The State of Chatbot Commerce in 2017
A few hundred messaging enthusiasts expand my perspective.
Last week at Jeff Keni Pulver’s Monage conference in San Jose, I was fortunate enough to be both an organizer of the chatbot track and eager attendee for the rest of the conference. Being embedded with a few hundred messaging enthusiasts and dozens of speakers covering chatbots, AI, IoT, messaging networks, and adjacent industries seemed like a good place to get the industry pulse.
How is 2017 shaping up for the chatbot economy and more specifically the commerce sector? It’s not 2016 anymore with all the chatbot hype and weekly postings and pitch decks featuring that graph.
Commerce Chatbots Panel Highlights
The most direct space insights we can get from a few chatbot companies themselves. Following their demos, I moderated a chat on the commerce space with Hipmunk and Cently by CouponFollow.
Interestingly, both companies started on the web and other channels before moving to chatbots. They had successful services people already desired and chatbots provided another channel to reach them. That’s an immediate distinguisher from many of the bargain basement chatbots from last year existing simply to be a chatbot and not providing noticeable value. Amir Shevat, Head of Developer Relations at Slack, nails it.
A bot is only as good as the service it exposes — Amir Shevat
Cently actually started in Slack where the strong developer experience was a big factor in the first platform choice. The more technical audience may also play a factor in comfort level with commerce through new avenues. Given the expanded reach, CouponFollow said it will officially debut for Messenger in late April as well. Cently allows consumers to easily purchase from hundreds of top retailers directly from the chatbot. Contrary to existing retail shopping chatbots like Spring, Cently has focused on making the checkout experience inside the Messenger application, and as seamless as possible. As a nice tie-in, the bot also applies retailers’ promo codes automatically by tapping into CouponFollow’s core coupon database.
Cently believes bot driven commerce is still new and being explored by mainly early adopters out of novelty though Amazon Dash, Amazon Alexa, and Google Home will bring more mainstream interest.
Both Cently and analysts looking at Kik believe in a potential future world where the shopping platform’s convenience and reputation may trump at least some individual store or even brand loyalty. Extrapolating early indications from WeChat support this theory.
Hipmunk, which debuted their chatbot last year, carried over their playful logo and branding to the chatbot space where personality can sometimes net surprising amounts of user loyalty and press. Conversation is a natural fit for a user expressing their travel preferences. Rather than face a giant grid of prices on a website, the bot gives you targeted options based on your preferences. It goes beyond what’s even possible in a filter by allowing open-ended questions for finding flights to a location over the summer or finding you a beach in April.
Starting in the pre-chatbot days, Hipmunk laid the groundwork individually reaching out and partnering with 50 vendors one by one to provide comprehensive customer choice and value.
Having a broad base of affiliates avoids concerns over playing favorites or limited choice that may or may not be in the customers’ overall best interest.
Optimizing for reduced agony in flight choices was an early goal of Hipmunk and fits well with a chatbot cornerstone of providing personalized value. Hipmunk also can proactively anticipate and suggest travel options for your upcoming meetings.
This is a fascination space and, as a frequent traveler, I find it interesting to consider what further data and preferences can be made available about flights. Age, type and comfort level of planes, service level, food choices, luggage costs, and more are places where customer experience is affected. A conversational experience is an excellent vehicle for this expression.
Checkout Is Not The Only Goal, Even In A Commerce Bot
Having a chatbot whose sole goal is a sale ignores everything that’s been learned over decades of sales and marketing funnels. You don’t want a single opportunity to make the sale. You build a relationship, re-engage, and may have another opportunity in the future. The chatbot form factor is a fantastic channel for engagement and for analyzing your customer behavior and tools like Dashbot make it much easier to save, query, and visualize this data. Learnings from the customers who did NOT purchase can be as important as information from customers that did.
“Conversational data is richer and more actionable than traditional web metrics” —Dennis Yang, Dashbot.
Elizabeth Snower and Audrey Wu combined their past chatbot experience from Kik and Imperson respectively in their fresh company CONVRG and brought some great insights to enforce the brand value of chatbots.
In one of their examples, 70–80% of users answered three question surveys (even open-ended ones) for a simple coupon due to an easy chatbot experience. The actionable data that can result from such a survey is incredible and the response rates have easily been 3x or more compared to traditional e-mail surveys.
What else can a chatbot for a commerce brand do?
“The best chatbots happen when you take your brand and personify it into your best salesperson.” — Liz Snower
As the oldest and largest chatbot platform, Pandorabots has over 245,000 registered developers and 300,000 chatbots created. Their CEO, Lauren Kunze, shared some impressive stats from some of the chatbots on Pandorabots. Further supporting CONVRG’s experience, Lauren notes that while email marketing has an open rate of 24%, a recent chatbot rocketed past that at 80%. The low friction, direct engagement, and novelty factors make chatbots an amazing way to learn about customers or potential customers.
Kik’s bot store has a strong fashion and consumer brand presence and initial indications are positive.
Peter Buchroithner from swell.biz is also betting on the market research potential of chatbots.
Conversational Marketing — The Evolution of Marketing?
Andy Mauro, CEO of Automat, thinks so and gave a compelling presentation. In short:
Rather than tracking users with pixels and cookies, which is imperfect to correlate at best and results in roughly targeted ads and e-mail campaigns, why not actually engage them, learn about them, and provide value in return in the form of products and experiences that actually meet their needs?
Creating well targeted drip campaigns is a lot of work and can’t come close to the segmentation and personalization potential of data streams willingly entered and processed with AI.
The following slide points from Andy’s session are reproduced verbatim as they are incredibly powerful:
Conversational Marketing
- turns awareness into action
- lets your content drive commerce
- nurtures bi-directional communication
- learn about customers in their own words
- is the first step towards an AI first world
Other Industry Thoughts
Last year Shopify put work into allowing their customers to easily create Facebook Messenger chatbots for customer support inquiries after acquiring Kit CRM. Today, Shopify customers can create Messenger bots that do many things including purchasing products. While a number of bold mainstream retailers like Sephora, H&M, and Tommy Hilfiger have tested the waters in evolving chatbot technologies, it may be some time before many other retailers are willing to invest significant resources, especially as most bots proceed with little conversational context and personalization being captured and the ROI just from purchases is not yet proven.
Industry analysts like Business Insider still consider direct sales and affiliate related transactions (bot-leveraged affiliate marketing) as key chatbot monetization strategies along with more intangible brand value adds.
Recently a Seattle-based company, ReplyYes, who specializes in technology for buying products via text message, raised $6.5 million for their conversational commerce platform.
Facebook accepting payments on Messenger and Paypal through Slack signal the interest for payment options.
eBay has been working on their ShopBot chatbot, which is currently in beta and intends to enhance it with computer vision, NLP, and more. A well-regarded article from Elaine Lee explores their design process.
CBM Feature: 11 Examples of Conversational Commerce
Voice-First Commerce
Aside from chatbots, there’s been major moves from giants like Amazon allowing purchases through its household Echo gadget (though it’s already had some hiccups). Companies are already leveraging the Alexa platform including Dominos, 1–800 Flowers, and Kayak which have already released Alexa Skills allowing consumers to quickly order pizza, flowers, or book travel.
Amazon’s low barrier of entry with their Alexa skills program is giving it wide potential reach.
Failing Fast And Iterating To Success Or Just Failing?
It’s hard to not mention the elephant in the room despite generally very positive signals. Many chatbots on Facebook at least are getting questionable engagement. It remains to be seen where it’s due to a lack of curation and eager but inexperienced developers/conversational designers or something more fundamental and long-term. Sarah Perez from TechCrunch, reported earlier this month that Facebook Messenger chatbots are now losing steam, and even prior to that she had struggled to find them useful.
Furthermore, Facebook is constantly changing their chatbot platform and seemingly looking for any way to maintain traction. Sarah Perez appears to believe chatbots are about conversations, while Facebook’s Mikhail Larionov predicts more structure and less conversation in 2017. Webview style interactions are becoming more prevalent. Slack is also building a richer UI into their messaging interactions:
- https://trello.com/c/jCPpCpP0/27-interactive-messages-stage-2
- https://trello.com/c/YiHuccUu/29-interactive-messages-stage-3
Perhaps each change is a step in the right direction, or perhaps it’s confusing the entire ecosystem (developers and users); only time will tell.
Given the later start of Alexa Skills compared to FB Messenger, it will also be interesting to see how that strategy plays out in voice. Will Amazon run into the same quality/engagement hurdles of Facebook or become the undisputed voice commerce solution through breadth?
What do you think? Is there a bright future for commerce in chatbots, or is this just passing fad?
Interested in having these kinds of conversations live? Monage is coming to Boston in October! Sign up to speak or get updates on the event.
