MENU

Chatbotched

By: Alex Sophocles & Cara Peckens, imre health experience team

Chatbots are the hottest topic in pharma marketing right now—every  brand wants one, but they lack the strategy a productive chatbot demands. However, if you rush to bring your brand’s chatbot to market, you risk squandering your sizable innovation investment on a mediocre, glorified FAQ hub that may be housed within a poor build that is half finished, or even broken. This rush to market for chatbots in healthcare perpetuates the myth that the technology can’t deliver meaningful interactions for users, on par with other digital or social media experiences.

The irony is that in spite of healthcare’s “chatbot frenzy,” technology has finally reached the precipice where chatbots can truly provide value to the user. But to leverage the technology effectively, marketers and technologists must start small, iterate, and adapt to their user base. Chatbots require a continual investment in order to be a useful product; otherwise bots will linger in the land of bad reviews. That’s why before we analyze what makes a good chatbot, we encourage brands to ask: why a chatbot?

“Instead of asking what makes a good chatbot, perhaps we should ask ourselves: why a chatbot?”

If you’re still reading this, perhaps you’ve decided that you’re willing to explore the investment it takes to make a purposeful bot. The difficulty in building a chatbot is not a technical one—it is an issue of user experience. It must be seamless in its availability and add value in its use. Here is a quick litmus test for consideration:

3 Bad Signs

—Is it easier to use Google to get the same information?
—Is the bot serving the same information that's on your website FAQ?
—Is your bot going to be able to do this and this and this and this?

3 Good Signs

—Does it reduce friction to already existing task or process?
—Does it complete a singular task?
—Does it connect you to an actual person at some point in the experience?

If you’ve passed the litmus test then here are some best practices to take into consideration as you continue to solidify your chatbot concept.

1. Users need guidance

When left with open ended questions; users tend to feel lost or drop off entirely. Lead users through the experience by providing sample answers, guided prompts, or quick replies.

2. Have a persona

With any new technology users tend to test its limits. Make sure your bot can handle questions about itself. This a great place to insert some brand personality and entertainment value for your users.

3. Provide a real
human fallback

At a certain threshold users will get frustrated with the experience if the bot can’t answer or understand the user. Make sure your chatbot is programmed to either directly connect the user with a human or have a human follow up with the user within a certain period of time.

4. Monitor, analyze,
and iterate

Chatbots are a living product. Watching real users interact with the bot for even just 5 minutes can shed insights on user paths and dialogue patterns you won't have anticipated. Think of it as a subscription you have to pay for to constantly improve the quality, not a one time deal.

5. Prioritize MLR collaboration

Since every response and conversation path will need to go through legal review and because you want the flexibility to update these paths regularly, establish guardrails ahead of time with your MLR team. You want to translate your review processes in a way that will help you avoid a production bottleneck, when you want to optimize your bot.

Chatbots seem simple in their production state, but the planning and strategy behind them can be monolithic. Put your user first, work backwards, and make sure that a chatbot is the right solution to the problem you are trying to solve. If it is the right solution, and you follow the necessary steps to build it properly, the climate is ripe for a brand to launch a successful bot that the world will notice. Do you want to be the brand that everyone wants to emulate? Start first by asking yourself:
why a chatbot?

How to Build a Chatbot

More Articles

Combating Compassion Fatigue

Pharma and the opioid epidemic

Empathic marketing: Swimming in the deep end

Share

Chatbotched