Marketing Type

Online Polling Creeps Toward Respectability

Polling, focus groups and the pseudo-science of customer research have been the bread and butter of many, many marketing companies for decades. As with other  industries, the Internet has chipped away at this market in various respects -- not least of which was establishing a culture of permission marketing that has made phone calls to your house during dinner from strangers a socially unacceptable practice (unless it's your local school district calling).

With an unnecessarily cutesy headline, the NYT today reports on the next assault on the consumer research racket in a story that focuses largely on the British company YouGov:

About Online Surveys, Traditional Pollsters Are:
C) Somewhat Disappointed

This is promising arena, I think. Traditional survey companies can charge upwards of $50 per respondent. Don't think for a second that equally sound (or unsound) results can't be had for much cheaper.

See also:

Polimetrix
SurveyMonkey
Amazon.com's Mechanical Turk
SurveyGizmo

 

More Mechanical Turk

Following up on a previous post about Amazon.com's investment in "artificial, artificial intelligence" -- which I skeptically titled "Mechanical Turk Is A Mechanical Dud" -- Jason Pontin of the NYT recently published a terrific piece on MTurk and the general lay of the land in the "crowdsourcing" arena:

Artificial Intelligence, With Help From the Humans

Excerpt:

Harnessing the collective wisdom of crowds isn’t new. It is employed by many of the “Web 2.0” social networks like Digg and Del.icio.us, which rely on human readers to select the most worthwhile items on the Web to read. But creating marketplaces of mercenary intelligences is genuinely novel.

By way of an update on my own experiment with the Turk: 

After the transfer of funds to my Amazon MTurk "Requester" account was FINALLY completed (a transaction that took the e-commerce giant over a week to complete), setting up a series of multiple choice survey questions proved to be quick. I offered $.09 for the completion of each of four questions, and $.10 for the completion of a fifth. Shooting for 1,000 responses to each I discovered that the initial pace of completions proved quite misleading. The longer the questions, or HITs (Human Intelligence Tasks), were up, the slower the responses arrived.

After a week or so, however, I did have 1,000 responses to each of the four $.09 questions. Strangely, the fifth (and most remunerative) question, STILL has not reached 1,000 responses 2 weeks later.

The other salient observations I can offer are these:

1) Despite MTurk's ability to allow individual respondent accounts access to each HIT only once (thus presumably preventing duplicate entries), responses submitted for one of my questions make it clear that  there are a fair number of individuals gaming the system by using multiple accounts.

2) The geographic breakdown of respondents to my survey ( self-reported) was as follows:

• 60% - US
• 23% - India
• 2.5% - UK
• 2.9% - Canada
• 11.6% - divided among 19 other countries  (the Philippines highest among them at 1.4%)

I'm not, by the way, discouraged from using the Mechanical Turk again, just a bit wiser about its best application. If you can use Amazon.com's Turk API to design a HIT that calls your own survey form, it does offer -- compared to most of the available methods -- an amazingly cheap way to gather impressions from a large, if unscientific, sample population.

Mechanical Turk Is A Mechanical Dud

Having fallen in love with its anachronistic moniker, it pains me to report that Amazon.com's "Mechanical Turk"-- a service intended to harness the vast collective processing power of individual human brains around the world --  turns out to be equally anachronistic in its technology.

The Mechanical Turk (Wikipedia entry) ostensibly allows anyone to post tasks that require human intelligence, as opposed to automated computer programs, to complete. An example might consist of an organization trying to analyze satellite photos of mountain ranges looking for evidence of a wreck. There are thousands and thousands of photos to process, but because it is difficult for computers to identify unpredictable patterns in images, it might make sense to upload the images into a series of HIT s (Human Intelligence Tasks) in Amazon's Mechanical Turk system, and to offer to pay $.05 or so per image to have a few thousand of your fellow humans glance over  a dozen photos each.  Similarly, if you work from home and have access to a decent Internet connection you might be willing to do a few dozen $.05 tasks in an hour of your spare time. 

In the same way that businesses and institutions can borrow the huge processing power of Amazon.com's servers for complex computer tasks through its Elastic Compute Cloud, individuals, businesses, and institutions are supposed to be able to borrow the Mechanical Turk for complex human tasks.

Why the Mechanical Turk is a Dud

Last week I decided to try out the Turk on a survey undertaken on behalf of a friend. Even though I have a pretty extensive account with Amazon.com already (seller, buyer, business credit, personal credit, reviewer, etc.), I extended my account to register with Amazon's Web Services and set about creating my first HIT. I needed 1,000 responses to a 5-question survey. To get 1,000 responses to each of the five questions, I planned to offer a generous $.10 per question.

No sooner had I created the first HIT, however, I was informed that in order to make the offer I had to have sufficient funds in my Requester account to fund it. Makes sense, right? To make the system work, requesters need to pre-pay. I expected to have to authorize a credit card, or PayPal [wait, that's owned by  a competitor...] transaction, or something quick. But here's the catch... I discovered that the ONLY way to fund a HIT request is through a bank transfer, which the help text explains takes "up to five days."

Wait a minute! This is Amazon.com, right? The pioneer of e-commerce -- the quickest, most sophisticated shopping site on the Internet. I can buy a $10,000 Plasma TV on Amazon.com in under 15 seconds. They already store my credit cards, bank account information, address, phone, and tax id -- probably more information about me than any other single business entity apart from one of the credit reporting agencies. But I still have to wait five days for a check to pass from my bank to them? For $500? That's really odd. An email to customer service (no phone number, no IM) yielded no explanation, just a repetition of the 5-day warning.

Seven Days Later

The best part is that here I am 7 days later, and the transfer is still pending. It's still pending even though my bank cleared the amount the day after the request. What that means to me is that some clerk in the Amazon accounting department hasn't gotten around yet to putting the little check next to my account that will finally authorize the transaction and allow me to complete the tasks I  NEEDED COMPLETED LAST WEEK.

Which is funny, because that, too, is a human intelligence task -- although it's one that most companies use computers for these days. I guess the folks at Amazon's Mechanical Turk have decided to practice what they preach.

Niche Topics: A Survey

I have been brainstorming this week in an attempt to figure out a few niche topics people care about that might be under-represented on the Internet. Rather than relying exclusively on my own very limited brainpower, I've decided to do what any good marketer should and launch a survey.

So please, friends, bloggers, countrymen, take my survey! [Click here]

If you finish the survey (which is mercifully brief), you will have a chance to win something you've always dreamed of... a bendable praying mantis replica.

Feel free to send a survey link to other blog-reading people you know.

Podcast of Cory Doctorow's lecture at Duke

Cryptonaut, of the RandomSignal podcast, has posted a recording (.MP3) of Cory Doctorow's recent address at Duke. Thanks, Jason!

(my previous entries on this topic are here and here)

A Reporter I Am Not

Well the battery on my laptop didn't last long enough to do any lucid reporting from Cory's lecture at Duke this afternoon, but I'm pretty sure writing and listening intelligently at the same time are to me what walking and chewing gum are to the protagonist of all the moron jokes. I found the material to be quite interesting and Cory is a bright, engaging, and provocative speaker who is also quick on his feet.  I found myself in agreement with him about 65% of the time (net neutrality, the idiocy of filtering Internet access in public libraries, the misguided nature of most regulatory attempts to address the uglinesses of the web), in disagreement with him about 20% of the time (absolute rejection of all attempts to establish reliable id systems, disdain for the crime-prevention capacity of public cameras, equating all governmental attempts to manage citizen information with Fascism), and confused perhaps 15% of the time as to what he was talking about (I think I mentioned that confusion is never far from me).

There were a number of bits of Cory's presentation that bear repetition, but without having taken notes I won't try to relate them just yet. Fortunately a former colleague, Cryptonaut of the Random Signal podcast, recorded the lecture and hopes to post the file soon. If so, I will return to the subject and offer a link and a few specific highlights for the curious.

Also worth noting that addressing Richard Lucic 's e-Commerce class again this year on the topic of Internet marketing was fun. Richard also teaches a class at Duke on new media in which non-technical students learn to cobble together podcasts, vlogs and the like. There's a great need for basic training in this area for sales and marketing folks in the private sector, many of whom have gotten the message that communicating with customers using new media and social networking is worthwhile but who just don't know where to start. It's great that his students are getting a chance to dabble in this stuff.

Privacy and Risk

So I am sitting at the Cory Doctorow presentation at Duke and, as is the fashion among bloggers (mostly bloggers of a more dedicated variety than I), I thought I'd jot down a few impressions as the event unfolds, live. For as long as my battery holds out at any rate.

First impression: Cory is probably around the same age as me, if not a bit younger, and looks much as his publicity photo depicts him. I had time to say a brief hello since we've never met in person, and he began almost immediately  talking over my head about a new configuration he's planning for his laptop/desktop synchronization. It's not hard to talk over my head of course.

Cory starts by reading, for comic effect, a typical IP disclosure clause from a contract that if taken literally would restrict one's very ability to think. He follows that with a quote from Mark Twain, which is an auspicious way to begin a lecture.

He's segued somehow into cognitive therapy -- which he seems to be in favor of -- but somehow I missed the transition. Now he's describing efficiency experts and the human desire for self-improvement.  I gather the connection is the application of information collection and analysis to the goal of improving one's life from an independent, rather than an institutional, standpoint. Now collaboration.... It all makes sense. He is, after all, first a science fiction author. Second an activist. As long as L. Ron Hubbard doesn't creep into the discussion, so far, so good.

Cory Doctorow at UNC and Duke

I have belatedly become aware that novelist, copyright reformist, and BoingBoing blogger Cory Doctorow is scheduled to speak on Thursday at both UNC and nearby Duke.

• Cory Doctorow at UNC, Chapel Hill campus, Thursday, February 22, 2007 ,2 p.m, Wilson Library [more from Paul Jones]
• From Myspace to Homeland Security: Privacy and the Totalitarian Urge , Duke University, Thursday, February 22, 2007, 5:00 p.m., Love Auditorium

Unfortunately I'm going to miss the first event, which looks to be fairly informal, because I'll be speaking on the subject of marketing at an ecommerce class at Duke the same afternoon. I hope to make the lecture at 5, however.

UselessAccount.com

The inimitable Jeremy Wagstaff, who writes the WSJ's Loose Wire column (from Jakarta, I think), offers an entertaining interviewette [subscription] with Jim Whimpey, (alleged) founder of a new service called UselessAccount.com:

Loose Wire: I see. So after the user signs up, what do they do next?

Mr. Whimpey: We live by the philosophy "less is more." Once logged in, users can edit their accounts, perhaps even occasionally forget their password. A huge focus on editing your account is what sets us apart.

Loose Wire: I see. But in terms of the actual service? What is it?

Mr. Whimpey: Er, it's an Account Creating and Editing Experience.

I think I may have signed up for this at some point, but I'm not sure...