E-commerce in 2017 isn’t what you think it’s going to be
This week the team at Semantics3 decided to take a day off and chill at the Yogis-thaan Cafe in Bangalore.
In the midst of our zen, we tried to predict what’s about to happen in 2017 — post-Election, hopefully while we still have a functioning world.
Here’s what we think it’ll look like.
In 2017, there will be a landscape shift in retail:
Traditional brick-and-mortar will shrink, with an increasing shift to online retail
This is kind of important. In all the years previously, traditional brick & mortar has dominated retail. From Sears, to Walmart, its always been about physical locations. Even mail-order catalogs failed to usurp brick & mortar.
But in recent years, e-commerce has started to creep up. Starting with Amazon Prime, Shopify, the rise of online streaming and shopping c0upled with the vast improvements in logistics networks, the stage is set for e-commerce to fundamentally take over retail.
But this also means that a tonne of immutable assumptions we previously made about retail are about to change forever.
Let’s ask a couple of questions here:
Traditional brick-and-mortar retailers are transforming into retail marketplace platforms
An area that we’re seeing considerable tectonic shifts is the increasing number of traditional brick-and-mortar retailers that are starting their own retail marketplaces.
This is an interesting development as it signals the shift of brick-and-mortar retail from offering a traditional retail shopping experience to more of a platform player — where they leverage their existing assets: supply chain, warehousing, and fulfillment to enable third-party sellers to sell their inventory through these assets.
This is likely to be a temporary shift though, as retailers try to bridge their move from brick-and-mortar to online retail.
In the future, it’s more likely that an increasing number of retailers will seek to divest themselves of expensive real estate assets they currently own — strip malls, flagship stores and prime location space and move more of their sales efforts online.
That changes a lot of ground realities — and it’s happening already. Suburb designs are not going to be centered around retail centers, downtown areas would feature less flagship stores, and commercial real estate values would drastically reduce, given less demand for floorspace.
In a fantastic trading of roles, more online retailers are also opening flagship physical locations — more to offer customers the chance to “showroom” products before purchasing — but with much smaller footprints.
Many retailers, like Amazon, even Modcloth are currently experimenting with this to great success: Modcloth recently opened a temporary “FitShop” in downtown San Francisco, which we loved
If e-commerce usurps traditional retail as the main source of sales, what would retail advertising and marketing look like?
For one, shoppable media is going to be a thing
Traditional marketing for retail has always been a bit weird. You spend millions of dollars on cable and TV advertising, but offer little means for people to instantly respond to these ads!
This is even more frustrating, given that more and more consumers are increasingly streaming their content online, and not watching traditionally scheduled TV programming.
Streaming media offers a powerful opportunity here — given that most people are going to be watching this media on their mobile devices, laptops, and on their Chromecasts or Rokus, there’s a huge untapped opportunity to embed purchase URLs directly into the content themselves.
Youtube already offers this functionality via their captioned video functionality.
We think that traditional types of marketing aren’t going to go away — but they’ll tweak their format significantly — making it easier for the target audiences to engage with the retailer more effectively, with better methods of measuring conversion.
AI is about to play a much larger role in retail than we previously thought
A few weeks ago, our fantastic Engineering team published a post about how we created an artificial intelligence system to create content automatically, using our existing corpus of blogposts as starter feeds
An initial run of this system generated a pretty good paragraph:
While it still looks pretty rudimentary, this is groundbreaking stuff — it means that we can actually have AI auto-generating content for e-commerce product pages; for e.g. product descriptions and copy which would otherwise take countless human-hours to create.
But that’s not the only thing — AI is already permeating other areas of our shopping life. From Alexa to Messenger Chatbots, Artificial Intelligence is making it more easier than ever to buy something without hesitation
While AI assistants like Alexa and Siri make buying stuff a wickedly cool experience — “Alexa, could you buy some toilet paper?” — a much more deeper application is the metadata collected by these applications that drive much of e-commerce personalization.
Truth, as it always is, emerges from large datasets
By tapping into the vast number of conversations people have with their Alexas, (and probably countless more conversations which Alexa is probably listening into), Amazon is able to build a rich, comprehensive view of its customers — what they like, what they are talking about, and more importantly, be able to predict what they might like next.
And all of this feeds into the almighty Amazon product recommendation engine.
This has come to a head with the launch of Amazon’s DSSTNE (pronounced ominously as “Destiny”), short for Deep Scalable Tensor Neural Engine , which is used to “train neural networks and generate recommendations that power various personalized experiences on the retail website and Amazon devices”
This leaves us with the final thoughts:
If you want your brick-and-mortar store to go online, how do you stay competitive in the e-commerce landscape? How do you price your products online? How much inventory is needed?
Data can drive these crucial decisions by providing price comparisons, real-time pricing, and catalog data from retailers.
E-commerce is not stagnant and neither should your products be.
Going back to the shift of retailers creating an online marketplace — this could be an enormous endeavor both on time and money to have sellers list their products.
A simple solution could be to have sellers use a one-click import of data on their products. Literally, voilà.
And with the ever growing influence of content marketing, you can power your content with rich product metadata to help generate the relevant content you need quickly, and target your audience with a fine-tuned message, and this can be leveraged even further with good AI.
Your audience will really appreciate it.
In the next few weeks, we’ll go into much more detail on how 2017 is going to be the year of e-commerce.
Written in San Francisco by the Semantics3 Team.
Published at: September 22, 2016