We Are Pleased to Present… Trada!

March 18, 2010 By 5 Comments

After almost 18 months, the whole team here is extremely pleased to present Trada to the world. As you learn about what Trada is doing, we hope you get as excited as we are. We fundamentally think we’ve invented, developed and continue to perfect a totally new way of doing paid search. It’s a way that aligns everyone’s goals so everyone can win. It works for big companies and small companies alike and it takes advantage of a fundamental principal: if organized correctly, all of us can be smarter than any one of us.

The Trada journey started almost 21 months ago literally as a light bulb idea. I was actually with my parents, on a boat in Scotland, heading into the only unnavigable whirlpool in the Northern Atlantic. The Corryvrecken has been known to have 30 foot swells and multiple whirlpools all at the same time. Sounds exactly like what you’d want to drive a boat into, doesn’t it? Well fortunately for me we had calm seas that day and the long boat ride toward the edge of Jura left me a lot of time to wander in my thoughts.

I had spent the last 6 months helping Verne Global start the datacenter industry in Iceland. After months of trips between Boston, Reykjavik and London (and enough Brannvin to last a lifetime) I decided to head back to Boulder and start something new. For the few months prior, Seth Levine, my good friend and many-time investment partner, had been tossing ideas around with me in the advertising space. Both of us felt that online advertising was the new Internet “wild west.” It had an incredible and consistent growth rate and an even more incredible pace of change. While the numbers in overall online advertising are staggering (>100B depending on what you count between search, display, video, sponsorships, etc.), it’s still a drop in the bucket of traditional offline advertising. If you include all print, yellow pages, radio, television etc. in traditional advertising it’s almost a side note. Nothing is more fun than building companies in these types of wild-west environments. There is usually a huge ability to either help a class of customer deal with increasingly complexity or to massively optimize the inefficiency that occurs when markets grow very fast. As we hope to prove to everyone, paid search had both of those characteristics.

As I was on that boat heading into the Corryvrecken I was thinking about my own experience managing AdWords campaigns directly. I was managing a campaign that was spending about 8k/mo. Depending on who you talk to this is either really big or really small (it constantly amazes me to read about companies spending $4M on paid search a month). I remembered two things, 1) how addictive AdWords was because of my affection for analytics and 2) how time consuming and complex it was to actually optimize the results into an affordable outcome. I was literally spending 4 hours a day inside of AdWords tweaking bid prices, changing ads, researching new keywords, etc.. And then it dawned on me – why should I become the expert when I know there are many people out there that already know way more about this than I do. How do I find them? And how do I know if they are doing the best they can? How do I pick one over another? If I pick one how will I know they are doing the best job that’s possible? Well, what if I could just have them all? And pay them when they succeed. Thus the idea for Trada was literally born in an instant, on a boat, heading into a whirlpool.

While the Trada marketplace has matured vastly since we launched in very privately in Sep 2008, this basics of this original idea have never really changed. We still fundamentally believe that a group of experts working on a complex problem that needs constant attention is much more powerful than any one of them. It’s partially about the diversity of thinking that occurs when you get a lot of minds working on something, and it’s partially about the amount of time a large group can devote to something.

To this end, we help advertisers find groups of experts through our marketplace. We change the way advertisers approach paid search by getting them to focus on what matters for their business (such as what is an affordable price they can spend to sell something) instead of getting mired in the details of keyword and bid prices. We give paid search experts a way to use the well-earned skills they have without having to find, recruit and manage customers. And we give everyone a financial incentive to succeed: advertisers get very sophisticated campaigns tailored to their business metrics and goals, optimizers are incented to beat advertisers’ stated goals by doing the skilled work of optimization, and Trada sits in the middle making sure the market runs smoothly. In the end, it’s a simple system to help everyone win while working on a very complex problem.

As we launch Trada we have decided on a tagline that describes what we aspire to be. Every day we work hard to succeed at living up to this summary. We hope we are and if we’re not, we’d love to hear from you about how we can.

Ridiculously Easy. Measurably Better. Powered by People.

Filed Under: Trada News Tagged With: ,

Google: A One-Click Pony?

March 17, 2010 By Leave a Comment

This Pony Needs Tricks

While there are countless daily articles about impending threats to Google’s hegemony (many of which are partially fair), what most people overlook is that Google has an immense amount of options in which to continue to grow its core business.

But “Google’s just a one-click pony,” you say. They make 90% of their revenue from search, and they have that maxed out, right? Here’s where I start to disagree.

First of all, Google actually does one thing very well: run massively scalable auctions letting advertisers match their ads to something they would consider a target. When people think of paid search, they think of bidding on keywords. What they are really doing is matching an ad to a keyword with the right bid price. If you replace the word “keyword” with “web page” or “video” or “mobile app” you see very quickly that what Google has built for AdWords actually translates very well into all the other mediums they are starting to talk about.

Display advertising, setting aside the process of creating image ads (which is something that Google helps advertisers do now) is simply the process of matching banner ads to websites or URLs that they are effective on (and pricing them correctly). Mobile applications are no different. Recently Google introduced a whole new set of tools for advertisers to self-service advertising in YouTube videos, video by video.  Guess what? It’s now a process of matching ads to videos, for the right price. As long as we keep coming up with different ways for Google to let advertisers target, Google’s got very deep pockets to go there. While paid search is relatively common now, Google is just barely scratching the surface of other advertising mechanisms they have in place.

One argument can be made though that Google has no way to target an ad to a person (e.g. a Facebook page). While this is true (and quite a powerful performance based marketing option), I think over time Google will do a lot to create more of a “presence” of a person on the web. I don’t know if this will be through their tools, Google Wave, Buzz or an acquisition of something, but once they do, they’ll have the last major existing targeting category down.

One downside to all of this opportunity is that the complexities of doing paid search will also translate into the other mediums they are exploring. Finding, managing, bidding and optimizing 1000s of keywords is the same thing as finding, managing, bidding and optimizing 1000s of URLS, mobile web pages, or videos. The complexity just comes from another variable. This is an issue that Google must constantly keep in mind.

To that point, the other major dimension that Google has to work with is expanding their customer base for existing products. While paid search is somewhat commonly known as a marketing option, it is very uncommonly used by small- and medium-sized businesses in the US and has even less penetration globally. There are a number of reasons for this. Primarily, I think Google’s tools simply have too high a learning curve for many business owners to use. You have to invest a lot of time and money to get them working well for you. That’s prohibitive. A whole generation of business owners would rather just pay $3,000 a year for a Yellow Page ad that takes 15 minutes to dictate over the phone, even if it never produced anything for them! While I won’t go on a long tangent about Trada, providing a model where SMBs have access to this kind of expertise is exactly what we do.

As Google goes forward, they have a massive opportunity to grow their core business and diversify what people see as their one-trick pony – paid search. By the end of 2011, we’ll see a significant part of their revenue contributed from non-paid search ads. I think their revenue will mostly still be from performance-based advertising but not necessarily keywords.

Filed Under: Paid Search Tagged With: , ,

The Fallacy of the Head of the Tail

March 15, 2010 By Leave a Comment
Heads or Tails

Heads or Tails?

A lot of companies that have tried paid search simply don’t have the time to explore and manage the middle and long tail of keywords. Matt Hessler wrote a great blog post about the value of the long tail and how Trada is designed very well to do this for you. But what about the head of the tail? Don’t the top 25-100 keywords produce the majority of your traffic? Why even bother trying harder? Lots of people point to these keywords, and say “I can just focus on those and I’ll do fine. I can spend a little bit of time, and get 50 to 75 percent of the volume I’d be getting if I explored the long tail.”

This is true if you’re only looking at clicks as the measure of effectiveness. It’s when you dig even one layer deeper that you see the fallacy of the head of the tail. Calculate some other metrics, and you’ll see what I am talking about:

  • Cost per Click: your average cost of a click in paid search
  • Conversion Cost: the actual cost of paid search traffic for each conversion (sale)
  • Conversion Rate: the conversion rate of traffic (e.g. % of visitors that buy)
  • Bounce Rate: a dull edged but simple metrics for traffic relevancy
  • Time on Site: how long does a visitor stay on your site when coming from paid search
  • Cost per New Visitor: what is the cost of a new visitor (slightly different than cost per click)
  • Pages Views: how many pages does a visitor look at when delivered from paid search

If you’re not familiar with some of these metrics, some of them can be calculated right in AdWord (or YSM) and some of them are available through Google Analytics integration (the last 4).

Once you start to examine these metrics you might find that head of the tail terms are not as effective as you might like them to be. Is your conversion cost in line with what you can afford or are you overpaying for every sale? Do many of the visitors from paid search traffic bounce right away (e.g. have a high bounce rate) indicating you weren’t what they are looking for?

In my view there are two types of head of the tail terms: branded terms (e.g. “Clorox”) and categorical terms (“buy blackberry”). There is much debate if you should bid on branded terms or not, so I’ll leave that for another post. If you look what’s underneath categorical terms, you’ll quickly see the problems that arise. Let’s assume for now that you’re doing the simplest thing possible: using broad match on categorical terms. Here’s a number of problems you’ll encounter:

  • Expensive bid prices: because of the paid search auction model, more common terms are more expensive. That’s because everyone can think of the same 50-100 terms that many people would type into Google for your products. There is a natural inflation of the prices of these keywords simply because of the demand for them. Can your business afford this price inflation?
  • Broad match follies: broad match is a dangerous beast. When you enter keywords into Google with broad match, you give Google (et al) broad license to match search terms to your keywords conceptually. This means that if the search engine believes there is a connection between what a searcher is looking for and your keyword, they will show your ads. Now all the search engines have some smarts but they also are imperfect. While “running shoes” might match a search term of “athletic shoes” it might also match a search term of “used running shoes.” If you don’t sell used running shoes then you’re going to spend some money on clicks you don’t want.
  • Negatives: many people that focus on head of the tail terms due to time constraints don’t have the time to put in negative keywords. Negatives allow you to tell the search engine “if this word appears in the search do NOT show my ad.” In our above example, if you had a negative keyword of “used” you’d avoid the situation I mentioned. But on a broad match categorical term the number of negatives that you need to put in to prevent all the crazy matches can be huge. Why spend your time putting in negatives when you can put in longer-tail terms!
  • Bid pricing: its very easy to spend a lot of money on head of the tail terms. The volume is high and if you’re not watching it your costs (and CPAs) can skyrocket. Head of the tail pricing actually requires much more diligent bid price optimization and attention. As the buyers ebb and flow on the web, the $0.80 cent bid of last week might actually need to be $0.60 this week. If you start to get conversions, what’s the optimal bid price based on the conversion rate? You need to carefully hand tweak these prices constantly (especially between ad networks)
  • Day parting: all the ad networks let you say what times of the day you want to run your ads. With head of the tail terms the need for dayparting is much more dramatic. Someone typing in “birthday cakes” at 3:00am is pretty unlikely to buy something. While this issue still exists with longer tail keywords, someone typing in “who sells overnight birthday cakes” at 3:00am is probably still a pretty viable customer.

All of this is to say that the head of the tail can actually require as much time to manage as the tail. One of the benefits of Trada is that our optimizers do that. Their goal (and incentive) is to find everything that works and to finely optimize it to the best outcome against your business goals. Having 10s of optimizers constantly watching these things (and knowing how to look for them) can be as valuable in saving you time and money as having them explore and optimize the long tail.

Filed Under: Keywords Tagged With: , , , , , ,

Ruby: A Love/Hate Relationship. Code is Fickle.

March 11, 2010 By Leave a Comment

The engineers here at Trada have written the marketplace application in a language called “Ruby,” and we use an extension to Ruby that makes for quick development and prototyping of websites called “Rails.”  You may have (not) heard of Ruby on Rails, but it’s all the rage in web development circles, requiring less overhead and a better framework than another competing framework that uses the programming language Java, called J2EE (or Java 2 Enterprise Edition).

Before I get started on my rant/praise of Ruby, let me avoid the spate of Ruby lovers piling on their scorn for my criticisms by saying after many tortuous months, I like Ruby.  There, I said it.  I still find things about it that are infuriating, but I’m pretty much assimilated at this point.  Resistance is futile.

I can’t think of a better way of developing a non-static website than rails.  Websites come in two varieties—static, like the Trada corporate site, and non-static, like the Trada marketplace or Amazon.com.  Static websites may collect information (like email addresses, sales leads, etc.), but they’re pretty much what they say they are—presentation, but not a lot of activity behind it.  Non-static websites are ones that do things behind the scenes called “business logic.”

Coding Since the Time of Angels

With that out of the way, let me give you a little background.  I’ve been a developer since I was a teenager—about the time that Farrah was still an Angel (if you have to ask what that means, it means I’m OLD).  I first programmed on a paper-tape machine using Basic, but wound up writing assembler for an Altair a friend of mine had to make it do rude things.  Yes, I’ve actually used Cobol and Fortran.  No, they haven’t been on my resume in years.  I taught Pascal (whatever that is) as a graduate student at CU.

I adopted the programming language C and added Intel 8086 and Motorola 68000 assembly languages after college, then C++, Java, C#, and about 10 other languages only geeks have heard of as they became available (Bourne shell, Forth, Logo, Perl anyone?).  About 8 years ago, I started hearing of a new generation scripting or interpreted programming languages—Python, Ruby, and Lua (if you use Second Life, you may run across Lua).  I learned all three and liked them in the following order Python, Lua and Ruby.

Why did I like Ruby the least?  It seemed like a hacker’s language (and I don’t mean that as a compliment)—it used special characters an awful lot and used the same special characters in lots of different ways, making the code hard to follow, much like the language Perl.  Perl is such an ambiguous language, that on more than one occasion, I’ve had to bring in a “Perl Guy” to write some scripts to use it.  When that Perl Guy left, I’d have to go out and get another Perl Guy.  Invariably Perl Guy 2 would be unable to understand Perl Guy 1’s code and would just chuck it and start over.  Same goes for Perl Guy 3, 4, etc.

[Read more...]

Filed Under: Trada News Tagged With: , ,

Why The Long Tail is Worth The Effort

March 10, 2010 By 7 Comments

Pursue the Long Tail (Rat Not Included)

Many people credit Chris Anderson with coining the term “long tail” in his 2004 Wired magazine article. His explanation of the long tail was applied to online merchandising, but it was quickly adopted by advertisers using Pay Per Click search marketing to drive traffic to websites with huge volumes of SKUs.

“Long tail” keywords are highly specific search phrases of three or more words that are less common than shorter more general terms for a product or service.

Most search marketers find that researching and implementing a long-tail strategy is both difficult and time consuming. Even with keyword permuters and other research tools, it is difficult for even a PPC expert to envision all the ways a user might search for their product or service. If you combine this hardship with the fact that typically the top 100 keywords in a campaign drive 60 to 90 percent of all clicks, the question becomes: why bother?

The answer to this is best illustrated by some sample campaigns in the Trada Marketplace. This advertiser has a fairly robust campaign with nearly 40k keywords. The top 40 keywords only (1 percent of the total campaign keywords) make up nearly 50 percent of the all the traffic for the campaign, however, these same top 40 keywords are only responsible for 30 percent of all of their conversions.

The lazy or unsophisticated search marketer may feel that this short keyword list is good enough because it covers the lion’s share of the market. In this specific case lazy campaign management and not developing the long tail would have come at a cost of 70 percent of the conversion volume.

The reason that long tail keywords are so effective is two fold. The first reason is because longer, more obscure search phrases typically cost less. As mentioned it can take a long time to develop a deep long-tail list and since many marketers don’t take the time the competition for these terms is lower driving down the cost. The second reason is that a search user who types a very specific query is more likely to click on your targeted ad and resonate with your product or service offering because it fits the specific need they searched for.

Trada was created in part because of our belief in the success of long-tail search strategies. While we understand that long tail strategies do not work for all advertisers we believe they can enhance most advertisers’ campaigns. We understand that it is difficult for any one person to generate an exhaustive list of long tail keywords. With the Trada model you have several optimizers working on your campaign at once, all bringing their unique perspective to the campaign. The result is a thoughtfully developed long tail keyword list that will help you uncover many low cost conversions that your current campaign would not have captured.

Cultivating a thorough long tail keyword list is not a complete panacea for underperforming campaigns. It is important that you aggressively manicure the head and the middle of the tail so that these top keywords are performing optimally. Because these keywords are generating the majority of the volume they must be constantly optimized with bid-price changes, ad copy and landing-page testing, as well as match type and negatives.

Unfortunately the truth about search is that there is no silver bullet or magic formula for success. Doing PPC search marketing well is about rolling up your sleeves and doing the work. For this reason, there are literally millions of search campaigns that are underperforming. Even campaigns run by marketers with the best intentions seem to wear their owners down. They require constant testing and analysis, exploration and refinement. When was the last time you wrote new ad copy, or explored an additional 100 keywords for your campaign?

If the answer is never or not in many months, then maybe it is time to see how your campaign could perform with group of PPC experts doing this work for you.

Filed Under: Keywords Tagged With: , , ,

Don't Be Afraid to Tell Us Our Baby is Ugly

March 9, 2010 By 9 Comments

I had coffee with a friend recently to discuss social media marketing and as I was updating her on what we were up to at Trada, she asked, “How do you collect feedback from users in Trada?”

For product managers like me, user feedback is the most fundamental aspect of our jobs. Unfortunately for many of us, it too often ends up being the most ignored.  It’s never purposely ignored – it’s something that tends to happen slowly, over months and years. We get busy with other projects and feel the pressure to get new features to market quickly. Before you know it, you’ve lost touch with your users and as a result they tend to lose touch with your product.

At Trada, as an organization we work hard to not let this happen. Even before Trada was “officially” founded, we started interviewing paid search experts and advertisers (many of whom were one and the same) to find out what they found difficult about paid search advertising and determine if Trada’s fundamental model of a paid search marketplace made sense. Niel and I had both run paid search campaigns in the past and we had our opinions about what advertisers struggled with, but it was important to hear directly from them to help us fully understand the problem we were trying to address.

Shortly after we released our first prototype of the Trada marketplace, we held a series of meetings (we call them Contradas) with paid search experts (we call them optimizers) to learn what they liked and what was lacking in the marketplace. Contradas typically included 8 – 10 optimizers, lasted for 6 – 8 weeks, with a 2 – 3 hour meeting once a week in the evening. Every Trada employee attended. We’d get together with our optimizers and eat pizza and drink beer with them while they told us what they liked, and what was lacking in the marketplace. This allowed us to iterate the UI and functionality incredibly fast. Since then, we’ve also held Contradas with advertisers to gather their feedback as well. Over the course of the last 14 months or so, we’ve held 4 Contradas – about once a quarter. It’s about time for another one.

If it sounds amazing that we could get 8 – 10 highly sophisticated and successful paid search experts to spend a significant amount of time each week for up to two months to review a (then) prototype application without being paid a dime, I have to admit that I was pretty surprised myself. The key is the structure of Contradas and the type of people we recruit to participate.

FREE PIZZA + NEW FEATURES = HAPPY CONTRADA!

To start, people who are great at what they do also tend to be intellectually interested in new technologies in their industry. If you do your part, they’ll give you an amazing amount of their time and energy. But they won’t do it forever – nor should they be expected to. For that reason, our Contradas were specifically time bound and our agendas were very clear and efficient, so members knew exactly what they were committing to, and for how long.

Also, people get excited when they see their ideas become a reality in a short period of time. Notice that I didn’t say, “when users validated or rejected our ideas”. While we definitely came to our users with prototypes to react to, we were careful to ask them open ended questions that often revealed completely new ideas that we’d never thought of before. If we’d simply said, “Here’s a new feature, do you like it?” we wouldn’t have uncovered those pearls that came out as a result of open ended questions and conversation. We also encouraged criticism. We opened every meeting by telling everyone, “Don’t be afraid to tell us our baby is ugly.” Rather than getting defensive of criticism, we embraced it as an opportunity to make the Trada marketplace better. We also offered free counseling to Trada employees to help them deal with the rejection. ;)

When we received feedback from optimizers, we built it into the product very quickly – usually within a week, or two at the most. We could pull this off because, as I mentioned earlier, everyone working at Trada attended Contrada meetings. Engineers heard directly from users, which allowed them to understand the use case quickly and start development without delay. The big caveat to doing this is that you have to have a rock-star engineering team who can produce high-quality work at breakneck speed. We’re fortunate to have that dev team at Trada.

Understand that not everything went perfectly. One of the things we learned quickly is that it’s very difficult to have Contrada members who are remote and dial/web conference into the meetings. There was simply too much conversation and white boarding going on in the room for them to be able to keep up with what was being said and provide the value they wanted to.

Also, I found that when the entire staff is involved in the meetings and they’re as passionate as Contrada members about certain topics, it’s easy to get overly involved in the conversation and start talking/debating employee-to-employee. This wastes valuable time with users and you don’t learn anything talking amongst yourselves. When we realized this was starting to happen, we made a conscious effort to recognize it and put a stop to it. You have all week to fight, er, debate with your teammates – when end users are present stay focused on them!

Finally, remember that a little “thank you” goes a long way. One of our Contradas took place near Thanksgiving, so we took the opportunity to buy everyone pies from a great little bakery across the street from our office. Another took place near Halloween, so we stuffed jack-o-lantern buckets with goodies and a book. Each only cost a few dollars, but went a long ways toward showing Contrada members how much we appreciated their time.

Now that our advertisers and optimizers are literally spread around the globe, it’s a bit more challenging to gather feedback in the same way we do with Contradas. Just today we had a discussion about which tools to implement into our application to gather user feedback and different ways we can include optimizers and advertisers in our product development process. We hope our optimizers and advertisers continue to give us the outstanding feedback they have in the past – and we’ll continue to work hard to make user feedback a core part of our development process.

Filed Under: Trada News Tagged With: , ,

What We're Reading This Week at Trada

March 5, 2010 By Leave a Comment

Here is a sample of what we’re reading. Have something you think we should be reading? Shoot us an e-mail.

When you Google our fearless leader, Niel Robertson, an awesome looking picture pops up in image results. We decided to create a Flickr group that puts the best looking face forward for all of our Staff members. Leave a note in the comments or reply to us @Trada to let us know who you think is the most photographic.

Filed Under: Uncategorized Tagged With: , , ,

Wikipedia’s Decline and the 7 Types of Human Motivation

March 2, 2010 By 5 Comments

There have been a lot of stories recently reporting on the slow decline of participation in Wikipedia. Most of the stories have come to the conclusion that people are just sort of not interested in helping anymore as there is an abundance of information available on all topics. The fun of building a global utility has been usurped by the utility of using the result. Another common comment is that as Wikipedia has grown, the friction of participation has gone up (due to the necessity of enforcing quality). One now has to log in to edit, there are more administrative processes involved in making changes, and some pages are simply uneditable unless you’re a sanctioned member of the crowd. While I think there are some elements of each hypothesis that are correct, I think what Wikipedia is suffering from is what other crowdsourcing companies are learning to take advantage of: humans have different motivation systems and you have to cater to all of them.

While there is some debate over semantics, most recognize Wikipedia as one of the first and most successful examples of the crowdsourcing model. Multiple authors create something bigger than themselves to someone (everyone) else’s benefit. Wikipedia got its start many years ago (March, 2001 to be exact) before crowdsourcing was even a common term. Since then the crowdsourcing industry has matured a lot, and its understanding of what works and what makes a crowd sustainable has matured as well. I think Wikipedia’s fundamental problem is that it hasn’t kept up with the times.

One of the most fascinating things I have learned from talking to all sorts of crowdsourcing companies in the last 18 months (Mahalo, uTest, Get Satisfaction, Threadless, etc.) is that the reason for human participation is both widely varying and also not singular. Surveying the crowdsourcing landscape, I have found 7 basic types of human motivation that crowdsourcing companies integrate into their systems:

1) Monetary Compensation: The company outlines what you need to do to make money. You do it. You make some money (the amount varies widely). Rinse and repeat.

2) Points and Rewards (Non-Monetary Compensation): You do something good or you contribute in a certain way and you get some quantifiable but not directly monetizable reward. Many companies have some type of “points”. Keeping score matters here (see below for more on this). Sometimes these rewards in the end are monetizable either by exchange (turning them in for prizes) or indirectly monetizable (using points to get access to more of the system).

3) Leaderboards and Competitive Standing: Many companies let you know where you stand against your peers. Leaderboards usually reflect you standing using some other form of motivation (earned money, earned points, badges, etc.). What’s key is that you understand from the physics of the system you’re using how to positively affect your leaderboard status. Leaderboards are always public. This plays to people’s desire to compete publicly.

4) Badges and Goal Completion: The system you work with defines levels of achievement or specific goals to complete and you are awarded something (even just a graphical badge) that denotes your accomplishments. Necessary in this motivation (like leaderboards) is that the badges are publicly viewable (this is the boy scout badges concept and is basically the way Zynga learned to dominate the social gaming industry) and a core part of Foursquare’s philosophy.

5) Reputation: The system has some mechanism (usually a combination of all the other things I mentioned above) to help you express to others (and self evaluate) what your reputation is in the system. Foursquare uses the Mayor concept, Zynga uses a ranking system for player titles and Mahalo uses a martial arts belt system. All of these approaches make it easy for someone to understand that someone has a general standing greater or less than them.

6) Community: You can participate in and communicate with a community of similar people interested in similar things.

7) Collaboration: You can work collaboratively with other people on something larger than you could create yourself and the results are publicly (or at least partially publicly) on display. Your group effort is visible.

I mentioned that in any group activity there is always a diversity of motivations that people have. In any long-term situation, people need multiple types of motivation to keep them engaged. If this sounds at all familiar, it’s because you experience it every day in your personal life (why should your online life be any different). A simple example of this is how people think about their jobs. A high salary is rarely a long-term retention strategy or the best way to get the optimal contribution from someone. Some people value job security, others value recognition, some value medical benefits, and others value being able to correlate their work directly to the outcome of the company (or a division of it). Most people favor one of these (study after study shows it’s not salary at the top of the list), but always place value on more than one.

In my view, usage on Wikipedia is declining because they don’t understand the dynamic that successful crowdsourcing is about multiple forms of motivation. They have relied on only one (or maybe two) dimensions for a long time: collaboration being the primary mechanism. Wikipedia’s premise is that the motivation to contribute to something that everyone uses is enough to keep people engaged for the long haul. I don’t believe this is true. I wouldn’t believe it was true if they had used money as the sole motivation either. I think Wikipedia had a good long run with one kind of motivation and now the wheels are starting to fall off. Clearly I am not ringing the death knell of Wikipedia, rather I’m rooting for them to adapt to the changing environment they operate in.

Why not have a leaderboard that tells the world (and me) how many contributions I have made versus others? Where is the quantifiable metric of what my contribution has done (even just page view counts on all the pages I have worked on)? Where is the structure for goal completion (“Your page is PR1 on Google!”, “You’ve contributed to 100 pages – you’re now in the Century Club”, “You’ve earned blackbelt status!”). All of these things are woefully missing (okay you can earn your way into the ranks of “Wikipedia Adminstratorship” but that’s not the point for a general contributor). You don’t have to look too far to see a gaggle of content creation companies nipping at Wikipedia’s heels with a multi-modal strategy.

Take Mahalo, or Demand Media, or Associated Content, or the Examiner. All of these companies in their own way are all about collaborative content generation. In varying degrees, they get the multi-modal experience right. I spent 30 minutes on Mahalo a while ago and contributed 3 answers to questions, made about 0.70 cents, earned 25 points, got three “Good jobs from their community” and was told what the next level of achievement was. If I was going to spend more time contributing content to a website, guess if I’d spend it on Mahalo or on Wikipedia?

This is a subject I will write a lot more about in the next few months. We think it’s critical to our long-term success at Trada, and I think it’s critical to the success of crowdsourced businesses. To be fair, we haven’t incorporated all of these 7 motivations yet and some of the motivations we have are at the early stages of being integrated into the core experience. But you can bet we’re working on it!

Of course not all crowdsourced businesses will need, want, or benefit from all types of motivation. Many companies will argue quite justly that they can build a powerful long-term crowdsourced business without a dime of monetary motivation. Those who don’t recognize that humans are not one dimensionally-motivated though will likely see the similar early growth and then plateauing curve that Wikipedia is now experiencing. The good news for Wikipedia is that they can change this. I will look forward to seeing my name in the Leaderboard.

Filed Under: Crowdsourcing Tagged With: , , , , , ,
Newer Posts »