YCRFS 2: New Paths Through Product Space

My first thought was “doesn’t Amazon already organize things independent of vendor?” But what we’re really looking for here is a recommendation engine, and Amazon really isn’t very good at that. As I see it there are three main steps which should be done in order. The more you know about ahead of time, the further you can skip ahead. Here are the steps:

Step 1. I don’t know what I’m looking for.

Maybe I’m looking for a good science fiction book. Or a gift for a 14-year-old niece. A deep gift recommendation engine alone could probably make a good startup. The engine should ask some probing questions and give recommendations—either specific products or product categories, depending on the type of question. This experience mimic the experience of having an experienced salesperson ask questions and point you towards good options. A beautiful browsing option would be a great way to conclude this step.

Step 2. I know what I want, but I haven’t decided which brand or model to get.

Here we need a deep comparison. The ability to rank priorities would be really helpful. Let’s say I’m buying a laptop. The site should ask me how much I care about weight, screen size, processing power, etc. It can then recommend models that fit within my budget. There’s a huge amount of domain-specific knowledge required to get this right for a variety of products.

Step 3. I know what product I want.

Here we need to give comparisons among vendors. Price, shipping speed, return policies, and other general customer feedback are important. This is where it helps to be a trusted third party.

This problem is a huge undertaking, which is why no one has come close to solving it yet. Frankly, unless the domain is very restricted, I don’t think a three person team will be able to do enough specific programming and testing to make it work.

The principles and mechanics would best be discovered in a very narrow marketplace, and one full of confusion. Electronics would be the natural stating place, but a less crowded market might also work.

Do you really find this to be true?

1) Go to http://www.newegg.com -- while brand/vendor is one selection criteria, so are many of the technical specs and options. They even offer advanced and power search.

2) Amazon is no different: While Brands are promoted (because consumers are trained to that) they offer power on the left side here (below brand): http://www.amazon.com/Digital-Cameras-Photo/b/ref=sv_p_4?ie=...

While I don't think that the statement in the RFS is accurate, there is no doubt in my mind that the online product finding experience has dramatic room for improvement and innovation.

One of the challenges of abstracting this layer is that who sells the product still adds or subtracts value from the purchasing experience because of those mundane details (shipping speed, reliability, returns, etc.).

There are various ways of dealing with this. Amazon [1] is already active in this space using what you might call the "user review" model combined with enforcing some exacting standards on their vendors. Already having a huge channel of course made it easier for Amazon to get vendors to accept this.

[1] Incidentally, I've never thought that Amazon did a great job with sorting products within its general categories, so I always search instead. But I do wonder, how many more people would window-shop at Amazon if their sorted product catalog was more interesting?

Shopping is often a pass time and a social experience, not a mission, yet most sites are optimized for the last "stage" of shopping--the stage where you actually make a purchase, or the penultimate stage--the one where you compare a few specific products. The last stages are the most profitable and easiest to implement, so that's why they came about first. But if you can grab a consumer at an earlier stage, when they're brainstorming bridal shower ideas, or when they're bored at work and checking out new cell phones, and if you optimize for discovery as well as sales, they'll come back and buy from you.

The challenges/pitfalls/competitors are numerous, obvious, and non-obvious. Good luck to the team that attempts this!

http://en.wikipedia.org/wiki/Affinity_analysis

That would make this an ideal project for people with machine learning/data mining experience, particularly from a retail/e-commerce background (Amazon, eBay, Wallmart.com or even a brick and mortar store chain such as Safeway).

Let's say you have a problem or a need or a want, and you're sure there is product somewhere in the marketplace that would help. You Google, and a handful of web sites pop up that offer promising solutions. But obviously you don't believe anything you read from vendors, so you check for online reviews. Then you wonder if the favorable reviews are planted by the vendor, and the bad reviews are planted by his competitors. Can you trust reviews from anonymous strangers?

You ask your coworkers and friends if they have ever used the product that you're interested in, and no one has. What now?

Shopping is broken.

How much more stimulated would the economy be if the people who have money, and are willing to spend it, could be reliably connected with the products that they desire?

What the world really needs is some way to connect you with the people who already use the sorts of products you want, and are willing to answer an e-mail or two about the topic.

About a year ago I had surgery to fix my voice. The information on the Internet about that particular surgery was outdated and didn't address my questions. The only way I could become a consumer of that surgery was by communicating directly with people who already had it, which I did. And since then I have answered questions for dozens of people who have the same questions that I had.

Likewise, as my wife and I make a zillion decisions for the home we are building, we prefer products and solutions used by people that we have spoken to personally. The Internet is virtually useless for any of the hundreds of product decisions we have made so far.

And what about choosing a destination for a vacation? You're much more comfortable if you have spoken to someone who visited the same place.

The obvious problem with connecting past consumers with potential consumers is that while people are generally helpful by nature, no one wants a million e-mails asking how they like their new can opener. So how do you strike the right balance?

Imagine a system that works like this: When you buy a product, you agree in advance to answer up to four e-mails from future potential customers, beginning no sooner than one year from when you make your purchase. It's totally optional, but agreeing gives you access to people who already bought the product you're considering today, to help you make your own decision. It would strike you as a fair deal.

For privacy reasons, this imagined system would disguise your e-mail address. And the system would have to be administered by some third party, not the vendor selling the product, or you wouldn't trust the strangers giving you advice.

Maybe you have a better idea for fixing shopping."

From: http://www.dilbert.com/blog/entry/shopping_is_broken/

Internet tends to be an open system where consumers are in the same way important as the product makers and sellers. Success is measured by community support and it is assured by complete transparency, trust.

Marketing your online products also cannot be done well without a strong community. A future webshop must be completely user and community driven, your duty as a webshop owner is to build trust and make your existing customers to connect to your next customers.

A Classic Store vs. Auction Based Stores

Following years will be about a 180 grades shift in e-commerce. Now you are presenting products to customers as a classic store, later customers will ask you for products and services.

Yabe.hu -- Race. For you -- is a new approach to shopping. As a customer you log in and tell your desires then vendors are racing to offer you the best solution. This is the only way to assure yourself you'll get the best deal.

You don't have to choose, listen to commercials, browse social shopping sites and do research to find what you want; just spit out a desire and let the others do the job for you.

Until this approach to shopping hits mainstream you'll have to do the classic way. Look for a shop who makes you feel comfortable, do the shopping, show your items bought and generate trust for other users and would-be-customers. Wait for your cut offered by the webshop as an affiliate program.

A Classic Webshop In Modern Clothes

   1. Browse to your favorite online (gadget) store.
   2. You are sure the frontpage is already taylored to your shopping needs (via an Automated Recommendation Engine)
   3. Select your products to be bought and check for the best price gathered from other (gadget) stores.
   4. Make shopping and later add comments, recommendations, how-to's, FAQs to your products bought.
   5. Other users will see your value added to existing products and when buying you'll get a fair cut from every sales.
   6. Too boost your revenues use widgets with your products bought on your blog or favorite social network, do the marketing and sales for the webshop you like and which pays back your efforts.

The companies that are best at these things are best precisely because they don't consider them to be "mundane".

In customer service and fulfillment, it's a fine line between being excellent and being out of business. These can be very difficult things to do, and the thinner your margins, the better you have to be.

And very few companies handle customer returns well. They keep getting better and better until they realize the best way to handle customer returns is to avoid them in the first place.

Also, I have a hard time imagining how big stores could choose to use a presentation system similar to that of other stores. Unless I misunderstood the original idea, it seems that selling presentation software is not the right way to address that problem.

Having a good understanding of co-op advertising could matter a great deal here. Retailers get paid by manufacturers for product placements... Someone going after this model could try to reach similar deals with both the manufacturers and the retail-fulfillment entities.

http://www.etsy.com/