Remember when e-commerce used to be about the technology? Never mind profitability ratios or shakeouts. In the beginning, it was all about the high-tech novelty of shopping online, the excitement of the “never before.” Isn’t that why we started paying attention in the first place?
However, the first wave of e-commerce companies found out the hard way that no matter how new or different their tech advances were, it was the quality of the online shopping experience that counted most, not how high- or low-tech it was.
That is still the case today. Take a look at what’s new in the e-commerce technology realm and it is clear that customer experience is king.
“The most significant innovations from a technological standpoint fall into two areas, one focusing on improving a site’s experience itself, and the other the actual customer experience,” Elaine Rubin, chairman of Shop.org, the online arm of the National Retail Federation, told the E-Commerce Times.
When it comes to improving site experience, nothing seems to generate as biga buzz in the industry as these three words: natural language search.
No matter how incredible your product, or how great your online customerservice, customers are far more likely to purchase from you if they are ablefind exactly what they’re looking for.
According to a recent Forrester Research study, 76 percent of tested Web sitesearch engines failed to show all relevant search results. Luckily fore-tailers, “archaic” Boolean logic search engines are on the way out.
“Natural language searches don’t require consumers to change the way they would ask questions in an offline retail store,” said Rubin. “They allow for less specific search parameters and a much freer type of search experience — with better results.”
Instead of conducting a key word search on an apparel site for “sweater,” for example, a natural language search lets consumers enter more precise and more conversational language into the search engine, such as, “Show me all your small angora turtleneck sweaters.”
Users can also ask follow-up questions torefine the first one, such as, “Which ones come in blue?”
IBM’s Conversational Services division, SpeechWorks International and others have been busy in the natural language market for some time.
Startup company EasyAsk claims that its engine achieves precise matches over 90 percent of the time. If accurate, those numbers have the potential to dramatically increase a Web retailer’s customer retention rate, as wellas its conversion rate.
EasyAsk says it enables e-tailers to learn from the searches that are conducted by its users. It gives companies the ability to conduct detailed analysis of search queries and results, as well as personalize the results and prices for each customer.
According to Rubin, natural language engines are already being employed by severalmajor e-tailers, mostly in the apparel and soft goods industries.
The biggest irony in the evolution of e-commerce technology may be the increasing efforts by e-tailers to recreate the offline shopping experience as closely as possible.
However, even sites like Amazon, that excel at creating a genuine customer experience, have yet to leverage one of the biggest advantages of offline channels: direct face-to-face contact with a salesperson.
If a new company called Finali has anything to do with it, that may be about to change.
Finali creates personal assistant characters, called netSages, that whileappearing live to site customers, are actually automated videos of realpeople offering help with virtually any aspect of the shopping experience, including policies, product availability and technical support.
Each video response by a netSage is prompted by a user’s selection from a series of help links, so customers feel as though they are receiving directfeedback from a real sales assistant. In addition, users can switch to alive online specialist at anytime during the process.
The netSages can be customized for each e-tailer, and they also collectdemographic data from the customer, with a success rate that Finali claimsis three times better than using standard request forms.
NativeMinds is another mover in the virtual interaction arena. The San Francisco-based company, which counts the Ford Motor Company and Oracle among its clients, builds and maintains automated online customer service and support agents called vReps.
Using the company’s flagship product, NeuroServer, vReps answer customer questions via two-way, conversational dialog.
According to Rubin, virtual representatives are the result of many years of psychological research on how people interact with computers and technology.
“It’s very smart,” said Rubin. “[Finali] realizes broadband isn’t hereyet and what will get people to buy more can all be treated through thisautomated sage. We’ll see a lot more of this kind of social interface.”
Dell, BMW, and Amazon are some of the major e-tailers already on Finali’s client list.
Could you ballpark the cost of the service , say, in a limited number of applications to allow small business web sites to know whether the service is within reach?
Agree that both databases are limiting, and that it’s the multiple product shoppers which are proving to be hardest to tackle. However, it is interesting that the new image-based search technology companies seem to have been overlooked. These technologies should not only make the shopping experience faster and more efficient, but should overcome database limitations AND encourage the impulse buy because users can compare VISUALLY the products which they are looking for. Language — either natural or boolean — will no longer be an issue, and that includes foreign language shoppers. I have only seen a few of these technologies work, but so far they are quite promising. (Lookthatup, eVision, Visfinity) (yes, I AM in the image technology field, but I don’t know enough to know if they will help the shopping experience). Anyone???
Natural search language has been tried before. And it doesn’t work better because it did not solve the real problems. The problem is not in the interface but in the database information. A search engine works well if and only if it has access to a well-referenced database with information pertaining to the users. To do so, you need: A huge relational database kept up-to-date by huge reference staff and a lot of willing customers that give you their opinions. That is not an easy task. For the rest of the features you talk about, in a year’s time you will understand that they do not bring anything useful.
Continue to search…because they do not serve the user experience and they will not increase the ROI of any e-tailer.
Let me give you a piece of insight — The average retailer makes over 80% of their profit not from the
guy who goes in and buys the one item he is looking for, but rather from the guy who buys
the shirt, a tie, and a great-looking pair of trousers that he really wasn’t even looking for. That’s
what e-commerce needs to strive for. Natural language only improves the typing experience, not
the shopping experience. When you can get the Internet to duplicate the ability to browse, not
search, the JCPenney catalog under 1 minute, then and only then will the Internet be able to
beat its current rival, the catalog industry. Because to compare the Internet with an in-store
experience, it’s a no-contest situation.