10 ways to make your chatbot strategy a success

When paired with knowledgeable internal human teams, like sales or customer service people, the AI-human combo can deliver high value to customers at any time, in any place.
By
Steve Desjarlais
July 2, 2019

The first generation of chatbots in 2016 were shiny objects at best. In 2020, conversational AI assistants and chatbots have become much more pivotal: they are now the heart and soul of a solid and sustainable digital transformation strategy that puts AI front and centre. When paired with knowledgeable internal human teams, like sales or customer service people, the AI-human combo can deliver high value to customers at any time, in any place.

From marketing to ecommerce to customer service, conversational AI is ushering in a new age of 1:1 communication that can unlock long lasting relationships that deliver value and the tailored experience mobile customers are craving for. There are many paths you can go down when building a conversational strategy, and it’s important to make the best decisions for your brand from day one. Here are 10 of our best tips to get your conversational AI strategy off to the best start:

1. Define your personality, tone and manner

As we increasingly shift towards text and voice-based interfaces, the idea of what makes a brand is shifting too. As customers increasingly connect with businesses via messaging apps like Messenger and WhatsApp, brands are moving away from catchy slogan and visuals in favour of developing a more holistic brand voice with a distinct personality that can easily adapt to the new generation of interfaces that focus on natural language.


In the future, the role of Chief Marketing Officers will be to act as a relationship designer — to humanize brands and give them a voice, literally. To create customer-centric experiences that feel more human, they will inevitably have to zero in on personalization, which is becoming the key to commercial success.

AI assistants will increasingly become the brand’s virtual ambassador, embodying it from digital head-to-toe. Everywhere you look, chatbots are becoming be a user’s first point of contact with the company, that’s why they need to be as relatable and universal as the people they’re trying to reach.

As these virtual assistants become the extension of your brand, it’s important to ask yourself “What kind of friend is your brand?” It may seem like a trivial question, but AI-powered chatbots become like the friend you go shopping with, and having a clear image of who that person is will dictate everything, from its conversational tone to the type of service you’ll provide.

The next crucial question to ask yourself is “Would you be friends with your brand?” Unfortunately, the answer is often no. The reason for this is that most brands make interactions about themselves instead of creating a dialogue that focuses on their customers’ needs and wants.

Like with real-life friends, nobody is looking to hang out with someone who talks about themselves for an hour without checking in to see how you’re doing. In the conversational world, brands need to become that altruistic friend who asks questions and is interested in the answer. If your brand feels like the annoying stranger crashing the party rather than a best friend, it means you still have work to do in recalibrating the relationship. In a messaging-first world, conversations will become the atomic unit of customer relationships.

2. Identify the right use case and Key Performance Indicator (KPI)

As with most business endeavours, it’s smart to start small and scale from there. In the world of AI, agility isn’t just another buzzword: Being able to adapt to ever-changing consumer needs is a core tenet of any worthy artificial intelligence and machine learning strategy. The important thing to keep in mind when it comes to conversational AI is that it doesn’t fit in the set-it-and-forget-it category. Much like your customers, it’s is an ever-evolving, learning and growing organism, not a project that is set in stone. It’s iterative by design.

When starting to design your chatbot, it’s important to assess your organization’s most pressing needs to identify its low-hanging fruit in terms of automation. For example, when talking to ecommerce business owners about holiday sales, we are often told that up to 80% of customer service questions amount to “where is my order?” Given the straightforwardness of this question, it’s a perfect use case to get started with AI automation in a customer service context. Plus, it’s easily quantifiable and therefore measurable (that’s what you want to go for to be able to quickly track and assess ROI and build an internal coalition around your AI project!).


When selecting the right use case to kickstart your conversational strategy, you should be keeping the following elements in mind:

A) You are starting with a clear objective with defined parameters
B) Implementation is fast and affordable
C) It won’t disrupt the core of your business
D) Results are quantifiable and measurable

In order to track visibly measurable quick wins, your business needs to have the right KPI. Once you’ve gotten things moving and can measure your endeavour’s success, you can start expanding your use case and scale within your organization. AI is the new disruption and, as with all change, more organizations are reluctant to jump onboard at first. As an agent of change working to future-proof your business, it’s well-advised to start small.

Before impacting the core of your business, you should focus on proving your assumptions with results and then using this first case study to expand your use case. It’s important that you find a tech vendor and platform that understands your business context and can align with these strategic imperatives.

3. Pick the right channel for your use case and audience

Sometimes, it’s not about what you’re selling, but more about where you’re selling it. To figure out the place that’s best for your company, you need to understand who your target audience is and engage with them on their favourite platforms. Sephora and H&M successfully took up this model in the early days of the chatbot revolution in 2016 by being early entries into Kik’s bot shop, which was a savvy decision because the messaging platform was a go-to for tweens and Gen Z shoppers. Campbell’s Soup made a similar move via Amazon Alexa, knowing that a large part of their base is made up of busy, cooking parents looking for voice-activated commands while their hands are busy chopping carrots.

To effectively choose the right outlet for your brand, you need to be keenly aware of all its facets and be openly able to ask yourself what its “place-ona” is.

Derived from the concept of “persona,” Canadian computer scientist and researcher Bill Buxton introduced the idea of the “place-ona” to capture how location can influence which types of interactive technology it makes sense for people to use.

Let’s look at these three scenarios:

  1. On a bicycle: Your place-ona is “hands busy, eyes busy, ears free, voice free.”
  2. In the classroom: Your place-ona is “hands free, eyes and ears free, voice restricted.”
  3. At home with friends: Your place-ona is “hands free, eyes and ears free, voice free.”

In the first scenario, the platform of choice would be voice. In the second scenario, text would be more appropriate. In the third scenario, either voice or text would work. That’s a key point to make: Bots should not compete against each other, but work together to create a unified, seamless, delightful user experience that creates value for whichever context a user might be in. Figuring out the best-suited platform for your brand will put you in direct contact with the people you’re actually trying to reach, which cuts down on a lot of wasted time.

4. Focus on value creation and relationship building

In the ecommerce world, you should actually prioritize value creation instead of sales themselves. This might feel counter-intuitive, but what you’re trying to do is build long-lasting relationships with customers. To do so, you have to put your brand’s helpfulness first. In order to increase conversion, you’ll want to engage in useful and pleasant conversations, without pushing the hard sale. We can’t stress this enough: Rather than just a company, you need to think of your brand as a virtual friend and, like with all friendships, time and relationship-building are key. Though it may seem like a circuitous approach, it’s actually the best way to achieve brand preference and maximize customer loyalty, not to mention your revenue.

Rather than being pushy with products, good strategies are ones that inspire users with personalized content, ask questions to create a more human bond or share videos and tutorials with users. Then, once customers are fully engaged and have zeroed in on what they truly want, you can smoothly push the sale.

5. Have a clear data mining and data governance strategy

AI consumes data for breakfast, lunch and dinner — it’s the supreme fuel that keeps the whole engine running and continuously built-on with new information. The first and possibly most important step in a viable AI strategy is to spark a great conversation with the user by asking them engaging questions. After getting to know them and what they’re looking for, you can leverage that conversational data to paint a comprehensive portrait of your customer, all the while providing fuel for your AI-powered personalization strategy.


However, the data collections must always come first, so think about your data mining strategy and architecture from day one. This organization from the get-go will focus your efforts in the right places: creating long-term value for you and your customers in a more egalitarian relationship.

To sum up, here are the three steps to create a viable Conversational AI strategy:

  1. Open up customer communication channels
  2. Track conversations and gather valuable declared data
  3. Train your AI on real-life datasets

P.S. Obviously, you need to mine data in an “opt-in” fashion to respect regulations and be GDPR-compliant for example.

6. Enable discovery with a strong digital strategy

Click-to-Messenger ads that drive directly to chatbots, like the ones available on Facebook, can start up a conversational flow between your business and a user that will guide them down a funnel (whether that funnel is acquisitional or educational remains to be seen). While you may not convert at each exchange, you’re still gathering precious data with every interaction, which is why your entire ecosystem should be geared towards engaging conversations.

You obviously want your chatbot strategy to be a success and, for it to perform well, it needs to be tied to a more holistic digital strategy that is at the very centre of your brand. In marketing terms, you should think of chatbots as conversational landing pages, the first thing a user will come into contact with when interacting with your company, meaning that the stakes are high.


Within this framework, the chatbot takes on the great responsibility of being an ad destination meant to qualify prospective customers and lead them down the acquisition funnel. For this tactic to be effective, it’s essential that you have a well-designed digital strategy with touchpoints seamlessly interwoven throughout the whole experience.

As there are many types of shoppers, there are also many ways for them to discover a Messenger chatbot, in both organic and paid-for ways:

  1. Organically: by clicking the “ Send Message” button on the company’s Facebook page
  2. Organically: You can add a widget on your website and a “Send to Messenger” button on your checkout page
  3. Organically: Users can click on the “Discover Tab” in Messenger to find bots within the platform
  4. Organically: By clicking on a brand’s micro URL that leads directly to a chatbot (m.me/brandname)
  5. Organically: Users might come across Messenger QR codes on out-of-home advertising or packaging, for example
  6. Paid-for: Companies can buy click-to-messenger Facebook ads


Even beyond these modes, there are many other ways to drive traffic to your chatbot. A chatbot is essentially a micro URL and can be passed along in a call-to-action button. Because of this format, emails and display banners can also drive to a Messenger chatbot. A text message with the link or QR code that you can snap on a physical product or image can also launch a chatbot. In the case of our client MAKE UP FOR EVER, we even leveraged their influencer roster with a referral link on Instagram to drive to chatbot.


No matter what your tactics and touchpoints are, the main point is to make conversation the core of your customer engagement strategy. Otherwise, chatbots will only be shiny objects with limited reach and impact, and not much to say.

7. Leverage omnichannel and micro-moments

Unlike Vegas, what happens online doesn’t always stay online. Chatbots are in the perfect position to act as a bridge between a website or app and a brick-and-mortar store.

For example, our client Decathlon implemented a function to sign up new memberships within their chatbot, but also drove customers to the opening of their flagship store. As it turns out, 10% of chatbot members made an in-store purchase during the first week of opening and the company saw a 41% conversion rate from the initiative. All of this was done within a strategy of planning for the long-run by fostering loyalty and excitement with their new and returning customers.

Click-and-collect functionality – where shoppers can make their item selections online and then pick them up either in-store or at a centralized collection point – is also one of the best omnichannel strategies we’ve seen when it comes to converting chatbot users into in-store foot traffic. In the end, it’s all about having a unified omnichannel strategy that puts mobile at the front and center and makes your brand readily available across channels.

8. Keep humans in the loop

AI should be used to augment your team rather than replace it. It’s important to remember that conversational AI is still in its early stages, it can’t effectively answer everything and still needs a helping hand from people. Also, a human touch is still necessary, as some customer questions might require a person’s input.

Though customers appreciate self-service to make their shopping experience more efficient, and employees love to eliminate redundancies in their work, there’s still a level of service that you can only get from a real conversation.

The chatbot’s main goal is to help capture leads and automate the basic FAQs to free up the human team to do the more knowledge-based work and allow them to step in at the right moment when the sale is near. They might not wear capes, but customers will surely see them as sales heroes.

When trusted sales teams and chatbots are paired together, you get a virtuous circle in which the AI will constantly learn and perfect itself by working alongside your best brand ambassadors, who will in turn be able to boost their productivity thanks to time freed up by the chatbot. With that extra chunk of time, the people in your company can zero in on exercising their emotional intelligence, using their product expertise and providing a human touch, leaving the robot work to the robots.

9. Nurture relationships and re-engage over time

Social media is clearly taking a turn towards messaging and, if you don’t want to take our word for it, just look at how Mark Zuckerberg is betting the future of Facebook on Messenger, Instagram and Whatsapp. There was a time not so long ago when community managers would create content calendars to engage their page’s fans, but those days are over. In the new world, AI assistants will act as the community managers that can engage with each individual customer on a personal level, with content and deals curated precisely to their preferences. In this way, social media’s shift towards messaging is providing everyone with the one-on-one attention they crave.

The key to a successful AI-powered chatbot is to play the long game and understand that these are sustainable tools rather than shiny new toys that will quickly be forgotten in favour of the next thing. Done right, a well-designed chatbot has the power to be a game-changer for your company, and that’s because its true power lies in building solid relationships with users, which certainly won’t happen in a day.

In the long run, the goal is not only to treat customers like V.I.P.s, but to start anticipating their needs thanks to the precious data you’ve been collecting throughout the journey. This end-goal is why it’s key to ask the right questions and engage clients with a long-term strategy so that you can unlock the full value of personalization over time.

10. Pick the right partner

Teaming up with the right people becomes essential when designing and kickstarting your conversational AI strategy as it will play a pivotal role in your entire business strategy in terms of sales, marketing and support in the years to come.

AI is disrupting just about every industry and it is evolving fast. Picking the wrong vendor could delay your go-to market strategy and reduce or annihilate your competitive edge in the space.


That said, the best way to get started is to go for it, but they key is to start small. You should move fast but responsibly. Before settling on a vendor, you can commission a proof of concept to quickly find out if they’re the real deal or just “vapourware.”

If you’re curious to learn more about Heyday’s enterprise-grade Conversational AI solution or need assistance in crafting your conversational strategy, you can book your free consultation session today. We’ll help get you on track, and let you steer the conversation.