Is there currently a bots backlash happening on company websites all over the world?
Well, there’s certainly no shortage of stories about bad bot experiences. In fact, you’ve probably experienced one yourself. Annoying pop-ups, even more annoying names, and suggestions that are completely off topic. While many companies have been quick to rush in on the cost-saving benefits and efficiency gains of chatbots, some have done so at the detriment of the customer experience. Not good.
Like any application of artificial intelligence (AI), properly deploying and using it are key factors for long-term success. Below we look at the elements of effective chatbot implementations and provide some tips for today’s companies to navigate the new chatbot landscape.
The Human Element
A recent ZDnet article covered the efforts by Facebook’s AI researchers to help chatbots be less repetitive, more accurate, and more emotional. This goes beyond adding a name and face to your bots. By bringing in more data sets and focusing on aspects like tone, mood, and texture, the researchers are expanding the scope of previous work to help chabots better interact with customers. The ultimate conclusion: Chatbots, which can offer big opportunities for improved customer engagement, still have a long way to go before being truly conversational.
Augmented Intelligence for Agents
To that end, a valuable piece of advice comes from Forrester analyst Ian Jacobs, who urges companies to stop trying to replace customer service agents with chatbots. Here at the APEX of Innovation, we couldn’t agree more. Of course, AI will automate some tasks, but the big contribution in the contact center is centered around AI’s ability to help agents, not replace them.
In a recent Forrester blog post, Jacobs reveals consumers’ preference for calling into a contact center or contacting a company over social media versus dealing with a bad chatbot interaction. In fact, it’s when chatbots can’t solve a customer’s problem that they get the most frustrated. At best, they may call your contact center feeling a bit annoyed. At worst, it’s an easy jump to your competitor’s website where you may lose them for good.
So, what can you do? According to Jacobs, below are some useful tips for chatbot implementations:
- Agent-facing chatbots: If your company is just getting started with chatbots, Jacobs recommends using them for internal purposes. They can be a great tool to enable agents to quickly tap into knowledge bases to better serve customers. Since agents are the solution’s first users at your company, you can reduce the risk of a bad experience making its way to the customer.
- Human-intermediated chatbots: Here’s where AI really helps the agent. In this use case, AI monitors conversations between customers and agents and offers up suggestions to improve engagement in the form of augmented intelligence. Today’s AI-powered tools can automate the creation of a specific customer response for the agent to immediately send or allow for the agent to make any edits first and then send it.
- Front-end chatbots to handle routine tasks and free up agents to resolve issues: Collecting all the customer information and context needed before an agent can actually manage a customer interaction presents a huge opportunity for time-savings via automation. This is one area where AI completely taking over tasks makes sense, saving agents valuable time and making for a more efficient customer experience.
- Intermingled workflows so both chatbot and agent do what each does best: Think of this one as a form of agent and chatbot collaboration. Building on the front-end chatbot approach, this use case enables a ‘back and forth’ between agents and bots to complete specific workflows. According to Jacobs, “The agent and chatbot can flex back and forth to tackle the portions of the interaction they excel at.”
There’s little debate: Chatbots are here to stay. When deployed correctly they can benefit both your customers and your agents. Success requires a long-term view and an approach that strikes the right balance between replacing routine tasks and augmenting agents with extra business intelligence.