It’s about relationships

The web isn’t about — or just about — content or communication, of course. It’s really about making connections — about, well, links. But to be valued, those links must be organized around trust. In short: It’s about relationships.

The greater the value of the relationship, the greater the value of those who enable those relationships. Those who figure that out succeed (and those who try to impose the old models and relationships fail): Google connects us to information at a low level with sufficient accuracy (we trust it enough) and succeeds thanks to high efficiency and high volume. eBay connects people in transactions at a higher value by creating a system of trust. Whoever uses online functionality to connect, say, employers to the right executives or homebuyers to the right homes — replacing today’s middlemen — can bring even more value.

The first services that come along tend to free up information that used to be controlled by middlemen: It’s what Craig’s List did to classified apartment ads, what SelectQuote did to insurance rates (as Freakonomics points out).

But in certain high-value transactions, no one has quite cracked that next level: using trust to make the right connection efficiently.

Look at jobs. As I’ve said often, the new, inexpensive, efficient marketplaces where buyer and seller meet (Craig’s List and Monster) replacing the old, expensive, inefficient ones (newspaper classifieds) are only a first step, a waystation on the way to a distributed future. Job listings and resumes are now popping up everywhere and anywhere online. So the next services to come along aggregate that (distribution and aggregation is the constant ebb and flow of this new world). Many services are now scraping jobs from any and all sources: Indeed, Simply Hired, Yahoo now, Flipdog was until it was scooped up by Monster, Workzoo still is after being bought by Jobster, and DirectEmployers’ JobCentral has been doing it for sometime (as Stone points out, via Fred).

Makes you think that you need the aggregator of aggregators, eh? But, no, what you need to sift through this, to find value and efficiency, is trust: Relationships. Of course, Linked In wants to do that but I still don’t get the service — that is, I subscribe, but I don’t grok it, because it isn’t efficient for me; it gives me relationships I don’t necessarily care about or that can even be a bother. Look, too, at Ev Williams’ frustration trying to find a service to manage his recruiting. No one has cracked this nut.

This riff is inspired — as much pondering often is — by Umair Hague, who said that scraping jobs is just a first step:

I think the real problem in the jobs vertical is a coordination problem – not a search problem. That is, how do you connect job-seekers and decision-makers economically?

Why do I think this? Think about what headhunters do – it’s equal parts coordination and search. Sure, they have to find the right candidates; but they also have to build relationships with the right decision-makers. So far, the basic model on the Net for jobs has only ever really addressed the first half of this equation. I’ve argued this many times before; archive it if you’re really interested.

Matching jobs and job-seekers is fairly liquid on the net – the relative gap, and where Yahoo could build a strong sustainable advantage fast, is the coordination half of the value prop (which has been ignored for a veeeery long time, most likely because it’s easier to build a few forms, a spider, and a database, than it is to build a whole new innovative coordination model).

LinkedIn, I think, gets this coordination problem, and has a particular take on it, which I think is kind of brittle and rigid (ie, transaction costs for referrals are pretty high, and expected gains aren’t that high on average). If you ask me, the dominant design will be (a lot) more plastic, and it does not have to be centered around ‘link to me’ style social networking (but I can’t say more unfortunately).

The problem is finding the right — the sufficient and efficient — level of trust needed to make the right connection, the appropriate relationship. A service that doesn’t provide enough trust won’t work (in the old world, it was the problem of putting your phone number in the paper to sell a car to people you don’t trust; in the new world, anything you do not trust is spam). But on the other hand, an effort to establish excess trust — to make me friends when all I want to do is complete a transaction (see Tribe or Linked In) — is too much; it’s inefficient.

I’m not sure how to do this in jobs. If I were, I’d create the business and make a damned fortune. I’d start by building on top of the aggregators and help people find the right jobs and get mutual recommendations specific to the task at hand. That’s how headhunters work: When they call me to get names for a search, I take the time to reply because maybe someday they’ll come searching for me. That’s also how it works with friends and colleagues: I will recommend or help the ones I respect and trust and it’s not wholly selfless; someday, maybe they will help me. But the networking is specific to the task at hand; I don’t want to become buddies with all these folks and have to declare myself friends with some ongoing, public obligation (that’s where Linked In stumbles). I want a system that lets me call on them when I need them and rewards me when they need me. I still don’t know how to do that. Umair’s right: That is the real challenge.

So let’s switch to another leg of the classified stool: Real estate. In no sane universe, of course, does a real-estate agent earn 6 percent commission for the services provided. Now Craig’s List and search engines can take away half their job by sharing information and breaking apart the monopolistic trust (in the bad sense of the word) that multiple listings services are: We can find out what’s for sale and how big it is and the price.

But what they don’t do is solve the trust problem: Real estate agents arrange for all those damned strangers to tour my house. At one level, it might be wise for a newspaper or Craig to simply offer a scheduling service (though if they take money from realtors, they are disinclined to do that): They could handle the hassle of matching schedules and then increase the level of trust by verifying the identities of the prospective homebuyers, by requiring and vetting credit-card information, for example. It would take a trusted agent to do that for both sides of the connection. On an even higher level, someone could start a new business that does nothing but arrange and schedule visits and physically accompany buyers on the visits: Say, a bonded moonlighting cop trusted to have my key does this to make sure my house and I are safe but without making any effort to actually sell my house (which he can’t do without a license); I could pay the same person to wait at home for me to let the damned cable guy in. As a homebuyer, I would prefer that to a Realtor: I am armed with all the information I need about the homes — now distributed and aggregated openly — and I don’t have to suffer through a sales guy’s unctuous blather.

Now take this same calculation to other areas.

Take it to cars: Auto web sites give me the information I need about a car; CarsDirect and an Edmunds give me the information I need about pricing. Since all the cars are the same, I really don’t need a trusted agent to sell me one (and I don’t trust him anyway). If federal law would allow it, I’d just buy the car directly from Toyota and buy repair insurance (aka a warranty) and get it repaired wherever I want to by whomever I trust. That’s the way that industry should work; regulation protecting old sales channels prevents it.

Take it to marketing: If you use this medium wisely, you don’t treat it as a medium. You treat everyone you find here as a consumer, as a customer, as a person. You find the people who help you improve your products (alpha consumers); you find the people who help you market your products (influencers); you find the people who help each other with your products (life is customer service); you make good products and get out of the way. If you’re lucky, you don’t advertise. You create a sufficient and trusted relationship with consumers to answer questions and solve problems and keep them coming back. Are you listening, Dell?

And take it to news: The first step is to break up the old, centralized, controlled marketplaces of information and distribute news everywhere (put it online). The next is to aggregate it or help find it (see GoogleNews). The next is to solve the problem of overload and confusion via trust (see blogs). Note that the journalists no longer hold a monopoly on trust (if they ever did); in a distributed/aggregated world, we chose whom to trust.

That is what the internet is about: not just distribution, not just information, not just transactions. It is a web of humans. It is about relationships.

Information + Aggregation / Trust = Relationships

: I STAND CORRECTED: j.d. in the comments shows why I avoided higher math like the plague. What I meant to say is:

(information + aggregation) / trust = relationships

It’s about relationships

It’s about relationships

: The web isn’t about — or just about — content or communication, of course. It’s really about making connections — about, well, links. But to be valued, those links must be organized around trust. In short: It’s about relationships.

The greater the value of the relationship, the greater the value of those who enable those relationships. Those who figure that out succeed (and those who try to impose the old models and relationships fail): Google connects us to information at a low level with sufficient accuracy (we trust it enough) and succeeds thanks to high efficiency and high volume. eBay connects people in transactions at a higher value by creating a system of trust. Whoever uses online functionality to connect, say, employers to the right executives or homebuyers to the right homes — replacing today’s middlemen — can bring even more value.

The first services that come along tend to free up information that used to be controlled by middlemen: It’s what Craig’s List did to classified apartment ads, what SelectQuote did to insurance rates (as Freakonomics points out).

But in certain high-value transactions, no one has quite cracked that next level: using trust to make the right connection efficiently.

Look at jobs. As I’ve said often, the new, inexpensive, efficient marketplaces where buyer and seller meet (Craig’s List and Monster) replacing the old, expensive, inefficient ones (newspaper classifieds) are only a first step, a waystation on the way to a distributed future. Job listings and resumes are now popping up everywhere and anywhere online. So the next services to come along aggregate that (distribution and aggregation is the constant ebb and flow of this new world). Many services are now scraping jobs from any and all sources: Indeed, Simply Hired, Yahoo now, Flipdog was until it was scooped up by Monster, Workzoo still is after being bought by Jobster, and DirectEmployers’ JobCentral has been doing it for sometime (as Stone points out, via Fred).

Makes you think that you need the aggregator of aggregators, eh? But, no, what you need to sift through this, to find value and efficiency, is trust: Relationships. Of course, Linked In wants to do that but I still don’t get the service — that is, I subscribe, but I don’t grok it, because it isn’t efficient for me; it gives me relationships I don’t necessarily care about or that can even be a bother. Look, too, at Ev Williams’ frustration trying to find a service to manage his recruiting. No one has cracked this nut.

This riff is inspired — as much pondering often is — by Umair Hague, who said that scraping jobs is just a first step:

I think the real problem in the jobs vertical is a coordination problem – not a search problem. That is, how do you connect job-seekers and decision-makers economically?

Why do I think this? Think about what headhunters do – it’s equal parts coordination and search. Sure, they have to find the right candidates; but they also have to build relationships with the right decision-makers. So far, the basic model on the Net for jobs has only ever really addressed the first half of this equation. I’ve argued this many times before; archive it if you’re really interested.

Matching jobs and job-seekers is fairly liquid on the net – the relative gap, and where Yahoo could build a strong sustainable advantage fast, is the coordination half of the value prop (which has been ignored for a veeeery long time, most likely because it’s easier to build a few forms, a spider, and a database, than it is to build a whole new innovative coordination model).

LinkedIn, I think, gets this coordination problem, and has a particular take on it, which I think is kind of brittle and rigid (ie, transaction costs for referrals are pretty high, and expected gains aren’t that high on average). If you ask me, the dominant design will be (a lot) more plastic, and it does not have to be centered around ‘link to me’ style social networking (but I can’t say more unfortunately).

The problem is finding the right — the sufficient and efficient — level of trust needed to make the right connection, the appropriate relationship. A service that doesn’t provide enough trust won’t work (in the old world, it was the problem of putting your phone number in the paper to sell a car to people you don’t trust; in the new world, anything you do not trust is spam). But on the other hand, an effort to establish excess trust — to make me friends when all I want to do is complete a transaction (see Tribe or Linked In) — is too much; it’s inefficient.

I’m not sure how to do this in jobs. If I were, I’d create the business and make a damned fortune. I’d start by building on top of the aggregators and help people find the right jobs and get mutual recommendations specific to the task at hand. That’s how headhunters work: When they call me to get names for a search, I take the time to reply because maybe someday they’ll come searching for me. That’s also how it works with friends and colleagues: I will recommend or help the ones I respect and trust and it’s not wholly selfless; someday, maybe they will help me. But the networking is specific to the task at hand; I don’t want to become buddies with all these folks and have to declare myself friends with some ongoing, public obligation (that’s where Linked In stumbles). I want a system that lets me call on them when I need them and rewards me when they need me. I still don’t know how to do that. Umair’s right: That is the real challenge.

So let’s switch to another leg of the classified stool: Real estate. In no sane universe, of course, does a real-estate agent earn 6 percent commission for the services provided. Now Craig’s List and search engines can take away half their job by sharing information and breaking apart the monopolistic trust (in the bad sense of the word) that multiple listings services are: We can find out what’s for sale and how big it is and the price.

But what they don’t do is solve the trust problem: Real estate agents arrange for all those damned strangers to tour my house. At one level, it might be wise for a newspaper or Craig to simply offer a scheduling service (though if they take money from realtors, they are disinclined to do that): They could handle the hassle of matching schedules and then increase the level of trust by verifying the identities of the prospective homebuyers, by requiring and vetting credit-card information, for example. It would take a trusted agent to do that for both sides of the connection. On an even higher level, someone could start a new business that does nothing but arrange and schedule visits and physically accompany buyers on the visits: Say, a bonded moonlighting cop trusted to have my key does this to make sure my house and I are safe but without making any effort to actually sell my house (which he can’t do without a license); I could pay the same person to wait at home for me to let the damned cable guy in. As a homebuyer, I would prefer that to a Realtor: I am armed with all the information I need about the homes — now distributed and aggregated openly — and I don’t have to suffer through a sales guy’s unctuous blather.

Now take this same calculation to other areas.

Take it to cars: Auto web sites give me the information I need about a car; CarsDirect and an Edmunds give me the information I need about pricing. Since all the cars are the same, I really don’t need a trusted agent to sell me one (and I don’t trust him anyway). If federal law would allow it, I’d just buy the car directly from Toyota and buy repair insurance (aka a warranty) and get it repaired wherever I want to by whomever I trust. That’s the way that industry should work; regulation protecting old sales channels prevents it.

Take it to marketing: If you use this medium wisely, you don’t treat it as a medium. You treat everyone you find here as a consumer, as a customer, as a person. You find the people who help you improve your products (alpha consumers); you find the people who help you market your products (influencers); you find the people who help each other with your products (life is customer service); you make good products and get out of the way. If you’re lucky, you don’t advertise. You create a sufficient and trusted relationship with consumers to answer questions and solve problems and keep them coming back. Are you listening, Dell?

And take it to news: The first step is to break up the old, centralized, controlled marketplaces of information and distribute news everywhere (put it online). The next is to aggregate it or help find it (see GoogleNews). The next is to solve the problem of overload and confusion via trust (see blogs). Note that the journalists no longer hold a monopoly on trust (if they ever did); in a distributed/aggregated world, we chose whom to trust.

That is what the internet is about: not just distribution, not just information, not just transactions. It is a web of humans. It is about relationships.

Information + Aggregation / Trust = Relationships

: I STAND CORRECTED: j.d. in the comments shows why I avoided higher math like the plague. What I meant to say is:

(information + aggregation) / trust = relationships

  • http://www.evolution-nextstep.com j.d.

    Did you mean
    (information + aggregation) / trust = relationships
    or
    information + (aggregation / trust) = relationships, which is what you wrote.
    Order of operations, you know. Big diff.
    ;)

  • http://www.simplyhired.com Dave McClure

    hey jeff -
    great piece, and definitely agree including trust & networking data can change the way we currently do other functions online.
    re: your comment (question?) on how social networking can help with a job search, you might want to check out how we partnered with LinkedIn to integrate their functionality on our search engine, SimplyHired.com. we made it easy to just click on a ‘who do i know?’ button on any of our 3.5M job listings, and use your LinkedIn network to find inside contacts at companies you might want to work for.
    for more on how it all works, check out Dave Taylor’s post or read the Simply Hired blog on the LinkedIn-SimplyHired mashup.
    regards & keep up the great blogging:)
    - dave mcclure
    SimplyHired.com
    less crap. more jobs.

  • http://www.evolution-nextstep.com j.d

    Just thought I’d have a bit of fun with you.

  • http://www.memeorandum.com/ Gabe

    I still don’t like the formula! Currently, the greater the trust, the lower the relationships. And even worse, when trust=0 relationships=infinity. I think the / should be a *, so when trust is 0, relationships=0.

  • http://www.evolution-nextstep.com j.d.

    That’s a good point, and in the light of your comment I almost think it ought to be ^ instead of * or /, so relationships tend to infinity exponentially with trust, to emphasize the importance of trust.

  • http://www.buzzmachine.com Jeff Jarvis

    Gabe, jd: I thought of that but I actually think it is divided by because trust gives us the subset — the creme le la creme — of a universe, no?

  • http://www.evolution-nextstep.com j.d.

    Then, the operation you want is “mod”, short for modulo. All the non-trustworthy folks are “modded out.”
    You must wish by now that you’d left the math out of it. :)