves rather than their company?

What’s missing? Well, Traackr of course

Above and beyond the quality of information, what’s missing here is the next step in emarketing: the ability to use all this social networking and sharing information to reach audiences in new ways. That’s because social networking sites are harnessing user influence strictly as an acquisition tool; but as we’ve seen, this is a bit circular. Our prediction is that ultimately this will fizzle out.

Influence needs its own marketplace so users can find out whether they are influential and capitalize on their influence, whatever it may be.

Influence needs its own marketplace so

a) Users can find out whether they are influential and capitalize on their influence, whatever it may be.

b) Marketers can hone in on different kinds of micro-influence and develop campaigns that are more precise and subtle than ever

Traackr does just that. We do not accept Malcolm Gladwell’s “Animal Farm” premise: some users are more equal than others. This, in fact, flies in the face of evolutionary wisdom. Instead, we believe that every user has influence of some sort. What’s important is define what kind of influence: Is it specific expertise? The cool factor? A number of influential connections? Traackr neatly defines these types of influence and create sub-groups in which users can calculate their influence and watch their popularity, reach and buzz scores fluctuate. We help users visualize influence as an asset that can be grown. This in turn can be quite attractive to marketers since now there are clearly defined clusters of influence that span many areas and many sites.

In other words, Traackr will be the marketplace for influence and a sweet spot for emarketers, but this will raise another set of issues: will emarketers be able to relinquish some control and let users make of the brand what they will? This is key to the influence marketplace, because you cannot buy influential users. You can only court them.

We will be presenting Traackr at the WebInnovators Group meeting tomorrow in Cambridge. Come see us!. It’s a free event:

http://webinno17.eventbrite.com/

Date
Wednesday, April 2, 2008

Time
6:30 pm ET – 9:00 pm ET

Location
Royal Sonesta Cambridge
40 Edwin Land Boulevard
Cambridge, MA 02141