Simplicity, Scale, Sexiness.
These are the 3 ingredients why I believe Google is on the path of conquering the world of chatbots. It’s very exciting to see how Google managed in just over a year not only to boost the developer community of DialogFlow (formerly known as api.ai) by 4x, but also continues to make it super simple to build chatbots. Google does this by introducing new, fresh ways of making bots that truly enhance the customer experience on all levels.
I got inspired by this session that I saw at Google Next ‘18 a week ago and wanted to share my findings and notes with other chatbot enthusiasts. Here we go.
fail are hot
If you Google for “why chatbots”, the first result you get is why they fail. That is because so far no company has been able to master the skills of making chatbots useful and accessible to a point they can provide people with a real enjoyable customer experience.
That’s about to change really, really fast.
Enterprises like to invest in chatbots to reduce operational cost. That way they can serve customers faster and more efficient. But did you know that also 60% of the customers actually prefer talking to a chatbot for getting things done? Gartner research shows that 50% of enterprises will spend more on chatbots than on mobile development by 2021.
In general, automating conversations using chatbots can be found in 3 main use cases:
- Connecting businesses to their customers around customer service and commerce
- Powering IoT devices like connected cars, TVs etc. – and make it possible to intelligently talk to these devices
- Connecting businesses to their employees so they can get easier access to data and insights, business intelligence and more
Google makes it extremely easy to build chatbots for any platform.
Using DialogFlow, you build the bot once and deploy to any channel like web, Facebook messenger, Google Home and Assist, or even a phone gateway.
DialogFlow handles the natural language understanding and conversation management, and you can use whatever backend to fulfil the user’s request. This flexibility and simplicity that Google brings to the table are key ingredients for me why Google gets a real competitive edge, compared to the competition.
In just over 1 year time, Google managed to scale the developer community of DialogFlow from 150.000 to 600.000.
This is not only growing but also accelerating at high speed. That’s a second reason why Google is winning in the chatbot market.
There are tons of solutions out there, some with even more whistles and bells than DialogFlow. But when asking developers why they choose DialogFlow over the others, it comes down to the quality of the Natural Language understanding and Machine Learning capabilities they receive with the platform. Building well-designed bots is having these 2 things aligned, and that’s what DialogFlow does really well.
And this is just the beginning, as DialogFlow is becoming more and more integrated with Google Cloud Platform since it has been acquired by Google in 2016. We will see more and more capabilities coming to the platform in the future, to keep the momentum and continue scaling.
This is the section where things get real. At Google Cloud Next ‘18, Google introduced 5 new killer features that support the Simplicity and Scale of where Google is going to with chatbots:
- Dialogflow Phone Gateway. This feature lets you convert any chatbot into an IVR system (Interactive Voice Response) in seconds. It provides telephony service, speech recognition, natural language understanding and speech synthesis out-of-the box. The speech is based on the WaveNet that make the conversation very natural.
- Knowledge Connectors. Really cool stuff. It’s not always easy, and pretty work-intensive to learn your chatbot new tricks. With Knowledge Connectors, you can now add any HTML page, csv or even PDF to your chatbot to build a knowledge base that can be used to fulfil user requests. Extremely useful for FAQs and no need for developers to help your chatbot learn new tricks.
- Automatic Spelling Correction. Users often use abbreviations,incorrect spelling or grammar. Thanks to the technology developed for Google Search, DialogFlow is now also able to correct user typos during conversations.
- Built-in Sentiment analysis. A very important part of providing a great customer experience is getting user’s feedback and emotional state of mind. Using the new built-in sentiment analysis in DialogFlow, you receive a score for every single query your chatbot has with the user. This makes it very easy to switch for instance from a virtual agent to a live one when needed.
- Built-in Text-to-Speech and Device Profile. Based on Wavenet technology, this feature delivers high quality speech responses right in DialogFlow. It’s extremely useful for telephony as well as other applications like cars, TVs etc. Device Profile will tweak the audio based on the type of device you use the bot at. It will adapt the speed of the conversation based on the device where you invoke the chatbot from (telephone, speaker, …).
It’s great to see how Google managed to incorporate DialogFlow in their Google Cloud Platform offering on such a short period. Also how they are leveraging the vast amount of proven features used in products such as Search, Google Voice and others to make developing chatbots simpler, more scalable and last but not least useful for humans.
I’m sure in a few months, the first result will change from “why chatbots fail” to “why chatbots rock”! 🙂
If you are interested to know more about chatbots, here are a few sessions worth watching from Google Next ‘18:
- How Dialogflow Enterprise Edition Can Transform the Enterprise Contact Center (Cloud Next ’18)
- Machine Learning-Enabled Chatbots (Cloud Next ’18)
- Intro to Building Serverless Apps for Google Home and Google Assistant (Cloud Next ’18)
- AI Powered Contact Center Analytics (Cloud Next ’18)
Get your own chatbot
Who we are? We are Fourcast, a Google Cloud Premier partner based out of Brussels with offices in Vienna and London. We help customers build cultures of innovation by implementing Google Cloud solutions such as hyper-personalised chatbots.
Contact us now, we’d be happy to talk about your needs and the possibilities of chatbots and Cloud technology for your company. We’re curious to hear about your projects and to see how we can help!