One of the most powerful abilities a chatbot has is that it can gather customer insights.
By monitoring your customer’s patterns, you can discover which products or services they prefer.
You may even uncover some common issues your visitors encounter when using your site.
“If there’s one reason we have done better than of our peers in the Internet space over the last six years, it is because we have focused like a laser on customer experience, and that really does matter, I think, in any business. It certainly matters online, where word of mouth is so very, very powerful.” ~ Jeff Bezos
We totally agree with Mr. Bezos, but note that he did not share with the world his methods for monitoring the customer experience (at least not in this quote!).
The devil, of course, is in the details…and we are confident that the “details”, or secret sauce, for many future companies, are chatbots that gather customer insights.
With chatbots, we can all be as successful as Bezos’ baby: Amazon.com, right?
Not everything in the world needs to be complicated, and yes, we can all be successful in this online world of abundance. Sometimes the best solution is the simplest.
At Gobot, our mindset is that if your customers have the answer, why not just ask them for it?
Prior to chatbots, there were no tools that allowed website owners to truly engage a customer and gather customer insights, so as to allow for data collection at just the right time and place.
Poll tools have been getting prettier, but they are dumb – they just “poll” and in no way engage.
Pop-ups are focused on gathering emails and contact information, and are just about as engaging as the sound of nails on a chalkboard (sorry about that, even the words are painful!)!
Let me give you some examples of how you can use chatbots to gain real customer insights.
You have a complex and/or large product offering and find that shoppers leave because they cannot easily find what they are looking for. In some cases, they don’t know exactly what they are looking for.
You leverage a chatbot to effectively guide your customers to the right products for them – this is called guided selling!
Guided selling involves asking your customers a series of questions and then, ultimately, based on what you’ve learned, sharing personalized product recommendations.
Customers love this assistance because it makes the shopping experience super easy and less daunting – boosting conversions and sales.
Above and beyond the immediate improvement in customer experience and sales, the data gathered and customer insight is tremendous.
By way of example, consider Joe’s camera store. In-store, all of Joe’s sales reps first inquire as to whether the customer is a hobbyist, in which case the sales rep will typically initially recommend an automatic camera, and then ask customers for their favorite photography subjects/types.
When dealing with a professional they typically focus on cameras with detachable lenses. Joe leverages a chatbot to recreate this guided selling on his online store with great success.
The guided selling approach gathers information and let’s you, the store expert, recommend products, thereby taking the burden off your customers.
The byproduct of this chatbot sales guidance is a trove of data.
Analyzing this data, Joe can start to understand the various personas shopping for his cameras, e.g., hobbyist photographers versus professionals, and can start to cater his online sales approach appropriately.
Joe can leverage his chatbot conversations to understand what camera features are key or are being sought out in particular.
In this case, Joe came to the conclusion that his site was attracting hobbyists so he lowered his expensive detachable lens inventory and stocked up on options more attractive to hobbyists, and watched sales soar.
Stay tuned for more information on this critical utilization of chatbots for guided selling.
Let’s now take a look at another way chatbots contribute to customer insights.
You’ve just changed and improved the design of your website. Certain employees at your company vehemently question the change.
You could test conversion of both versions of your site, but this would not be definitive, as there may be other variables in the market affecting your conversion rates.
So you set up a chatbot that is triggered to appear exactly when your customer scrolls to the new portion of the website and ask:
Gobot: “Hi there. We changed the background color of this section from yellow to green.
What do you think?”
- love it
- hate it
- prefer the original version
After about a week, and 1,000 responses, Gobot reveals the following results:
- love it (860 responses, 86%)
- hate it (50 responses, 5%)
- prefer the original version (90 responses, 9%)
And so, you proceed with the final the style change, which most customers apparently love, and silence all the company naysayers!
And you do so even though your site brought in less revenue than last week, properly attributing that dip to three new competitors with low ball pricing, and not to the new and improved color palette of your site, which actually minimized the dip.
Here’s another scenario that demonstrates how effectively bots can gather customer insights.
You’re debating internally whether to post a sale for a pair of red high-heel shoes.
Your partner is convinced that the sale is unnecessary because the shoe design is so comfortable, sexy, and attractive, that customers would purchase it at a relatively high price without any sale.
Initial sales have not been strong, however, and so you want to discount.
You compromise with your partner on a limited term sale with the following chatbot in place that is triggered to appear immediately after a purchase is confirmed.
Gobot: “Thanks so much for your purchase! The red high heel shoes are very popular this season.
Would you say that today’s sale was the primary motivator of your purchase?”
- Absolutely, otherwise too pricey
- Not at all, would have purchased it either way
Almost all customers ultimately choose a), and so you extend the sale and cash in with volume!
Customer insights are the key to online success.
-Bravo, bravo, encore, encore – your audience cheers!
OK, one more example of how adroit chatbots are at gleaning customer insight:
You run an online supermarket.
You are concerned that if you include too many organic products on your site, you may turn off many customers who don’t eat organic and resent the high prices that the organic label typically comes with.
At the same time, you want to offer organic products which appeal to a larger customer base, because of the higher profit margin.
You have a filter for organic but you recognize that most people don’t use it because your supply of organic products is not sufficient in all areas.
What do you do? Use a chatbot to gather customer insights, that’s what you do!
You trigger your chatbot to appear as soon as the customer hits your supermarket site:
Gobot: “Greetings shopper! Do you shop organic at all?”
For customers who choose a) your chatbot discreetly shifts them to a version of your site that highlights your organic selection.
For customers who choose b) your chatbot shifts them to a version of your site that offers non-organic, aka, conventional items.
For customers who choose c), your chatbot shifts them to a version of your site that offers a mix of both organic and conventional options.
Everyone is catered to, and we all sing Kumbaya on the way to the bank!
-OK, OK, just one last story:
You run an online liquor website, which has been generating a modest but stable revenue year-over-year.
Your goal this year is to increase sales and take your site to a whole new level. Should you focus on your brandy buying customer base?
Add Vodka Red BullTM to your offerings? Before you engage in a new sales plan, you ask yourself who your customer base is.
And you are floored when you realize that after all these years selling booze, you have no clue!
So you put away your silver pocket flask, and scream at the top of your lungs, “Chatbots, help me out here!”.
Fully sobered by your epiphany, you arrange for a chatbot to survey your customers.
Gobot: Good choice! We already know you’re 21 or over! For 5% off of your purchase, would you please tell us which age group you fit into:
User chooses: a. 21-25
The chatbot results reveal that most of your customers are between 21 and 35.
Based on this critical information, and because you know that the younger crowd is lovin’ those energy drinks(!), you reduce your brandy visibility and you premiere Vodka Red BullTM for the first time – and watch your newly focused sales sky rocket! ‘
Did someone say “Party?”
This online liquor store example was based on a video I saw by my marketing idol and guru, Gary Vaynerchuk.
Gary started his business empire with an online wine store.
He’s an incredible role model!
Check out his hustle here.
As I have learned from Gary, customer insight is absolutely key in sales.
Why wait for a customer to leave your site to reach out for feedback?
With Gobot’s assistance, it has become that much easier to emulate Amazon’s laser-like focus on the customer experience!
Now, saying “I love you” the first time can be a daunting experience, right?
You might think that expressing such a selfless beautiful sentiment to someone you really care about would be the easiest, most natural forthcoming thing in the world.
The cold truth is, however, that it is HARD, very HARD.
Not because of the words per se, moreso the possible response!
What if she doesn’t say I love you back?
What if she just stares at you dumbfounded, unsure of how to respond?
Worse, what if she outright rebuffs you with an, “I’m just not that into you.” Wha’?
Rejection can be devastating, and so this is why people often don’t readily volunteer their feelings or inquire with others as to their feelings for them.
Many businesses suffer from the same rejection and fear-driven dynamic.
They work day-in and day-out on their business offering and their website but do so blind as to how they are being received.
Rather than outright ask customers their reactions, they play the wait and see game.
Wait and see if revenue increases.
Wait and see if more subscribers sign up, etc.
Often they wait too long, and the damage cannot be undone.
Enter the online SURVEY.
More and more smart proactive businesses are becoming evolved in their approach to its importance.
Facing fear head on, businesses are starting to survey customers to see how they feel about their offering so that they can adapt as necessary – and win in the market!
A simple question like “How likely is it that you would recommend our company to a friend or colleague?” followed by button choices numbered 0-10 (or 0-5 or using happy/sad/neutral emoji faces) is a very straightforward and efficient way to collect feedback from your customers.
More and more sites are starting to get on the bandwagon, and you are seeing surveys like this pop up all over the internet.
These survey tools help you gauge the loyalty of a firm’s customer relationship.
They are an alternative to traditional customer satisfaction research and high scores often correlate with revenue growth.
0-10 survey tools increase engagement because they are so easy to understand and respond to, but there is DEFINITELY room for improvement!
Here’s the honest truth: It is easy to click a number from 0-10, true, but why should a website visitor invest the mental energy to decide how they feel about your site?
What motivation are you providing other than the possibility that you might actually listen to the feedback and make responsive changes some time in the distant future?
For many visitors, this is simply not enough.
This world is all about give and take.
If you are going to expect a visitor to give…well, you should be actively thinking about what to offer in exchange.
0-10 surveys offer simplicity and require minimal visitor effort yet the survey tool in isolation, which is how the tool is typically displayed, is simply not designed with this basic understanding of human psychology in mind.
So if you’ve been using a typical 0-10 survey, now you are feeling a little guilty SELFISHLY asking for user feedback, without offering anything in return.
Sort of like going to a dinner party empty handed – how embarrassing!
So you start looking around to find a survey tool that goes a bit further, that allows for a dialogue with your visitor so you can be generous and give a little.
Your intentions are good, but your effort is wasted because no such survey tool exists…UNTIL NOW!
Chatbots allow you to embed your 0-10 survey directly into an engaging chatbot dialogue, so engaging that your visitor will not hesitate to respond to your survey.
The beauty of the chatbot’s version of the 0-10 survey is that it provides additional branches for collecting user feedback.
If a visitor gives you a score of 0, you can ask for ways to improve.
This type of feedback is more actionable than just a score.
But nothing speaks louder than an example.
The bot below is designed to appear as your visitor scrolls about one third of the way down your homepage.
Based on their rating, customers are classified in 3 categories: detractors, passives and promoters.
Gobot shows the numbers of customers that fall into each category.
Detractors gave a score lower or equal to 6.
They are not particularly thrilled by the product or the service.
They likely will not purchase again from the company, and you could expect them to damage the company’s reputation through negative word of mouth.
Passives gave a score of 7 or 8.
They are somewhat satisfied but could easily switch to a competitor’s offering if given the opportunity.
They are unlikely to spread any negative word-of-mouth, but are not enthusiastic enough about your products or services to actually promote them.
Promoters answered 9 or 10. They love the company’s products and services.
They are the repeat customers, are the enthusiastic evangelists who recommend the company products and services to other potential buyers.
Above and beyond the number of detractors, passives and promoters, Gobot’s analytics also include a survey score, ranging from -100 to 100.
The survey score is determined by subtracting the percentage of customers who are detractors from the percentage who are promoters.
At one end of the spectrum, if all of the customers were answering the question with a 9 or 10, then the survey score would be 100.
On the other end of the spectrum, if when surveyed, all of the customers gave a score lower or equal to 6, this would lead to a survey score of -100.
The survey score is helpful in that it can motivate website owners to improve and provide the best customer experience possible.
The ultimate objective here is to convert customers who were less than happy or unimpressed into promoters who will put the good word out and help create increased revenues and profits.
The beauty of Gobot is that it not only allows you to survey where your customers are on the scale, it also allows you to engage them so as to actively start to shift them from detractors to promoters!
The bot below includes a survey and responds to detractors in a proactive way.
Gobot: Sorry to hear, but thanks for your feedback.
Gobot: How can we improve?
[user enters: “your live chat operators don’t know your products!”
Gobot: Duly noted!
Gobot: We will definitely change that!
Gobot: In the meantime, we want to make it up to you.
Gobot: Enter your email and we’ll send you a promotional code for 10% your next purchase”
As Leonardo Da Vinci in his ultimate wisdom said so many years ago, “Life without love, is no life at all”.
So we say, never fear rejection – with chatbots, proactively survey your customers and arrive at that dinner party with a bottle of wine, your appetite, and hey, if you’re single, maybe you’ll meet the love of your life.
Enough with surveys.
Let’s talk about polls and social media sharing.
What do polls and share buttons have in common?
Not much, but they do suffer from similar problems, poor conversion and utilization!
Interestingly enough their problems arise for opposite reasons! Allow me to elaborate.
There are many tools out there for polls, surveys, and quizzes.
For purposes of this blog entry, we will be lumping all of these tools under the term polls for convenience.
Put simply, existing polls are typically slide-ins that present a series of questions to a website visitor, and tabulate the results for the website owner.
While the graphics and user interface choices are pretty sweet, all existing poll tools suffer from the same problem as screen highjacking pop-ups: they lack warm-up and charm.
There is something abrupt about the way the poll slide ins appear and demand answers, don’t you think?
Sort of like a painfully awkward teenager without any dating experience walking up to a pretty cheerleader in the middle of her dance routine, introducing himself, and asking her to the prom!
It’s no wonder that pop-ups and poll slide-ins have such low conversion rates – X that intruder immediately!
Share buttons are also typically ignored, but for the opposite reason.
They’re like the wallflower at the dance, too intimidated to even approach the cheerleader during a great song.
It’s possible the cheerleader would welcome an invitation to cut a rug, but the wallflower won’t leave the security of his/her immediate clique to approach the cheerleader.
Dating analogies aside, polls and share buttons have not realized their full god-given potential, because the former is too direct and in your face, while the latter hides in the corner and blends in with the wallpaper.
Chatbots have the potential to address these shortcomings of existing polling and share button tools.
Eeeny, meeny, minie, moe…ok, let’s address polling first.
It’s somewhat startling how naturally suited chatbots are for polling.
Just like the poll tools out there, they are more than capable of asking questions, but how they ask the questions makes all the difference!
First, chatbots can start with some small talk, rather than jump right into business.
Second, chatbots interact via messaging, which is a much more natural medium for conversation than a slide in.
Third, chatbots can engage and offer buckets of real value before asking for something back, e.g., answers to poll questions, etc.
One scenario we recently discussed focused on involved a poll regarding the visitor’s response to a redesigned website – remember the controversial shift from yellow to green?
But the utility of polls goes beyond gaining customer insight.
Polls can be used to compliment your visitor’s experience and make your content much more engaging, thereby dramatically increasing participation and response rates!
As they say, sometimes it is better to give than to receive, right?
But why bother explaining when I can just show you?
Let’s use the chatbot’s simple-to-use polling functionality to illustrate how naturally capable a simple, humble, down-to-earth bot can handle polling and compliment your visitor’s experience.
Let’s say you own a math tutoring website, and set up your bot to periodically appear with entertaining (although admittedly corny) math jokes during instruction.
You also use your bot to present math problems to keep the readers engaged in the instruction.
Ok, great job completing section two!
Try answering this short problem to test your knowledge
What is the positive square root of 64, minus negative 2?
- a) 8
- b) 6
- c) 10
You choose choice b).
- a) 8 (incorrect) (10%)
- b) 6 (incorrect) (30%)
- c) 10 (correct) (60%)
The positive square root of 64 is 8
And, 8 – (-2) is the same as 8+2=10
Remember, two negatives make a positive! The enemy of my enemy is my friend.
Better luck next time.
So, your bot engaged the student by presenting an extra math problem.
Your bot then not only shared the correct answer, it gave the student a sense as to how other students answered the same question, and then explained how it arrived at the correct answer.
And gave memory-enhancing association!
Chatbots, we love you!
Now, imagine your reader is on your blog about politics.
About 2 minutes into the blog, and as he scrolls to the section about former Exxon CEO, Rex Tillerson, your bot is triggered to appear.
Crazy week in politics, huh?
So many of our readers are intrigued by Trump’s pick of Rex Tillerson for Secretary of State.
Let me know how you feel and I will share the poll results in my next blog entry!
- a) Love this pick
- b) Hate this pick
- c) Hesitant, yet will wait and see
Your faithful reader loves your political blog, but he’s not the type to love or hate without giving them a chance, so he chooses choice c).
As soon as she makes her choice, the chatbot reveals that, to date, readers stacked up as follows: ‘Love this pick’ (60 readers, 30%), ‘Hate this pick’ (80 readers, 40%), and ‘Hesitant, yet will wait and see’ (60 readers, 30%).
You find this information intriguing because you were always curious about this blog’s readership profile.
Also, the fact that this particular blog entry was posted just 2 hours ago and already 200 readers have voted makes you feel like part of an active community!
The following week your faithful reader makes sure to take the time to read the next blog entry because he is curious as to what the final poll result break down was.
Chatbots not only enhanced the content of the blog you were writing, they also enticed your reader to come back the following week for the final poll results!
I would also dare say that most readers would not have even participated in the poll if it was presented using current day poll tools and methods.
In contrast, the chatbot was timed to appear at the perfect time and place, warmed up the participant to the poll with small talk, and then offered real value to the participant by sharing both immediate results and the promise of additional results in the next blog entry.
Polls, quizzes and surveys are everywhere, but most are ignored just like run-of-the-mill pop ups – because they lack charm, warm-up, and context.
Use a chatbot to introduce these factors, and watch your humble chatbot snowball start a veritable avalanche in poll responses!
Chatbots can also rock your share button world!
There is no question that number of shares has become synonymous with popularity and value.
So yes, including share buttons to facilitate sharing of your links or materials is key.
However, wallflower share buttons rarely see any action.
Sometimes all it a takes is a bit of warm-up and the right gentle encouragement to prompt a visitor to share.
And just like the dance situation with the cheerleader discussed above, each situation is unique and may demand a slightly different share prompt.
In my opinion, the best time to invite your visitor to share your content is after you have done something to impress, and chatbots create a universe of such opportunities.
For example, in the blog scenario above, right after your reader responds to your poll regarding Tillerson, your bot could say:
Thanks for participating, please get your friends to participate by sharing (and show share buttons).
Let’s say that you own a barbeque equipment shop.
Because of the power of your chatbot’s personal engagement, visitors who arrive to check out your smokin’ spice rub might just end up leaving with a $900 smoking range!
Recognizing that your customers might want to broadcast their purchase ahead of their barbeque parties, your bot could appear with:
Congrats on your range!
Do I smell a BBQ party on the horizon?
Let your friends know by clicking here:
One could even present a slightly more aggressive share suggestion.
Consider this alternative chatbot wrap-up script:
Congrats on your range!
Do I smell a BBQ party on the horizon?
Let your friends know by clicking here and get 20% off your next purchase!
This level of assertiveness may not be consistent with your brand.
The beauty of chatbots, however, is that you can craft your approach with the precision of a surgeon’s scalpel.
Gone are the days of the one-size-fits-all poll and share buttons tools!