The Basic Flaw Inside Every Chatbot


Chatbots are a relatively new addition to the tech landscape.  Last year at this time, they simply weren’t on the radar, for the most part.  Of course, Amazon’s Alexa has been around for a while, and Apple’s Siri for even longer than that.  Chatbots really took off, though, when Facebook released the API platform for its own chat service, Messenger.  That one act spurred developers to create over 40,000 chatbots in eight months.  They’re still making them, too!  And if you count those created on platforms such as Kik and Telegram, add thousands more to that figure.

The function of one chatbot is pretty much like that of any other — after receiving input from a user, it gives information to that user via a text message.  The aim of a chatbot, though, can vary widely.  Some provide beauty tips, some allow you to order food, and still others satirize political candidates.

How do you make your chatbot stand out from the rest?

Chatbots are mostly the same behind the scenes, as well.  They all do the same thing.  A chatbot accepts input, then provides information based on that input.  Thus a user’s experience in one chatbot is pretty much a duplicate of another.

So what makes a bot special?  Or more importantly, how does a developer create a chatbot that’s distinctively different than another?

An abundance of bots

This question is becoming increasingly important as more and more bots vie for users’ attention.  Forty thousand chatbots is a huge market to try and get a share of.  And that’s just a drop in the proverbial bucket, compared to the deluge coming.  The exact number can’t be divined, of course, but take a look at how they’ve taken off:  Since the number of chatbots has grown by roughly 20,000 in the last five months alone, it’s a safe bet that the number of chatbots will soon outstrip the number of apps available.

Ever heard that “There’s an app for that!”?  Pretty soon we’ll all be saying “There’s a chatbot for that!”

Chatbots scale very well

Say it’s Saturday night.  You open Facebook Messenger and visit the Uber bot.  You use it to order a ride for yourself and two friends.  You go to a restaurant and get the seat you reserved while you were on Messenger.  It’s exactly what you wanted — non-smoking, near the door, a table for three.  After you eat, one of your friends tells you about Louis Vuitton’s new fall fashions.  You open another chat, connect to Messenger again, and look up the company chatbot.  You start by looking at shoes, but on impulse you buy a new jacket instead.

This scenario is not at all far-fetched.  And you never have to leave Facebook Messenger to get it.  By now, messaging platforms are practically ubiquitous, so any brand can have one-on-one, personalized interactions with their customers.  And this can happen on an enormous scale.

Yet for this to happen — much less for it to happen well — there needs to be a richer experience for the customer.  Right now chatbots are menu-based.  And even though they provide a conversational interface, the actual exchange is anything but.  This circles back to the design of chatbots, their basic functioning.  Generally speaking, answers are scripted.  A user can’t type their own free-form input.  Instead they look at several structured answers, and choose one of those from within the chat message.

When you narrow the bandwidth, you focus the message

This leads to a certain sameness in the user’s experience, when moving brand-to-brand.  A lot of this is because of the medium.  By itself, text doesn’t offer much differentiation.  Take a chat from Telegram and it looks a lot like a chat in Facebook Messenger.  This is a real challenge for a brand that wants to make itself rise above the others by creating a unique experience for consumers.

Authentic conversations are the missing piece

But consider for a moment that the problem might not lie in the interface, but rather with the brands themselves.

The basic flaw in the chatbots available right now isn’t the fact that they’re not providing flashy interactions, or even that they lack the visual design possible with a regular web page.  Rather, brands are wasting the chance to engage their customers with one-on-one, personalized interactions.  Instead they offer tired interactions that are about as memorable as reaching out to your bank.  In other words, today’s chatbot interaction may as well be a menu-based, touch-tone call.

John Schwartz, Washington Post writer for science and technology, once said “That’s part of the reason that filmmakers like Woody Allen or Steven Spielberg shoot movies in black and white. When you narrow the bandwidth, you focus the message.”  Because chatbots lack the standard neat interactions and good-looking visuals, they force a brand into examining the thing that truly matters to an audience:  The conversation with the users.  More specifically, the content of those conversations.

Just selling things to consumers is so 1990s.  A brand needs to invest in the relationships it has with its customers.  A brand has to create affinity with themselves and their customers.  A consumer will be much more likely to buy from a brand they already have a relationship with.  And brand affinity is built the same way it is built with other humans:  Through personal interactions, through conversations.

And that’s the prime ingredient:  Conversation.  True conversation, as opposed to scripted, menu-driven “chats.”  And conversational content is crucial if a chatbot is to be effective.  Companies like Amazon, Microsoft and Google understand this.  They have hired multitudes of writers, including comics and novelists.  These people are helping to develop company chatbots so that each one has a unique voice, an individual personality.  These things are what make chatbots compelling; only by using voice and personality can they be good brand ambassadors.

It works both ways, too, which is a good thing for brands that lack a personality or a voice.  It is a chatbot’s duty to convey the personality of a brand, yet due to the hard work of the creative people that write their scripts, they may put more personality into the brand than might have been there in the first place.

Chatbots represent the next boundary in brand marketing and consumer engagement.  They give brands an exquisite opportunity to forge durable bonds with their customers.  But the first step is crafting an experience that is genuinely compelling.  Up until now, users have been willing to overlook the basic flaw of chatbots; they’ve been putting up with the boring interactions.  Now, though, it is time for chatbots to evolve.


Where’s the Killer Chatbot?

Image courtesy TechNode

Let’s face it: Most of the chatbot experiences today are pretty wretched. They’re stilted, artificial and in some cases downright affected. Natural language processing is still in its infancy, and has a long way to go before sounding actually “natural.” Or truly understanding natural speech, for that matter.

This is due in part to the difficulty of designing a user interface around a conversation, which is non-hierarchical in nature. When talking to another person, the steps don’t always flow naturally from one to the other. This kind of design is also fundamentally different than either a mobile or web interface. Additionally, we have yet to develop a general-purpose AI which can accept a user’s open-ended input.

UIs are crafted to serve linear experiences

It is incumbent on chatbot creators, therefore, to pick out engaging patterns of interaction. Building on and around these will enable developers to create whole experiences that will delight the users.

So how do we work around the limitations of a conversational UI, knowing the above?

About the UI
Up until now, User Interfaces have been crafted for a linear experience, not a random one. In other words, after the user comes to the page, a specific sequence of events typically happen, at least in terms of ecommerce. First they search for an item or two. Those items are then added to the user’s cart. They enter payment information, check out and leave the site.

A chat based UI is completely different from either a web or mobile interface. One of the biggest stumbling blocks is that the customer can initiate the procedure in different places. Say they want to buy tickets for a movie. The customer can ask a bot “What’s playing around 8pm?” Another valid starting point can be “I want three tickets to Trolls at the Regal on Little Texas Lane and Congress.”

So we see that a big challenge for anyone wanting to design a chatbot is that the path a customer will use to reach their goal (in this case, to purchase tickets) is not known beforehand. The chatbot has to assist the user and provide the desired answers without needing a discussion to progress in a straight line.

AI is Not Yet Ready
The next big stumbling block for chatbot developers is that a true AI that works on a variety of inputs is still a long way off. AIs themselves are not especially new, but they are new to the consumer marketplace. One AI-like construct that bot creators use a lot is the Simple Linear Tree, which forces the user down a predetermined path. New AI routines might also be used, but these are not true AI. They simply match patterns against pre-programmed conditions, in an effort to determine a user’s intent.

What we think of as AI is not truly AI

Generally speaking, these work well enough when there are a finite set of ways a user can interact with a bot. But as developers are finding out, user input can be totally random. This leads to situations where a bot can get unexpected input that it can’t handle. So without better tools, a better AI, it’s all a matter of hunt and peck. Or worse, finding the linguistic needle in a haystack of possibilities.

The Solution: Modify, Publish, Iterate, Repeat
So how does a bot developer succeed with the limited tools they have? The best path is not already defined, given the variety of inputs. Neither the number of inputs nor their content is known. There has to be a quick, iterative path to successful completion, and it has to be low-cost as well. A developer needs to be aware of how their bots are responding to the inputs provided by the user. With this knowledge, they can then iterate on what is already there. Any blocks between the user and their goal need to be addressed.

Users must be able to easily understand your conversational UI

Experience has shown that the best tools for the iterative method are bot native. This means they are able to understand the complexity and nuance of a conversational interface, and are able to translate them into clear metrics. Conversely, it also means the user is not simply dumped into meaningless dialogues or dashboards.

Marketing teams can use these tools to pinpoint groups of similar users, then connect with them through personalized messages. Creative and editorial teams can use them to address messaging that may be off-brand or that doesn’t have the desired tone. Business leaders can use them to provide a detailed picture of their efforts without the use of an engineering team and a data scientist just to “run the numbers.”

It’s important to have a conversational UI that’s easy to understand. It’s also important to iterate quickly on this. Being able to do these things will assist business leaders to grow differentiated bot-native arms that can leverage the great power found behind the conversational interface.

8 Truly Helpful Chatbots


There’s a lot of hype around artificial intelligence today.  And chatbots are one of the primary ways the consumer is made aware of AI.

Chatbots are surprisingly easy to create.  And the deluge shows no signs of slowing down, thanks to the hype, the ease of creation, accelerating investments, and the rush of developers.  Over 225,000 bot developers have created more than 300,000 bots, reports Pandorabots, a leading chatbot development platform.

In this case, though, there is lots of supply and not so much demand.  The rush of supply seems largely due to the “cool factor” of bots, and is not driven by anyone asking their favorite brands to start using chatbots.

Chatbots do well when focused on one task

Dr. BJ Fogg is the founder of the Persuasive Tech Lab at Stanford University.  When looking at his model of behavior, we see that in order for any shift to be made in consumer behavior, any given chatbot needs to be easier to use than the method being replaced.

Unfortunately, they’re not there yet.  Chatbots are still clunky.  As of this writing, it’s still easier to call, text or email the company in question.

At this point chatbots seem to do well when they focus on making a single task easier.  There are a few that realize the promise of AI, and that are able have more reasonable conversations by using context.  At this point, though, many users will still have only bare-bones conversations with most bots.

The sheer number of bots out there presents users with a problem:  How is anyone going to know which ones are truly helpful?  Sure, some bots may well assist a company’s cost-cutting measures.  Still, there is a lot of work needed before any customer gets a measurably better experience.

Right now there are a very few chatbots that provide some real utility and actual value.  They provide an inkling of what an interaction with an actually intelligent bot might be.  You might notice a theme among the bots picked:  They’re more realistic when the user wants to ask one of a group of prearranged questions.  They’re at their best when the tasks they handle are specific and clearly defined.

So here’s a list of the eight most useful chatbots out there today, in no particular order:

1.  Howdy’s function is to help you run your meetings in Slack.  Its founder, Ben Brown, created it with an eye to customization, so that everyone can use its talents of digital assistance and automation.  Howdy will reach out to participants and ask them a group of questions to prepare for the meeting.  It automates information collection and distribution.  It makes meetings shorter by leveraging a shortcoming of human nature — Howdy can talk to many people at once, whereas humans can’t.  These answers are then delivered to the meeting leader, and all participants, if the leader chooses to do so.  It sure beats copying and pasting the same questions over and over again.

2.  HealthTap was launched on Messenger not too long ago.  It aims to make good healthcare easier to reach by more people.  Not only does it allow a user to get a referral from real doctors, but it also assesses results of tests.  All that, and you can ask it health questions as well.  If other users have asked similar things, then it will show you those results.  Of course, health information can be highly personal, so HealthTap is ideal for those people that have no inhibitions.  The company does state that all questions and identity information is kept confidential and anonymous, but the fact that the company may associate a personal Facebook account with certain medical conditions may be a turn-off for some that prefer to keep a tight lid on their health information.

3.  Sephora chatbot on Kik.  Although it currently exists only on the Kik chat platform, the Sephora chatbot makes it easy for consumers everywhere not only to shop for products, but to get beauty tips as well.  Acting like the best in-store assistant, it provides three prime attributes of great customer service:  It responds right away, it’s highly educated, and it’s always available.  If you’re a member of Kik, it’s like having access to a Sephora team member in your back pocket.

4. provides Amy, a personal assistant with one function:  Amy takes on the task of scheduling meetings, so that you can focus on more important tasks. has been so successful with Amy because she does only one thing.  But that one thing she does extremely well.  She’ll schedule meetings without effort, but won’t take notes or solve any other problems for you.  Dennis Mortensen,’s founder has made a huge bet on the thoughts that bots are terrific for any very small task.  Due to this laser focus, the bot needs to be just about perfect.  Amy is their first attempt, and she makes scheduling conversational and easy.

5.  Assist integrates with several on-demand services such as Uber, GrubHub, OpenTable and Lyft.  A user can communicate with Assist through Facebook message, a text or by using Assist itself.  They can then get a ride, book hotels, or order take-out.  Assist uses input from users to improve itself with every question.

6.  Ozlo is a more personal AI.  At this time available only for iOS, he helps the user find information on their phone both easier and faster.  As with other bots, engage Ozlo in a dialogue text conversation, and it will give you answers on Google Now-like cards.  Deeper links are also provided to the websites and apps where it found the information.  The bot is partnered with other websites and apps such as Foursquare, Yelp, OpenTable and Zagat so he can provide great recommendations.  Right now Ozlo is being trained to help people find drinks, eats and entertainment — things that we all do daily. You can get more information about Ozlo from Charles Jolley, one of its chief designers.

7.  Pana blends AI with real humans to create an always-on travel agent.  Planning a trip is made easier, simpler, more personalized and cheaper with Pana’s skills.  Pana can also be your personal concierge, providing vetted choices for places to go and food to eat.  Since booking travel these days is mostly a matter of searching the same criteria on many sites, Pana makes booking easier and less time consuming.

8.  Birdly is a chatbot that acts as a bridge between Slack and Salesforce, making business-related data accessible to entire teams.  Directly on Slack, one thing Birdly can do is manage expenses.  This makes things more efficient, as it recognizes data input from invoices or receipts, allowing users to spend less time on expense reports.  Birdly also provides analytics, efficient alerting and customer information in one place.