We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 – August 3. Join AI and data leaders for insightful talks and exciting networking opportunities. Learn more about Transform 2022
The market for engaging chatbots continues to heat up as consumers increasingly head online for purchases, creating a need for enterprises to offer a differentiated customer experience.
But good, differentiated chatbots are hard to build. With the influx of users and heightened demand, companies have had to turn to any number of places to get started with chatbot technologies. And those places can either be brittle or dominated by players that don’t offer much room for customization. For instance, companies might leverage the virtual assistants offered by Google Assistant, Amazon Alexa and Facebook Messenger. They also might turn to the offerings provided by chatbot providers like IBM Watson, Five9 or Gupshup.
Chatbots need to have it all
These bots are blossoming in functionality as well as in other essential areas such as business sales, analytics, customer service and overall end-user satisfaction. But the architectures of these offerings can sometimes be limiting. Some of them are one-stop shops that force companies to build the backend and storage within the same system – seen in offerings like Intercom and Drift. Or there’s Zendesk’s, which is popular among companies that use Zendesk customer support, but which don’t have open end-points to other ticketing systems or live chat services.
After several years of shake-out, one technology approach is becoming increasingly popular. Many companies are leveraging open source backend technologies, like natural language understanding (NLU) from Dialogflow and Google Cloud Platform, for example. Then they’re adding feature-rich, modularized, frontend tools on top. These tools empower enterprises to integrate, deploy and train AI chat and messaging into their own custom web and mobile experiences.
One new offering seeking to fill a void here is Botcopy, a cloud-compatible custom user interface for web messaging systems, which has just secured late-seed funding.
“What Botcopy customers tend to have in common is they believe there’s value in controlling and managing their own cloud platform services. We couldn’t agree more,” Botcopy’s Head of Product Alexander Seegers said. The challenge, he explained, is that some of these organizations don’t want to compromise on having a best-in-class chat solution for their websites and mobile apps. “So from the beginning, our focus has been on making that frontend piece easy to create and manage.”
Other examples of such offerings include FeathersJS and TalkJS — both of which require coding expertise.
AI chatbots are typically used to guide conversations along the best path. One way they do this is by collecting data and making suggestions to the end-user – a tight machine learning feedback loop that gets better with time. For example, let’s say the chatbot learns that new buyers are most likely to ask about payment options when they are looking at a product page. Once the chatbot has that insight, it can optimize the page by moving payment options, like PayPal, closer to the top. Ultimately, it boils down to lining up users’ preferences with what’s on offer.
The statistics surrounding AI messaging only help to further cement its place in the modern ecosystem. Chatbot adoption is estimated to save the healthcare, retail and banking sections alone upwards of $11 billion annually. Other predictions on the value-driven growth of AI-assisted virtual messaging indicate that 50% of all knowledge workers will utilize some form of a chatbot daily.
Modular vs. custom solutions
As the potential for disruption from AI-assisted messaging increases, so, too, do its methods of implementation and deployment across websites, mobile applications and services. Blending the backend AI services of GCP, Azure and AWS, with modular frontend components, has also led to greater flexibility of innovation from companies other than Facebook, Google, Microsoft and Amazon.
The SaaS market has a rich history of counter-balancing the need for speed to market on one hand, with feature-rich customization on the other. In this way, fast-to-market solutions and inexpensive user licenses often battle with the high knowledge and expertise demands that it takes to retrofit broad parts around specific business processes.