Stay marketing-savvy and tech-savvy. Get the latest in martech by subscribing to MarTech Today.
Bot-to-bot marketing is coming soon. Are you ready?
Martech's future isn't only about bots and agents talking to consumers. Here's what it means when they start talking to other bots and agents.
In the past year or so, a new marketing channel has emerged around bots and intelligent agents.
This includes voice-based intelligent agents like Google Home’s Assistant or Amazon’s Alexa, and chatbots that interact largely through text conversations. Marketers are beginning to plan their conversational strategies and logic for this channel.
But another channel — maybe it should be considered a sub-channel — is about to emerge. It’s when the bot or agent, fulfilling the needs of the human user, is interacting with another bot or agent instead of searching the web, or a knowledge base, or a profile.
Just when you thought digital marketing strategy had reached its stasis, we may be in for another major reinvention.
In a few years, ArcTouch Chief Experience Officer and founder Adam Fingerman predicted, “most interaction will be conversational.” His company develops apps — and now bots — and he’s conducting a session on “Bot-to-Bot Marketing” at our MarTech Conference in May.
In a common scenario today, you might ask Google Home to find a hotel room under $450 in Chicago near the Loop.
Google Home searches the web, possibly including reviews on TripAdvisor.
To support this user-to-agent/bot interaction, Fingerman told me, “marketers will have to spend more time on reputation management” to ensure that what Google finds is generally good. Hopefully, that will translate into being one of the top results, since a voice-based agent is likely to return only a few.
He also suggested brands may need to provide more structured data — possibly in a specific industry-based taxonomy — that makes attributes more transparent so Google Home can do apples-to-apples comparisons about, say, hotel rooms.
The importance of loyalty
Another tilting factor can be loyalty points, which Fingerman said will become more important since you may direct Google Home to Hilton, where you’re a loyalty program member.
Eventually, Hilton will want to become more of an active player in this new kind of user interaction. It might pay for top placement when “find me a hotel in Chicago” comes up in a search by an agent or bot.
So far, these are fairly standard marketing tools — reputation management, positioning, loyalty programs, paid placement — in response to what is essentially a voice-controlled intelligent search engine.
But this can change when the bot/agent acts on your preferences, either stated by you or learned over time from the decisions you make.
That’s because the natural evolution of this preferences-based activity by a generalized bot/agent is a personal bot or agent that acts only on your behalf. Perhaps it can search for anything, or perhaps it specializes in hotel rooms. In that case, you might also have bots/agents for finding a great restaurant, or getting the best concert tickets, or whatever.
Analyst David Raab pointed out to me that, when you are represented by a personal agent/bot, “brands will have to speak to your preferences.”
But those aren’t necessarily the same kinds of preferences that, say, guide the personalization of web content when you visit a particular publisher’s site. Or that cause Netflix to show you more thrillers, because you watch so many of that movie genre.
It’s not yet clear exactly how your personal bots will express your preferences, except in their selection of products to recommend to you. Perhaps a preferences profile will be available from the bot.
At the very least, this means marketers’ systems will need to track what your bot(s) chooses, in addition to what you choose. This can lead to new kinds of targeting segmentation, where advertising or marketing systems are segmenting and targeting not just you, but your bots.
And it could be more complex than that, because it may not be entirely clear the extent to which your personal bot actually represents the entirety of your preferences.
As Raab told me via email:
Consumer-side bots will need to be trained. A bot might just listen to you order pizza a few times to get some idea of how you do it, and then take over, perhaps asking for coaching during future interactions if it’s not sure what you want. People will vary in how much control they’re willing to delegate, and those preferences might change from one day to the next.
You might have personal bots with different authorities to represent your different interests, just as you might have a separate real estate lawyer, a divorce lawyer and a tax lawyer. Or you might have a bot with a deeper understanding of what you like in shoes, compared to another one that only finds slippers for you.
For one thing, Raab noted, this could lead to a “competition among consumer-side bots,” where independent providers — incentivized by a percentage of sales — offer personal bots with differing intelligence levels or industry specialties.
Beyond the generalized intelligent agents/bots, and the personal agents/bots, it’s quite possible brands may develop their own. (Agents/bots might overlap with platform-specific applications like Amazon’s skills, which some brands have developed, but agents/bots can offer platform-independence and act as a general brand interface.)
Hilton’s bot might know how to match the attributes sought, how to offer the right kind of incentives, or even how to negotiate.
Brand bots “will need rules for how to interact with [either] people or bots,” Fingerman noted. Rules that don’t yet exist, and which may eventually have to conform with some middleware provided by Google, Amazon, Apple or others.
Inherent in all of this, both Raab and Fingerman point out, is a new incarnation of trust. It might simply be another form of brand trust — i.e., I trust Amazon and therefore I trust Alexa.
Or it might be earned trust, where you learn to trust a personal bot provided by a third party because it returns good choices, the way you might learn to trust Pandora’s taste in how it programs a thematic radio station.
Raab pointed out that all of this belongs to a broader vision of autonomous technology acting on our behalf, described as “Digital Cohesion” by network solution provider Juniper Networks.
In a white paper on the topic, the company broadly defines the term as:
“… a future in which applications connect and self-assemble to deliver compelling mega-services that enhance our lives.”
In one implementation, for instance, a system-wide intelligence monitors a diabetic man’s medical sensors and knows when to intervene and what services to call.
Implied in this vision, of course, is that bots and agents are talking to bots and agents.
So, it’s not just about finding the best nearby pizza. It could be almost every kind of consumer, health or business choice will be mediated by intelligent bots/agents, probably layers of them. Marketers are just beginning to get a sense of that emerging ecosystem.
And you thought the digital universe was settling down.