E-commerce Chatbots: Five Examples That Increase Sales and Revenue.

Picture this. It’s 11 PM on a Tuesday night.

Having just remembered that it’s your nephew’s birthday party on Friday, you’re sitting at your laptop wondering what the hell seven-year-old boys like.

And you’re feeling mildly panicked.

With no gift and no time to go shopping, you turn to the Facebook page of a brand you know kids love for inspiration. But the number of options available are enough to terrify even the most confident of decision-makers.

You need help.

And you find it by clicking on the ‘Gift Suggestion’ button, where you’re directed to Facebook Messenger.

Here you meet a super-friendly, emoji-loving chatbot who knows all the right questions to ask about the birthday boy — like his age and specifics about the kind of things he’s interested in. And your budget, of course.

The Facebook chatbot’s cogs start turning and it offers up several personalized suggestions based on your answers. A couple of clicks later, the birthday gift has been ordered and is due to arrive before you leave for work on Friday.

A success for everyone involved.

Why?

• You got the help you needed when you needed it, long after the customer service team had left for home.
• The brand not only made a sale using targeted questions, but it also found itself a new, very satisfied customer who was highly likely to return to the brand in the future.

But this isn’t just a made-up example of how an ecommerce chatbot could work. This is the exact situation I found myself in about two months ago, and the messenger bot in question who came to my rescue was LEGO’s very own ‘Ralph’.

What started as a device to help holiday shoppers with their gift buying, Ralph soon became a full-time member of staff after driving over 25% of the social media sales. But that’s not all. He also enjoyed a conversion rate that was 8.4 times higher other Facebook Ads and helped LEGO reduce their cost per conversion by 31% (vs. other conversion-based ads).

So, Chatbots… Tell Me More.

At its most basic, a chatbot can be defined as: ‘A computer program designed to simulate a conversation with human users, especially over the Internet.’

Chatbots work by running the text you write through an algorithm to figure out what you want, before responding with an answer that’s based on your keywords.

They can be used for pretty much everything — booking a train ticket, ordering a takeaway, getting help with a clothes delivery, and much more. And they can be found within most existing messaging sites, like Facebook Messanger, KIK, Line, Youtube or Instagram, and social commerce platforms.

There are two main levels of intelligence when it comes to chatbots:

1. Smart Chatbots

These are the chatbots that use artificial intelligence (AI) to communicate. Users can type or speak freely into the platform, and the bot responds accordingly. They understand questions and commands (to a certain degree) and over time, the more the AI chatbot is used, the more ‘intelligent’ it becomes. However, this knowledge is usually only limited to a certain area.

Example — Mitsuku

A four-time winner of the Loebner Prize, chatbots don’t get much smarter than Mitsuku. This ‘18-year-old from Leeds, UK’ is full of jokes, general chit-chat and even opinions about current affairs. Her responses are witty, cheeky and often pretty well-informed… for a robot.

2. Scripted Chatbots

These are chatbots that follow a predetermined path. They usually only respond to specific commands or require users to answer multiple choice questions in order to move on to the next stage of the conversation. This means there are a limited number of outcomes of the conversation.

Example — Tech Crunch

The chatbot from digital publisher Tech Crunch helps users wade through the magnitude of stories it publishes daily to find the content they actually want to read. It asks targetted questions and users just need to click the option they’re interested in, and even specify how often they want to receive stories.

By putting the reader in the driving seat, it increases the positive brand association they’ll have with the brand and the level of engaged traffic to the website.

But Why do Businesses Need Chatbots?

The ways in which AI chatbots can help businesses are endless. First and foremost, they’re highly effective in increasing productivity — which means lower overheads and higher profits.

And come on, who doesn’t want to increase their sales and revenue?

But that’s just the tip of the iceberg:

1. Time-Saving

With chatbots, there’s no hanging around. They can give users answers to their questions or complete request in no time. This means that more people are being helped quicker and more efficiently.

2. Money-Saving

While chatbots do require some initial investment, they’ll save a business a lot of money in the long run. As the system is automated, it cuts down on staffing costs — robots never get tired or need a vacation. Plus they can handle thousands of queries at once.

3. Accessible Anytime

Don’t you just hate being put on hold while waiting on an operator to help you with an inquiry? Or waiting days for a company to respond over email? Chatbots slash this waiting time as they can give users answers immediately — during work hours or even at 3 AM at night. 24/7. 365 days a year. Also, by providing instant assistance, they can increase the likelihood of a sale if customers have a question about a product.

4. Reduces Errors

There is no human error with chatbots as there are no humans involved. Once programmed properly, they’ll always give users the right answers. So this cuts out the worry of things going wrong and keeps customer relationships smooth. However, there is a learning curve. And you should expect a few issues when setting up new chatbots and chat related solutions.

5. No ‘Bad Days’

Sometimes people get out of the bed on the wrong side. That’s just life. But not with chatbots. With them, you’re guaranteed 365 days of politeness and in some cases, light banter. This puts customers at ease and makes their experience more satisfying.

6. Build Trust

As 53% of people who currently use messaging apps to interact with brands are more likely to shop with a business they can message, it would be silly not to consider bringing it into your marketing strategy.

Five of the Best Chatbot Examples

While there are many Instagram, KiK, Facebook and Youtube chatbots out there, here are some examples that have seen a great ROI.

Marvel

Chatbot type — Fixed response

Chatbot goal — Turn engagements into direct ticket sales

To boost tickets sales for both Avengers: Infinity War and Ant-Man, Marvel wanted to try something new and bring the box office to where fans hung out: social media.

And they enlisted Jumper.ai to help.

Using personality-packed images and gifs to build hype and short persuasive copy, they prompted users to comment on the post using the hashtag mentioned in the post.

Here they were directed to a DM checkout chatbot who seamlessly took them through the sales process and the tickets were booked in no time.

The campaign was a huge success, both organically and with paid support, and used clever retargeting to engage previous users. Best of all, the lowest conversion rate from hashtag engagement to tickets purchased was a staggering 50%.

Key Takeaways:

• Engaging imagery and gifs made the brand personality shine.
• Using a DM chatbot checkout made it super easy for fans to buy their tickets instantly.
• Smart targeting and retargeting was used to engage users who initially failed to convert.

Patrón Tequila

Chatbot type — Fixed Response

Chatbot goal — Offer personalized drink suggestions with the brand at the core

An extension of the Cocktail Lab experience on their website, Patrón launched the Bot-tender (robot bartender — geddit!?) to inspire people to get more creative with Patrón over the summer.

To access a wealth of delicious cocktail recipes to suit their taste, users simply had to click on the button that reflected the vibe they were looking for on the Twitter post. They were then directed to a Twitter chatbot, where ‘mixologists’ Stephen and David were able to make tasty cocktail suggestions based on the answers the user gave to their questions.

Thanks to the campaign, Patrón exchanged over 120K messages with fans and made thousands of personalized drink recipe recommendations. And in addition to this, they were able to use the captured data to inform future campaigns and design new recipes.

Key Takeaways:

• The brand was the key component of every suggested drink.
• The campaign capitalized on the popularity of mixed drinks at summer events.
• The data collected from fan responses will be extremely helpful in creating future advertising campaigns.

Etsy

Chatbot type — Fixed Response

Chatbot goal — Offer personalized 24/7 customer service

While most of our other chatbots have been quite fun and playful, this example focuses more on the functional side of the technology.

Quality customer service is the name of the game here, and it’s something that Etsy has nailed with its Twitter DM offering. With the use of clear call-to-action buttons, users can resolve issues or find out more information in mere minutes.

As chatbots aren’t restricted to traditional working hours, Etsy can be there to help customers day or night, and fix any problems immediately. This increases the likelihood of the customer retaining a positive relationship with the brand, despite any issues that may have arisen.

Key Takeaways:

• By putting the most popular queries at the forefront of the chatbot, the brand made it extremely easy for customers to find instant solutions.
• Offering immediate solutions has helped the brand build trust with customers, and has encouraged repeat sales.
• As the chatbot is able to deal with thousands of queries at once, the brand has saved a huge amount of time and money on customer service resources.

Ben & Jerry’s

Chatbot type — Fixed Response

Chatbot goal — Offer free samples in a hugely engaging way

To launch their new Pint Slices and encourage conversion-led sampling, Ben & Jerry’s took to social media. Using Click-to Facebook Ads and Instagram stories, they built hype around the campaign and gave out QR codes to fans to claim their free sample at local vending machines.

With the help of jumper.ai, they streamlined this process by introducing a chatbot. The bot not only engaged with fans in a friendly, personalized way, it was also used to capture data.

The campaign was a huge hit. There was 5 times more engagement on social over this period with over 13,000 people interacting with the brand. All 5,000 of the free slices were snapped up (unsurprisingly, because well who doesn’t love ice-cream) and sales over delivered by 20%.

Key Takeaways:

• Using a variety of different social media platforms meant that the brand could engage with customers no matter where they hung out online.
• Paid media posts allowed the brand to connect with people who weren’t fans, and encourage them to sample the product.
• Offering a QR code to collect a free sample meant that the brand could capture lots of data that will allow them to target new fans in the future.

H&M

Chatbot type — Fixed Response

Chatbot goal — Turn personalized style suggestions into sales

While Kik may not be as popular as Facebook Messanger or Whatsapp when it comes to messaging platforms, it still has an impressive user base of over 300 million people. And over 40% of those are US teenagers. So this made it the perfect space for H&M to connect with its customers via a KiK chatbot.

Acting as a personal digital stylist, the chatbots asks the user to answer a series of multiple-choice questions to determine their preferences, before making bespoke recommendations. This could be an entire outfit or the latest trending pieces. There’s also an option to see outfits that other bot users have created for inspiration and to then vote on them.

By offering this experience via a chatbot, shoppers can easily and almost instantly find the clothes they’re looking for without having to wade through all the stock online. The brand also benefits enormously from the exchange via insights about the customer.

Key Takeaways:

• Multiple choice style questions enabled the brand to make really personalized suggestions to fans.
• This style of questioning also saved the customer a lot of time wading through the thousands of garments available on the website.
• The brand now has the knowledge they need to retarget the shopper in the future with specific, highly-relevant recommendations.

And one last bonus example… Casper

Chatbot type — AI Response

Chatbot goal — To increase brand affiliation and entertain

Launched a few years ago, mattress company Casper created an SMS chatbot to keep insomnia sufferers company on nights they couldn’t sleep (from 11 PM — 5 AM). Users simply messaged the number and the Insomnobot-3000 would reply and chat with them like a friend, in a light-hearted and sometimes tongue-in-cheek way.

While the goal of the chatbot wasn’t to increase sales and revenue (the whole point of this blog)… it is a useful, if strange, example of how brands can use chatbots to connect with potential customers in relevant spaces, in a new, innovative way.

Key Takeaways:

• While this chatbot doesn’t directly sell to people, it brings them closer to the brand.
• The brand conducted a lot of research to provide relevant responses to the things that keep people awake at night — work, family pressures or questions about the universe.
• Interacting with the chatbot via phone offered the company a way to contact users about future promotions.

What Does the Future Hold for Chatbots?

Year on year, chatbots are getting more and more intelligent.

Jumper users are already leveraging the power of automated chatbot checkouts to drive more sales. And you can join them by setting up a free account by clicking here.

While chat solutions have come a long way since the world’s first ever chatbot ELIZA in 1966, there’s yet to be one that’s so clever, you can’t tell whether you’re speaking to a human or a machine.

So to this day, the Turing Test (the holy grail of the industry) remains unbeaten. But there are a lot of incredible minds still working on it, so watch this space…

Originally published at https://blog.jumper.ai on April 23, 2019.