Are you a robot? Hell yes! … better get used to it

Christopher McLachlan
6 min readDec 3, 2016


Companies need to get ready to deal with robots
that help customers to manage their service providers.

November 16 2016 was probably a memorable day for the pricing team at Comcast. It was the day that TRIM officially launched a service to renegotiate rates of Comcast customers using a chat-bot. These customers can watch the TRIM-bot negotiating on their behalf (auto-pilot mode), but can also jump into the conversation any time by switching to “manual flight” mode. To make things worse for Comcast, but even more fun for their customers, TRIM offers a game overlay for its auto-pilot mode … so they can play Tetris, Brick Breaker, or Snake while the robot negotiates their cable TV prices in the background:
It was never easier for Comcast to lose touch with their own customers.

… but why?

One step back: As many companies also Comcast, over the last years, has developed a broad set of tools, skills and procedures to ensure that their websites, apps and channels are highly responsive to customers’ needs. The goal is to handle searches, questions or requests efficiently and effectively across all channels and even to anticipate customer moves. Tools like IBM Watson, or Adobe Analytics (and many more) can now provide insights in real-time, like which valuable customers show intentions to switch and how they can be won back. Never before had marketeers so powerful analytics, so advanced content management systems as well as responsive channels at their hand, allowing for personalized customer interactions at scale. McKinsey calls this technological enablement the Holy Grail of Marketing as it provides value to customers and results in better sales and less marketing spend. But this “holy”” scenario only describes how technology helps companies to better address and serve customers (#machine-to-customer marketing).

Companies like TRIM, however, show that the tables are starting to turn: Customers now can use the same powerful technology to manage supplier interactions on their behalf … better, faster, with less hassle. They can tap into the collective learnings of hundreds of similar customer contacts that are being analyzed. If a robot learns which buttons to push in a conversation to get higher discounts or better freebies, this learning will be accessible for all other upcoming negotiations right away. Imagine TRIM’s chat-bot continuously being fed and trained with new Comcast responses and developing scenarios that show the highest likelihood to succeed. This is when predictive customer behavior analytics of Comcast face their new counterpart: predictive company behavior analytics of customer robots (#machine-to-machine marketing). The obvious question for Comcast (now) and many other companies (soon) is:
How to deal with this new dimension of customer interaction? How to deal with such a highly knowledgeable, well trained, and fast learning counterpart?

The easy response would be not to allow robots to act on behalf of customers. Applications like reCAPTCHA from Google are commonly used throughout the web to manage (restrict) website access. Their goal is to avoid seemingly “malicious” robots from opening and managing accounts or from collecting information. However, reCAPTCHA’s claim “hard on bots, easy on humans” seems shortsighted for a world where robots will increasingly take over tasks with a clear mandate of their users. Not every robot accessing websites means fraud or a denial-of-service attack. Moreover, it seems strange if first companies equip themselves with all analytics and intelligence tools possible to optimize their business processes, and then exactly those companies would react irritated if their customers do just the same: streamline their personal processes, optimize their time.

… and now what?

“Embrace change” is probably easier said than done for many companies, but it is what it ultimately boils down to. As with the mass-market introduction of other technologies (web, mobile, …), also the machine-learning driven empowerment of customers offers substantial benefits for those companies that understand the new rules of engagement. Two examples: Customer robots scanning numerous alternative offers will not only be able to look at the price, but can run scenarios on total cost of ownership and dig into the small print of terms and conditions - ruling out those suppliers that try to use tricks. Price comparison websites, typically sitting between the customer and their supplier, might loose their appeal if robots start to compile such top-x lists based on the individual preference criteria of their users.

So, how to get started?

  1. Think how and where this will happen first in your industry
    Start with asking yourself, where a set of “digital helping hands” could make the largest and/or the easiest difference for your customers? Is it in the way they (have to) buy products and services? Or is it about their involvement and own efforts in service delivery? Also think about the way, your own organization is programmed and can be “hacked”. TRIM basically chose a very narrow topic, that will create noticeable benefits in an easy way at scale. Moreover, they chose a channel that is ideal for Natural Language Processing and, thereby, is ideal for using a chat-bot to “hack” the way customer retention has been designed so far.
  2. Understand mandate, mission, and setup of customer robots
    Just to be clear: Having a deep customer understanding is and will be key. But now is the time to start adding robot behaviour analysis to the task list of marketeers. As robots will increasingly support parts of customer decision making, it will become just as important to understand how they are setup: Which parameters do they reflect, which databases do they tap into, which algorithms are used, and what are the robots allowed to do vs. when does it need to check with the user. With a deep knowledge on robot design, companies will know what topics to focus in sales and service conversations.
  3. Shape the conversation with both: Customers and their robots
    Ultimately, the goal of bots like the one from TRIM is to help customers make better decisions and to take away some of the chores of executing it. To do so, they should increasingly represent the full set of needs, values and concerns that customers have which go well beyond price only. This is an open invite to companies to engage with customers in a discussion beyond service fees and rate plans. This is the opportunity to ensure that a broader set of decision parameters is reflected by customers’ robots. Small prints of contracts, or sustainable sourcing of raw materials and labour are likely starting points, to allow for a more differentiated assessment of the offers at hand.

So, the dramatic shifts we expect in the way customers and companies will interact, will lead to a similar shift in the way marketeers need to interpret and inform their role going forward. However, the fundamental goal of marketing - to facilitate transactions and create value for both: customers and the companies - holds more than ever.

If you liked the read, please show some robot love and hit recommend ;-)
You want to read more: Please check out my blog

Music video by Beck performing Hell Yes. © 2005 Interscope Records



Christopher McLachlan

Customer driven, hands-on, full stack marketeer and commercial strategist at - helping SMBs to stay on top of their utility matters.