Target piloting Amazon Alexa voice search rival

Oct 07, 2016

Dan O’Shea

Through a special arrangement, what follows is a summary of an article from Retail Dive, an e-newsletter and website providing a 60-second bird’s eye view of the latest retail news and trends.

Target and three other retailers have partnered with natural language processing (NLP) startup AddStructure to roll out what an AddStructure co-founder describes as a “white-label” spin on Amazon’s Alexa virtual assistant, according to a Chicago Tribune report.

Chosen in June to participate in the Target + TechStars accelerator program, AddStructure has already worked with Target on a six-to-nine-month pilot program and is set to begin similar pilots next month with L’Oreal, Under Armour and the online antiques marketplace, 1stdibs.

“We’ve always been focused on language, so how are people searching for things, what are the priorities around different products through analysis of user reviews,” AddStructure co-founder Will Underwood told the Trib. “Now it’s just about using that knowledge of natural language to allow people to interact.”

During the accelerator program, AddStructure got to work with a vice president as a mentor and collaborated with three departments within Target to learn more about integrating new technology into a retailer and perfecting its product.

When combined with machine learning and artificial intelligence, NLP allows chatbots, virtual assistants and voice-activated search engines to interpret and understand customers’ questions and needs, helping them find the products they’re looking for as well as other items that might interest them.

Some aspects of NLP are in Amazon’s Alexa virtual assistant (which powers its voice-controlled Echo speaker), or at least in the systems backing up Alexa. In addition, office supplies retailer Staples is testing a souped-up version of its Easy button that leverages NLP and related schemes, and last December, The North Face said it was tapping Watson, IBM’s NLP platform, to help online browsers use natural conversation to receive outerwear recommendations tailored to their needs.

On Sept. 16, Etsy said it acquired Blackbird Technologies, which uses NLP to understand complex search queries as well as machine learning that analyzes user behavior and unstructured data to suggest relevant and personalized search recommendations.

“We believe we can enhance the buyer experience by making search quicker and easier and by surfacing even more relevant, tailored product recommendations,” said Chad Dickerson, Etsy’s CEO, in a statement.

DISCUSSION QUESTIONS: Will natural language processing combined with machine learning take the online shopping experience to the next level? Do you see applications for voice search technologies at the store level as well?

Please practice The RetailWire Golden Rule when submitting your comments.
"NLP in both voice and text are the logical evolution of human-machine interaction."
"People don’t want to search, they want answers."
"NLP will unquestionably become a key piece of eCommerce online as well as in store, as more digital assistants become viable."

Join the Discussion!

11 Comments on "Target piloting Amazon Alexa voice search rival"

Sort by:   newest | oldest | most voted
Cathy Hotka

NLP is in its infancy, but the results are already pretty impressive. I saw a demonstration of IBM’s Watson a few months ago and it’s stunning. The question going forward for retailers is whether customers will get used to using it and, if so, how retailers who don’t have it will survive, since younger customers are devourers of all things tech.

Sterling Hawkins

NLP is in a long list of technologies that are poised to revolutionize online shopping. IoT (think Amazon Dash buttons and smart reordering), VR and 3D printing are all in the running with countless R&D and VC dollars going towards advancing the technologies. NLP will unquestionably become a key piece of eCommerce online as well as in store and more digital assistants become viable.

Ken Lonyai

I have been working in the NLP, artificial assistant, and bot arena for the past five years and there’s a reason why I say tired old click-and-touch interfaces will be replaced. NLP in both voice and text are the logical evolution of human-machine interaction. Click-and-touch will be around for a long time yet, but conversational interfaces will grow and in a few years become dominant. Well-built with powerful AI, these technologies make interactions more useful and convenient for consumers and more effective for businesses. The video example is a bit cumbersome and most automotive voice activated systems have a ways to go, but well-executed conversational interfaces are a reasonably enjoyable and efficient paradigm.

You don’t have to take my word for it though, look at what Google just put a huge investment into.

Keith Anderson

People don’t want to search, they want answers.

There’s a race among technology companies to provide them and if retailers don’t own or influence the platforms that deliver the answers, they stand to be seriously disrupted.

Max Goldberg

NLP has a long way to go before it accurately processes consumer requests and delivers what the consumer wants 99.9 percent of the time. That said, computing is moving towards an NLP world. Apple just build Siri into OS Sierra. Google just released NLP-driven devices. The key to successfully launching NLP at the retail level is accuracy. Nothing is more frustrating than having to query a machine two or three times with the same question.

Shawn Harris

I typically use a term called “frictional cost” as a measure of what shoppers are willing to tolerate on the path to conversion. The more intuitive and easy a retailer can make that path the better, right? Natural language processing combined with machine learning can help in reducing frictional cost. Help me find what I am looking for faster, while serendipitously offering up other finds. One area of opportunity in-store for NLP would be voice-based wayfinding (“Hey Target, do you have Alpo Healthy Puppy in THIS store?”) for both customer self-service and associates’ productivity.

Ken Cassar
Ken Cassar
Principal, Cassarco Strategy & Analytic Consultants
4 years 15 days ago

I am bullish on the power of voice-controlled user interfaces, particularly in the home and in the car, where typing requests can be difficult or impossible. Imagine if, while driving, you could easily ask a device to find “‘the Pampers closest to me’ or ‘the cheapest 12-pack of Sierra Nevada beer in zip code XYZ.'” Of course, any of these initiatives are going to have to be at least as good Amazon, which has a natural advantage due to its dominance of the e-commerce space, and Google with its virtual ownership of search. The biggest question is whether retailers like Target might be better served throwing in with Google to counter Amazon — a clear competitor — or getting into one of the most intense tech arms races on the planet.

Lyle Bunn (Ph.D. Hon)

NLP is one of the more exciting areas of our X + Y = Z world, in particular as our amount of digital information available is now unfathomable and exponentially growing. Product information, planograms, databases, operating applications and commentary offer the raw information for questions to be answered. The question of Big Data is simply “what do you want to know right now?”. Getting this answered has been a black hole of investment, so the real science of NLP is curating response elements so that a correct decision can be made. Baby steps are leading to strides.

Ben Ball

Solo NLP initiatives will be limited in their usefulness. They may well follow the path of solo payment apps — inconvenience devolving into irrelevance. Retailers will initially be better off to maximize their position with “The NLP Trio” (Alexa, Siri and the newcomer Google Now, which needs a better name). Consumers will value breadth of access from a single app they are comfortable with and won’t want to have to adapt to a dozen different retailers’ apps anymore than they wanted to learn to use a dozen different mobile payment apps. This will eventually change, however, when “voice-activated” becomes generic. It will become the way we interact with practically everything in the IoT.

Personally, I’m really looking forward to being able to ask my Samsung refrigerator to pour me a glass of Far Niente or a Glenlivet on the rocks.

Adrian Weidmann

NLP is an exciting frontier on the way to defining what it is to be truly digital. This has tremendous potential for replacing traditional analog keyboard interfaces. We live in a nanosecond world and shoppers want immediate answers and the gratification that comes with instant responses. Speed and accuracy will win the day. Brands that curate the voice libraries to address the widest range of questions will be the next winners in the Voice Engine Optimization (VEO) game. I haven’t experienced IBM’s Watson on this but I suspect it’s extremely impressive — right on the verge of spooky.

Larry Negrich

The proof is always in the execution. This has great potential to be a useful technology for both online and in-store shopping and I hope it will be applied to a number of existing issues for shoppers. For instance, the extremely simple task of in-store product location, which I thought would have been solved long ago, continues to be an issue. Kiosks, apps have been assigned to handle this but in the end store layout variability, stock location, product names continue to impede efforts to solve this simple task. So, NLP has great potential to be helpful, but retailers would be wise to have the right (boring, old school) systems in place to give NLP a chance to succeed.

"NLP in both voice and text are the logical evolution of human-machine interaction."
"People don’t want to search, they want answers."
"NLP will unquestionably become a key piece of eCommerce online as well as in store, as more digital assistants become viable."

Take Our Instant Poll

What’s the likelihood that natural language processing will be integrated into the online shopping experience of most major retailers over the next five years?

View Results

Loading ... Loading ...