Ultimate Guide to Creating a Chatbot Persona for Your Brand

This content has been archived. It may no longer be accurate or relevant.

From Userlike:

Generic bots are out, unique bots are in. Finding the right chatbot persona for your brand takes effort, but it’ll help make the rest of the design process flow.

Chatbots often give a first impression of your company. Whether it’s greeting your website visitors or helping customers with their orders or inquiries, chatbots communicate with your customers in a direct, personal way.

Coming up with a unique persona will not only help you build stronger personal connections with your customers, but it will add fun and clarity to your development process. This post will cover why a chatbot persona is so necessary and steps for getting started.

. . . .

Why chatbots need a persona

Chatbots are for humans, by humans.

People naturally project human traits onto everything. If you cut corners on your bot’s persona, chances are users will assign it one anyway. But the results could be unfavorable.

A rich persona is memorable and can make your bot feel more like an extension of your team. In return, this could help reduce workload for your human employees because users will feel more comfortable speaking to it instead of asking to be forwarded.

Personas make script writing easier

When creating your chatbot conversation flow, a backstory and persona will make it easier to find its voice. It’s an approach often used by screenwriters and authors to help drive their character’s dialogue and actions.

A chatbot persona doesn’t need a whole family tree and fabricated trauma to perform well, but a motive, experience and possibly even age will add character.

Take Cleverbot as an example. We don’t know how old it is (I asked and all I got was a “Good. How are you”) nor its background, but we do know that it’s clever in every dry sense of the word. This fuels its interactions. If you try to ask anything personal, it deflects with some witty remark. Its persona is simple but clearly defines its conversation style.

. . . .

Promote your brand’s identity

Technology can’t replace the human need for personal connections. Intelligent AI may be your solution to pain-free chatbot service interactions, but people want more than that.

Without a persona, a chatbot can seem empty and cold. Creating an identity modeled after your brand builds empathy with users and mirrors the personalized engagement they receive from your team.

If your business is online or rarely ever customer-facing, then a likable chatbot person may help your brand stick in a customer’s memory. They likely won’t remember the exact conversation they had with your bot, but they will remember how pleasant — or irritating — it was.

The industry demands it

We’re surrounded by Conversational UI. Chances are one of the devices you own has a digital assistant built into it, likely with a pre-determined persona like Siri or Bixby.

It’s no real surprise, especially once you consider that consumers prefer text communication over phone calls and face-to-face service.

But it’s not easy to write for Conversational UI. That’s why more and more companies are putting out job ads in search for conversation designers to do the tedious creative work for them. Every chatbot writer and creator needs a jumping point though, which is where personas come in handy.

Link to the rest at Userlike

PG understands that some existing chatbots are so engaging that some individuals (undoubtedly possessing closely-related personality types) have been spending more and more time with their chatbots and less and less time with human beings.

Of course, some individuals possessing an authoritarian temperament want one or more government agencies to regulate videogames and chatbots and everything else so modernity never causes uneasiness in their minds. These are the same people who wanted to regulate the Internet for the same reason a couple of decades ago. PG suggests they’re no match for even a moderately intelligent AI.

Most of us already know one or more individuals who have become addicted to playing one or more videogames (teenage boys, of course, but also an increasing number of other groups of people) and spend a great deal of their spare time playing videogames. To the point that the virtual world seems more engaging than the actual world (“meatspace”).

PG doesn’t regard himself as a member of this cohort, but will admit to interacting with Alexa on a regular basis for the purpose of giving her the task of reminding him of things like an appointment with one of the small fleet of physicians who minister to PG’s aging physical self.

Or even more important, Alexa reminds PG about birthdays, to-do lists and when he needs to pick up Mrs. PG from one of her outside-of-Casa-PG pursuits.

7 thoughts on “Ultimate Guide to Creating a Chatbot Persona for Your Brand”

  1. Methinks the OP’s headline was written by a chatbot. Only a chatbot could have the kind of boundless optimism combined with arrogant self-regard to make a claim concerning an “ultimate” plan in a rapidly-changing field.

    Wait a minute: Election Day is a week from today. Never mind the second sentence — but I still suspect the chatbot wrote the headline (and that says everything about the reliability of the OP itself).

  2. *Raises eyebrow*. I will need to see this in action. Every time I see a chat function, my goal is to quickly establish if I’m talking to a human or an NPC. If an NPC, my next goal is to reach the “talk to a human” option.

    Note, I’m not against a VI. I wouldn’t mind something like AVINA from the Mass Effect games: tell me directions, store hours, what’s nearby, and basic stuff like that. Basically Google maps, calendars, or a FAQ, but with ‘tude and perhaps an avatar. But when it comes to “support,” if I’m actually contacting you that means I’ve reached the “talk to a human” portion of my troubleshooting.

    • Several online sites are honest enough to let you know you’re dealing with a chatbot.
      Not all/will be because the point of the chatbot is to “not* have a human on call.

      This will be most attractive to smaller operations: It might be to save money, not having a knowledgeable person on hand, or having better uses for the ones they do have. The choice might be between the ‘bot and no support.

      (Of course, there will be operations like Amazon.com that have never offered real-time interactive support.)

      Amusingly, the best user support interaction (and most complicated/extensive) I’ve had recently was with MS–an update that downloaded but refused to install–and that took bare minutes to go from unhelpful ‘bot to friendly and competent human(?). If that was a ‘bot text chat it was way more advanced than bingchat. 😉

      The time is coming (and soon) when a chatbot *will” be more helpful than a human working off a canned script for first level support and beyond. No choice then.

      • Step 2 of “talk to a human” is “short circuit the human’s use of a script,” so I can get on with the actual problem solving 🙂 I’m against bots because it’s just an unnecessary intermediate detour to getting to Step 2. By the time I contact a company I’ve already Googled and visited forums or watched a video or something. Talking to the human is actually a last resort for me.

        Of course it will not surprise me if companies did eliminate the human in favor of the low-cost machine. In that case, though, my dollars would have to be spent only on the most popular of products that already have an extensive user base, so I *can* find customers who’ve already encountered whatever the problem is, and have written tutorials and whatnot. I would be less willing to take a chance on a newcomer, or be an early adaptor. I’ve just run into several situations where the people who make a product are terrible at explaining (in written form) how to use it, so crowd-sourced support is my only option in that case. Savvy startups would be wise to emphasize their customer service.

      • F. – I think the chatbots will take over first-line technical support in a lot of places pretty quickly. It’s not unusual for first line tech support people to rely on scripts and/or structured problem-solving software to enable their employer to deal with 80%ish of the tech support calls without spending much money for salaries or expensive tech solutions.

        The more complex problems or large corporate customers go to the more experienced and/or better-educated employees in tech support.

        AI’s will be able to handle the script and/or structured problem-solving tech support issues pretty quickly and the low-paid employees will be replaced as quickly as their employer can roll out enough bots.

        With more complex problems, AI may provide some assistance, but this is an area with far more one-off problems than first-level support deals with. That said, all the recorded verbal interactions between advanced tech support and customers with complex problems are likely to be fed into the tech support AI system along with any keystroke-based problems and solutions.

        But I could be wrong.

        • No, your read is in line with existing automation and labor trends. The choices are robots (harware or software and nothing.)

          What handwringers forget is the US (and most of the developed world) is in the midst of a demographic downsizing. Boomers are retiring and the number of people reaching working age 20-60 or so is less than the number going away. This trend will continue for another ten-twelve years and since the people leaving are experienced and skilled the gap is not going to be filled by the uninvited storming the borders. (In fact, not only are they unequipped for the open jobs, the jobs they might be equipped for are no longer existent.)

          Automation by hardware or software is necessary just to meet the minimum needs of the labor market which is why there is such a stampede towards “AI”; it is the right tech at the right time, much like PCs were the right tech for the late 80’s and 90’s.

          As is, the US is in better shape that europe and east asia; for all that can be said about millenials (and that’s a lot) they exist in large enough numbers (75M vs 78M boomers at peak) to keep the economy rolling (with automated tools to assist). Few other countries can claim that.

        • One thing: it isn’t just chatbots that will fill in the gap from retiring experienced employees. As I keep harping, chatbots are the least important use of the new tech.
          Here’s a bigger one:


          “The project will create AI copilots to assist staff at customer companies as they design new products, and organise production and maintenance.

          It examines information gathered by Siemens and helps customers quickly create, improve and debug complex automation codes and shorten simulation times at their factories and other facilities.
          Schaeffler has been using generative AI to help its engineers programme industrial automation systems like robots. It intends to use the Siemens Industrial Copilot to reduce production downtimes at its plants.

          Tasks that previously took weeks to complete could now be completed in a matter of minutes, Siemens said.”

          Note that Germany is one of the countries at highest risk from their demographics.

          In a related report Microsoft today released their copilot add-on subscription for their big corporate office app subscribers at $30 a month per user. Estimates are that if they get 18% of subscribers to sign up for Copilot it’ll be a $10B a year product by 2026. Two years.

          Microsoft is not alone in driving “AI” into the enterprise market; they just got there “firstest with the mostest” and are making big money off “AI” while the competition is still testing prototypes and the idiotpoliticians run around like scared chickens, with no idea what to do or even if they can do anything.

Comments are closed.