Adams is describing how marketing (tries to) work. Marketing seeks to identify customers.
In a manner of consideration, Adams is seeking to acquire better and more targeted search results, or skills in searching for products or services. Google Adwords tries to offers something close. But the Googlers are not very good at this level of targeting. (Yet?)
As for specifically contacting buyers, that's expensive for commodity gear. And what happens when you find and proffer a product to a customer, and the customer turns around and purchases that product from the lowest bidder; if you can't automate this and keep the customer acquisition costs (very) low, a business ends up a variant of the Brick-and-Mortar cost differential.
The middle ground - some form of a trusted concierge or consulting service or search service - seems to be one of the few potentially approaches from both directions; back to what amounts to a (trusted) middleman in the purchasing process.
Google AdWords, if managed properly, does a very good job at this. Take his example: For example, let's say you're looking for new patio furniture. The words you might use to describe your needs would be useless for Google. You might say, for example, "I want something that goes with a Mediterranean home. It will be sitting on stained concrete that is sort of amber colored. It needs to be easy to clean because the birds will be all over it. And I'm on a budget.
So say I am retailer of stylish, affordable outdoor furniture, among many other products (that is what is important in the query, more on the Mediterranean part later). I might have an ad campaign dedicated to patio furniture. Within that campaign I might have an ad group for plastic furniture (easy to clean).
That ad group would include a mix of exact and broad match keywords/keyword phrases like patio furniture, budget, cheap, affordable, stylish, along with negative keywords like expensive, wood, etc. The ads would deliver him to a landing page with links to hopefully appropriate furniture. On that page he can then see the various items that meet his basic qualifications - stylish, affordable, easy to clean - and click through to the items that look like they would match his Mediterranean home, with its amber colored concrete patio.
Reading through what I just wrote: I don't know much about the furniture business, but Mediterranean might actually be an important and useful kw to advertise on...
But Google is the concierge you're talking about. It has the description of the product on its page, and it has the description of what the shopper wants (your search terms). The algorithms just aren't as smart as a human being, and you rightly pointed that a (smart) human being is expesive.
The key problem is one of cost. As it is right now the customer pays the costs of specifically linking up with a supplier for a transaction. The supplier pays for some of the costs of the customer's search via advertising, but generally the supplier only pays for one-to-many interactions, while the customer pays for the one-to-one interactions.
The thing is that the customer can afford a few small one-to-one transactions. It's not hard for them. However, the supplier likely wouldn't be able to afford the costs. Lots of one-to-one transactions don't scale.
The only problem with broadcasting my info to a vendor for some patio furniture is that next week they are going to call me about some nice tiki torches. Then, I'm going to get some snail mail from their buddy the BBQ salesman, and his buddy the meat vendor will come a knockin' too.
Then, one of those guys is going to sell my information to totally unrelated businesses, so I'll start get calls from people trying to sell me auto warranties. I'll get endless catalogs from all their combined efforts, and I'll never be able to read email again, because there will be too much spam in it.
Finally, I will end up subscribed to some porno websites I never heard of, and since I don't want to explain that to my girlfriend, I guess I'll just keep using the canvas folding chairs that I have.
But if the only way they could contact you was through the specific "want ad" that you posted, wouldn't that help some? You're not asking them to come to you, you're asking them to reply to that ad. They shouldn't need to know anything else about you except what you posted in the ad. This assumes that the Broadcast Shopping company that is acting as the mediator between you and the sellers isn't going to sell your contact information, but that should be made very clear, "Hey, the only way a seller can contact you is through this one ad, and you can turn the ad off as soon as you get the offer you're looking for".
What if that related or unrelated business that wants to approach you based on your ad for patio furniture, have to pay for that privilege. That might definitely make those auto warranty guys have second thoughts.
The first challenge I can think of for someone creating a service like this is spam. If I put out a request for patio furniture, how does your service prevent all sorts of other vendors from replying to my request and trying to sell me something other than patio furniture?
So users want a list of items, without massive duplication (many shops selling the same toaster), while at the same time knowing the best price for each item. Seems Amazon have this nailed with their multiple retailer options.
Maybe I'm being cynical, but for the site he describes you'd get a whole ecosystem of "recommendation experts" or similar who will auto-spam replies on ToasterCo's behalf, and you'd also end up with standard keywords emerging, once people start to find words that correlate well with good replies.
It would probably work best in areas where companies sell their products direct, and are small enough to give personalised replies. Large corporations would surely give their sales people a set of model answers and you would end up with replies indistinguishable from spam.
So you institute a ranking algorithm that also watches for spam to insure that all of a particular vendor's replies aren't either the same or just filling in a template with keywords from a request and otherwise the same. That way, if a vendor makes too many unpersonalized offers, they could get warned, lose karma, whatever. Whereas if a vendor really targets someone's needs well, even if they don't have the price or exact item that the buyer wants, that buyer might still shoot them some good karma for trying.
In the enterprise world, this is known as "lead generation", and these queries for specific items would be "qualified leads" (meaning the buyer has specifically expressed interest in the type of product you're selling). It is a very profitable segment of the industry. That said, there are an awful lot of problems bringing the model to consumer goods.
The instant gratification problem that naval pointed out in the comments...Most folks have a buying process something like: Want, shop around for a few minutes to a few hours, buy, have it within a week. Sometimes it includes going to see it in person at a brick and mortar. Waiting for a bunch of people (most of whom will be spammers) to submit proposals, is tedious and boring...like being at work. A lot of enterprise product buyers even try to avoid that process. I have a Dell small business sales rep, who usually gets me better prices than the website, but I rarely go through him because the process is tedious. It wastes more of my time than the money I save. I imagine if I were buying dozens or hundreds of machines, it would begin to be worth it.
The spam problem is always underestimated by people who don't deal with it professionally. By the time Scott Adams sees the spam problem in his own life, it has already made it through several anti-spam preventative systems. This is a new way to spam, and will require new ways to fight it. I suspect one of the lead generation models would have to be used...as dpatru suggested, vendors paying a small amount to be displayed to the user would be a good path. A vendor would probably happily pay a few pennies, or even dollars, to reach someone that they honestly believe would be a good fit for their product. In the enterprise world, you only need a few vendors, a few buyers, and very expensive products, to make the model profitable. In the consumer space, you need all the vendors to make consumers really happy, and an awful lot of consumers to make it worthwhile for vendors to monitor and respond to leads.
It's an interesting take on how vendors in this highly distributed world can find out what people really want. Chinese manufacturers make almost everything we buy these days...but what items people need and want, fashion, trends, perception of quality and beauty, are different across cultures.
But, I don't think it will win out over automated recommendation and search engines. Automated tools will get better (and have gotten dramatically better in the past ten years) at helping folks find the right items. Amazon is already pretty darned good at it...I usually know which of a handful of items in a genre I want within a few minutes of beginning my search. At worst I read a few reviews, and then know. The Amazon model has the benefit of having other customers telling me about the products...I have a deep mistrust of companies telling me about their products. Some are honest and tell you exactly what you need to know, but most are hyperbolic to the point of being nonsensical.
In other words, this is one of those ideas that sounds neat, and is a really interesting thought exercise, but I really doubt I would ever use such a service, as a buyer or a seller.
Well - the idea seems sound. It's the same concept that makes stackoverflow, serverfault and others work. If there's a problem, it would be the target audience / moderation... Is there a stackoverflow clone for internet shopping already online?
The point about too many choices is just not true. I have a few online stores I purchase from. And eBay. Sometimes Amazon, but me living in Canada means that it's mostly restricted to books.
The average online store is so badly put together that it is the equivalent of shopping in a dirty dark alley in terms of confidence :) So, I don't. I find I keep going back to the same stores over and over. eBay wins because there is a buyer protection scheme in place.
The the point about thousands of stores to purchase from only holds for items costing less than I'd care about losing.
The problem of too many product offerings (the paradox of choice / analysis paralysis) is something I'm trying to address with my startup Kallow.
The issue with Adams' solution (which may work for some) is that customers sometimes don't know exactly what they want, and have a hard time describing it, especially when it comes to things like personal electronics. You are essentially asking them to design a product for themselves, which sounds kind of cool, but will leave a lot of people confused.
customers sometimes don't know exactly what they want, and have a hard time describing it,
I don't want to minimise the problem quoted, but that is not the entirety of it. There is still a problem for people who do know what they want, due to the one thing that is hard to quantify: whether the device performs its function well.
The toaster example is particularly apt. I can start off knowing precisely what I want: Say, I want a white 4-slice toaster with wide (bagel) slots and a defrost function for under $150. Ta da! If that is all there was to it, I could be finished in 3 minutes - it's not actually that hard to find one (though not with Google - you need a real e-commerce site like Amazon).
The problem is that when you do find a few toasters that match your requirements on Amazon, most have reviews like "stopped working after 2 months" "completely uneven toasting" "doesn't pop toast" etc etc. So then you have to wade through review upon review on model upon model in order to find the least sucky toaster. That is when the toaster hunt becomes a time sink.
Now there are sites like consumer reports who are supposed to give you a recommendation based on some actual testing, but of course they are too slow moving - whatever model of toaster they recommended 18 months ago is no longer to be found.
I respect Adams' opinion generally, but I don't understand why he thinks the buyer would get the sorts of responses the buyer wants. Why wouldn't the buyer just get spam (or, at best, directed advertising of the sort that Google already has)?
If advertisers were able to respond well to the important keywords in his description "easy to clean", "Mediterranean", "amber", "budget", they'd already be doing so with Google ads.
Do people really have trouble finding the things they want to buy online? Whenever I do a search for information on any kind of product, I get a ton of ads offering to sell it to me. This seems like a solution in search of a problem.
I think the point was that there are too many options:
Now you shop on the Internet, and you can buy from anywhere on the planet.
The options for any particular purchase approach infinity, or so it seems.
So while you may get tons of ads offering to sell you something, how much effort is spent determining which is the best match for what you want. It would be better to get 5 results that are really close to what you're looking for than 500 that are sort of what you want.
It's just not going to happen. My approach would be to start with the expensive ad-hoc shopping services that already work for rich people and figure out how to standardize, commoditize and mass market them.
Any idea that has an element of "replace humans with ai" to it is ultimately in direct competition with Google. Which means both that your service will not be more useful than Google and that they will crush you whenever they want to.
The path to value, for everyone other than Google and Amazon, is to not replace humans but instead make them more powerful.
I expect Hunch could easily grow this way: a mixture of introductory questions to assess your purchase interests and priorities, leading to an increasing proportion of sponsored 'questions' that are offers/pitches tuned to your revealed preferences.
In a manner of consideration, Adams is seeking to acquire better and more targeted search results, or skills in searching for products or services. Google Adwords tries to offers something close. But the Googlers are not very good at this level of targeting. (Yet?)
As for specifically contacting buyers, that's expensive for commodity gear. And what happens when you find and proffer a product to a customer, and the customer turns around and purchases that product from the lowest bidder; if you can't automate this and keep the customer acquisition costs (very) low, a business ends up a variant of the Brick-and-Mortar cost differential.
The middle ground - some form of a trusted concierge or consulting service or search service - seems to be one of the few potentially approaches from both directions; back to what amounts to a (trusted) middleman in the purchasing process.
And a concierge service has inherent costs.