There’s no denying Amazon.com leads the way with artificial intelligence (AI) in retail and e-commerce. The company’s advanced flywheel model rethinks how AI improves digital customer experiences (CX) and has ingrained the very idea of AI into how Amazon lives and breathes.
Retailers may be intimidated looking at Amazon’s suite of new AI-powered products and services. Rather than feeling outperformed, however, retailers should get excited. They may not be able to replicate the exact solutions Amazon (and other industry leaders like Microsoft and Google) is bringing to market, but there’s ample opportunity to learn from these advancements and mirror their intent at scale.
Amazon Points Other Retailers in the Right Direction
For the attuned retailer, Amazon’s new products are highly indicative of a rising trend in cloud services and user experiences designed for a post-text internet. Web engagements have historically focused on reading and writing, but today’s internet is evolving to preference vocal communications as well as visual mediums beyond text like pictures and GIFs.
Let’s break down that environment in terms of e-commerce. Not long ago the online purchase journey occurred exclusively on a retailer's website. However, the way consumers approach e-commerce has shifted dramatically in recent years. Shoppers are browsing on the go, making purchases through social media, and blending in-store and online capabilities. They’re popularizing new channels and devices, elevating e-commerce to include more than just screens.
The increasingly post-text digital world raises more questions about CX best practices than it answers. Reviewing the features Amazon announced in its most recent earnings report, however, can guide retailers immediate investment plans and research. These products include:
- Amazon Transcribe: converts text to speech
- Amazon Translate: translates text between languages
- Amazon Comprehend: explores text for insights and relationships
- Amazon Recognition Video: analyzes visual and video content
When parsed into their specific capabilities, Amazon’s new services highlight two key trends: the rise of natural language processing (NLP) and computer vision.
NLP in a Post-Text World
Consumers will make 50 percent of all searches via voice in 2020, meaning that retailers must pivot to embrace this new mode of search. The channel isn’t 100 percent viable for most customer journeys yet, and voice commerce is still an outlier. However, voice search is here to stay.
To get ahead of this trend, retailers must eliminate friction from search wherever possible. For example, NLP is a worthy investment because it ensures new AI-powered devices listen and communicate like humans, leading to fewer errors as well as solutions that understand what users are asking the first time.
This is crucial because consumers are choosing voice search where it drives convenience. Consumers no longer have the patience for typing requests on tiny keyboards. Rather, they’re turning to voice commands to search online. In fact, nearly a quarter (21 percent) of people using mobile devices for voice search are doing so just to avoid typing.
As a result, many of consumers initial interactions with retailers happen via voice. Devices such as Amazon Alexa and Siri make it easy and pain free to purchase low-risk, routine items via voice. As Amazon demonstrates, making the best first impression demands strong NLP capabilities. Of course, the retail giant’s new features have broader aims to address the multilanguage and geographical diversity of its user base, but the bottom line is that convenience remains king.
Looking forward, retailers should consider solutions that better enable effective voice interactions in real-world scenarios. For example, far-field speech recognition is an important addition to any voice-powered solution, as it allows the device to separate background noise from actual user requests. It’s also important to begin investing in updated search engine optimization (SEO) strategies that account for the rise of longer-tail keyword searches and question words.
For instance, while a browser might only be willing to type out “pizza” on their smartphone, the same user searching via voice may ask “What are the best gluten-free Chicago pizza restaurants open Sunday after 5?” Voice better accommodates details users would otherwise uncover though additional link clicking.
Computer Vision in a Post-Text World
Second to voice-enabled search is image-enabled search. Retailers are still in the early stages of developing purchasing possibilities and goals, but image search is already transforming the digital experience. Doesn’t it sound nice to take a picture on your smartphone, upload it to Google, and then discover similar products or experiences?
Mastering image search requires retailers to follow Amazon’s example and explore computer vision — i.e., the process through which machines analyze the visual world to spot trends and make decisions. On a scaled level that the average retailer can more realistically tackle, this means thinking through how AI-powered machines will take on image search, as well as the image-based search actions consumers will be willing to complete.
Remember, people are using these new modes of search with a single goal in mind — optimal convenience. For example, image search could provide major value around technical purchases. Currently, if a consumer needs to replace a part in her dishwasher, for instance, she must pull out tools, identify the piece, find it in the company’s product catalog, and then place an order. Why jump through all these hoops when instead she could simply snap a quick picture and upload the image to Google?
Companies like Pinterest already support image-enabled search options and prove that such experiences are possible. Popular apps like Pinterest will help speed up consumer adoption, and will also raise the expectations for effective, diverse search options.
What does this mean for retailers? They must have an extensive product catalog with high-resolution, image-rich content that can be dynamically and automatically tagged, organized and resurfaced for easy search efforts. In a sense, image quality and completeness will become new SEO standards for image-enabled search.
A Post-Text Internet Raises the Stakes for Retailers
As consumers grow more comfortable with voice and image search for everyday needs, they’ll naturally gravitate toward retailers that provide these features and away from those that do not.
Search engines like Google and Bing will follow suit. As browsers demand and popularize new search modes, search engines will tilt their algorithms to prioritize companies willing to play along. This means if a retailer hasn't provided multiple high-resolution images to satisfy product feeds or cannot update its SEO practice to accommodate unique voice searches, it’s unlikely it will show up at the top of consumer queries, if at all.
The stakes are high, but so are the rewards. Retailers that invest in post-text digital experiences will deliver a pleasing and positive experience that consumers come back for.
Ed Kennedy is a commerce strategist at Episerver, a company that unifies content and commerce in one platform.