Wednesday, April 11, 2018

The Bot Plan: Your Guide to Making Conversations Convert

Posted by purna_v

Let’s start off with a quick “True or False?” game:

“By 2020, the average person will have more conversations with their bot than with their spouse.”

True, or false? You may be surprised to learn that speaking more with bots than our spouse is precisely what Gartner is predicting.

And when Facebook’s Mark Zuckerberg says “messaging is one of the few things that people do more than social networking,” it requires no leap of faith to see that chatbots are an integral part of marketing’s future.

But you don’t need to stock up on canned peaches and head for the hills because “the robots are coming.” The truth is, the robots aren’t coming because they’re already here, and they love us from the bottom of their little AI-powered hearts.

Bots aren’t a new thing for many parts of the world such as China or India. As reported by Business Insider, sixty-seven percent of consumers worldwide have used a chatbot for customer support in the last year.

Within the United States, an impressive 60% of millennials have used chatbots with 70% of those reporting positive experiences, according to Forbes.

There’s no putting bots back in the box.

And it’s not just that brands have to jump on board to keep up with those pesky new generations, either. Bots are great for them, too.

Bots offer companies:

  1. A revolutionary way to reach consumers. For the first time in history, brands of any size can reach consumers on a personal level. Note my emphasis on “of any size.” You can be a company of one and your bot army can give your customers a highly personal experience. Bots are democratizing business!
  2. Snackable data. This “one-to-one” communication gives you personal insights and specificity, plus a whole feast of snackable data that is actionable.
  3. Non-robot-like interaction. An intelligent bot can keep up with back-and-forth customer messages in a natural, contextual, human way.
  4. Savings. According to Juniper Research, the average time saving per chatbot inquiry compared to traditional call centers is over four minutes, which has the potential to make a truly extraordinary impact on a company’s bottom line (not to mention the immeasurable impact it has on customers’ feelings about the company).
  5. Always on. It doesn’t matter what time zone your customer is in. Bots don’t need to sleep, or take breaks. Your company can always be accessible via your friendly bot.

Here in the West, we are still in the equivalent of the Jurassic Period for bots. What they can be used for is truly limited only by our imagination.

One of my most recent favorites is an innovation from the BBC News Labs and Visual Journalism teams, who have launched a bot-builder app designed to, per Nieman Lab, “make it as easy as possible for reporters to build chatbots and insert them in their stories.”

So, in a story about President Trump from earlier this year, you see this:

Source: BBC.com

It’s one of my favorites not just because it’s innovative and impressive, but because it neatly illustrates how bots can add to and improve our lives… not steal our jobs.

Don’t be a dinosaur

A staggering eighty percent of brands will use chatbots for customer interactions by 2020, according to research. That means that if you don’t want to get left behind, you need to join the bot arms race right now.

“But where do I start?” you wonder.

I’m happy you asked that. Building a bot may seem like an endeavor that requires lots of tech savvy, but it’s surprisingly low-risk to get started.

Many websites allow you to build bots for free, and then there’s QNAMaker.ai (created by Microsoft, my employer), which does a lot of the work for you.

You simply input your company’s FAQ section, and it builds the foundation for an easy chatbot that can be taken live via almost any platform, using natural language processing to parse your FAQ and develop a list of questions your customers are likely to ask.

This is just the beginning — the potential for bots is wow-tastic.

That’s what I’m going to show you today — how you can harness bot-power to build strong, lasting relationships with your customers.

Your 3-step plan to make conversations convert

Step 1: Find the right place to start

The first step isn’t to build a bot straightaway. After all, you can build the world’s most elaborate bot and it is worth exactly nothing to you or your customer if it does not address their needs.

That’s why the first step is figuring out the ways bots can be most helpful to your customers. You need to find their pain points.

You can do this by pretending you’re one of your customers, and navigating through your purchase funnel. Or better again, find data within your CRM system and analytics tools that can help you answer key questions about how your audience interacts with your business.

Here’s a handy checklist of questions you should get answers to during this research phase:

  • How do customers get information or seek help from your company? ☑
  • How do they make a purchase? ☑
  • Do pain points differ across channels and devices? ☑
  • How can we reduce the number of steps in each interaction? ☑

Next, you’ll want to build your hypothesis. And here’s a template to help you do just that:

I believe [type of person] needs to solve [problem] which happens while [situation], which will allow them to [get value].

For example, you’re the manager of a small spa, whose biggest time-suck is people calling to ask simple questions, meaning other customers are on hold for a long time. If those customers can ask a bot these simple questions, you get three important results:

  1. The hold time for customers overall will diminish
  2. The customer-facing staff in your spa will be able to pay more attention to clients who are physically in front of them
  3. Customers with lengthier questions will be helped sooner

Everybody wins.

Finally, now that you’ve identified and prioritized the situations where conversation can help, you’ll be ready to build a bot as well as a skill.

Wait a minute — what’s a skill in this context, and how do they relate to bots? Here’s a great explanation from Chris Messina:

  • A bot is an autonomous program on a network
  • A chatbot is a bot that uses human language to communicate
  • An AI assistant is a chatbot that performs tasks or services for an individual
  • A skill is a capability that an AI assistant can learn

Each of them can help look things up, place orders, solve problems, and make things happen easier, better, and faster.

A few handy resources to build a bot are:

Step 2: Add conversation across the entire customer journey

There are three distinct areas of the customer decision journey where bots and skills can make a big difference.

Bot as introducer

Bots can help your company by being present at the very first event in a purchase path.

Adidas did this wonderfully when they designed a chatbot for their female-focused community Studio LDN, to help create an interactive booking process for the free fitness sessions offered. To drive engagement further, as soon as a booking was made the user would receive reminders and messages from influencer fitness instructors.

The chatbot was the only way for people to book these sessions and it worked spectacularly well.

In the first two weeks, 2,000 people signed up to participate, with repeat use at 80%. Retention after week one was 60%, which the brand claims is far better compared to an app.

Adidas did something really clever. They advertised the bot across many of their other channels to help promote the bot and help with its discoverability.

You can do the same.

There are countless examples where bots can put their best suit on and act as the first introduction to your company:

  • Email marketing: According to MailChimp research, the average email open rates are between 15% to 26% with click rates being just a fraction of that at approximately 2%–5%. That’s pretty low when you compare that to Messenger messages, which can have an open rate of well over 90%. Why not make your call-to-action within your email be an incentive for people to engage with your chatbot? For example, something like “message us for 10% off” could be a compelling reason for people to engage with your chatbot.
  • Social media: How about instead of running Facebook ads which direct people to websites, you run an ad connecting people to bots instead? For example, in the ad, advise people to “chat to see the latest styles” or “chat now to get 20% off” and then have your bot start a conversation. Instant engagement! Plus, it’s a more gentle call-to-action as opposed to a hard sell such as “buy now.”
  • Video: How about creating instructional YouTube videos on how to use your bot? Especially helpful since one of the barriers to using this new technology is a lack of awareness about how to use it. A short, quick video that demonstrates what your skill can do could be very impactful. Check out this great example from FitBit and Cortana:

  • Search: As you’ve likely seen by now, Bing has been integrating chatbots within the SERPs itself. You can do a search for bots across different platforms and you’ll be able to add relevant bots directly to your preferred platform right from the search results themselves:

Travel Bots

  • You can engage with local businesses such as restaurants via the Bing Business bot that shows up as part of the local listings:

Monsoon Seattle search with chatbot

The key lesson here is that when your bot is acting as an introducer, give your audience plenty of ways and reasons to chat. Use conversation to tell people about new stuff, and get them to kick off that conversation.

Bot as influencer

To see a bot acting as an effective influencer, let’s turn to Chinese giant Alibaba. They developed a customizable chatbot store concierge that they offer free to brands and markets.

Cutely named dian xiao mi, or “little shop bee,” the concierge is designed to be the most helpful store assistant you could wish for.

For example, if a customer interacting with a clothing brand uploads a photograph of a t-shirt, the bot buzzes in with suggestions of pants to match. Or, if a customer provides his height and weight, the bot can offer suggested sizing. Anyone who has ever shopped online for clothing knows exactly how much pain the latter offering could eliminate.

This helpful style is essentially changing the conversation from “BUY NOW!” to “What do you need right now?”

We should no longer ask: "How should we sell to customers?" The gazillion-dollar question instead is: How can we connect with them?

An interesting thing about this change is that, when you think about it for a second, it seems like common sense. How much more trust would you have for a brand that was only trying to help you? If you bought a red dress, how much more helpful would it be if the brand showed you a pic of complementary heels and asked if you want to “complete the look”?

For the chatbot to be truly helpful as an influencer, it needs to learn from each conversation. It needs to remember what you shared from the last conversation, and use it to shape future conversations.

So, say a chatbot from my favorite shoe store knew all about my shoe addiction (is there a cure? Would I event want to be cured of it?), then it could be more helpful via its remarketing efforts.

Imagine how much more effective it would be if we could have an interaction like this:

Shoestore Chatbot: Hi Purna! We’re launching a new collection of boots. Would you like a sneak peek?

Me: YES please!!!

Shoestore Chatbot: Great! I’ll email pics to you. You can also save 15% off your next order with code “MozBlog”. Hurry, code expires in 24 hours.

Me: *buys all the shoes, obvs*

This is Bot-topia. Your brand is being helpful, not pushy. Your bot is cultivating relationships with your customers, not throwing ads at them.

The key lesson here? For your bot to be a successful influencer, you must always consider how they can be helpful and how they can add value.

Bot as closer

Bot: “A, B, C. Always be closing.”

Imagine you want to buy flowers for Mother’s Day, but you have very little interest in flowers, and when you scroll through the endless options on the website, and then a long checkout form, you just feel overwhelmed.

1-800-Flowers found your pain point, and acted on it by creating a bot for Facebook Messenger.

It asks you whether you want to select a bunch from one of their curated collections, instantly eliminating the choice paralysis that could see consumers leave the website without purchasing anything.

And once you’ve chosen, you can easily complete the checkout process using your phone’s payment system (e.g. Apple Pay) to make checkout a cinch. So easy, and so friction-free.

The result? According to Digiday, within two months of launch the company saw 70% of the orders through the bot came from brand-new customers. By building a bot, 1-800 Flowers slam-dunked their way into the hearts of a whole new, young demographic.

Can you think of a better, more inexpensive way to unlock a big demographic? I can’t.

To quote Mr. Zuckerberg again: “It’s pretty ironic. To order from 1-800-Flowers, you never have to call 1-800-Flowers again.”

Think back to that handy checklist of questions from Step 1, especially this one: “How can we reduce the number of steps in each interaction?”

Your goal is to make every step easy and empathetic.

Think of what people would want/need to know to as they complete their tasks. For example, if you’re looking to transfer money from your bank account, the banking chatbot could save you from overdraft fees if it warns you that your account could be overdrawn before you make the transfer.

The key lesson here: Leverage your bots to remove any friction and make the experience super relevant and empathetic.

Step 3: Measure the conversation with the right metrics

One of my favorite quotes around how we view metrics versus how we should view metrics comes from Automat CEO Andy Mauro, who says:

“Rather than tracking users with pixels and cookies, why not actually engage them, learn about them, and provide value that actually meets their needs?”

Again, this is common sense once you’ve read it. Of course it makes sense to engage our users and provide value that meets their needs!

We can do this because the bots and skills give us information in our customers’ own words.

Here’s a short list of KPIs that you should look at (let’s call it "bot-alytics"):

  • Delivery and open rates: If the bot starts a conversation, did your customer open it?
  • Click rates: If your bot delivered a link in a chat, did your customer click on it?
  • Retention: How often do they come back and chat with you?
  • Top messages: What messages are resonating with your customers more than others?
  • Conversion rates: Do they buy?
  • Sentiment analysis: Do your customers express happiness and enthusiasm in their conversation with the bot, or frustration and anger?

Using bot-alytics, you can easily build up a clear picture of what is working for you, and more importantly, what is working for your customer.

And don’t forget to ask: What can you learn from bot-alytics that can help other channels?

The future's bright, the future's bots

What were once dumb machines are now smart enough that we can engage with them in a very human way. It presents the opportunity of a generation for businesses of all shapes and sizes.

Our customers are beginning to trust bots and digital personal assistants for recommendations, needs, and more. They are the friendly neighborhood machines that the utopian vision of a robotic future presents. They should be available to people anywhere: from any device, in any way.

And if that hasn’t made you pencil in a “we need to talk about bots” meeting with your company, here’s a startling prediction from Accenture. They believe that in five years, more than half of your customers will select your services based on your AI instead of your traditional brand.

In three steps, you can start your journey toward bot-topia and having your conversations convert. What are you waiting for?


Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!