Google Next 2018 Recap – Google bets big on Enterprise Communications
At the Google Cloud event GoogleNext 2018, there were a lot of announcements. While the event is not as big (yet) as Oracle World and Dreamforce, it was big, occupying all of the Moscone Conference Center, in addition to venues in Yerba Buena Gardens and the Palace hotel. This conference is focused to how to use the Google cloud for business deployments.
For the cloud group, there were plenty of new areas, Kubernetes for container deployment, new AI and machine learning, security, manageability, new applications and more. Overall, it is clear that Google is becoming a major player in the cloud computing services world. In fact, Google claims that one percent of all enterprise workloads already run in the Google cloud (in addition to over 20 percent of Internet traffic that transits Google networks).
But the real news was around the communications space where Google made two major announcements, one expected and the other a bit of a surprise.
First, Google announced their implementation of machine learning into the Contact Center AI (CCAI) service, provided by Google from the cloud. CCAI is designed to be more than just a machine learning API, it already includes appropriate data models for contact center as well as tools to enable both virtual agents (bots) and agent support. The key is using the two key components of the Google tech stack, Natural Language Processing (NLP) to interpret what humans say (or write), and the search capability. In other words, finding the right answers to a customer query is very similar to a web search request, something that Google does reasonably well.
The result is a set of APIs that enable the applications of NLP advanced speech recognition and contextual information-based search to deliver contextual answers and information to a customer as a bot (Virtual Agent in Google terminology) and suggestions for agents.
The application of industry leading Machine Learning/Artificial Intelligence (MA/AI) technology to the call center has benefits for virtual agents, live agent assist, and in analytics. There were several partner announcements, but a few were in a session where they discussed how they were taking advantage of the Google CCAI platform to deliver new value. Mitel, Genesys, KPMG, and Chatbase (a start-up sponsored by the Google accelerator) demonstrated their innovation using the underlying technology. Mitel and Genesys showed how the machine learning can improve the virtual agent/bot experience and how agent assist could be used. Mitel discussed the use of ML/AI in analytics, enabling a much broader machine-based initial analysis versus the current statistical methodologies employed in most Contact Centers. With CCAI-based analytics, the ML/AI capability does the initial categorizing and defines where focus should be spent in human analysis. This has the capability of focusing analytics that involve a human to increase effectiveness of analysis dramatically. In addition, KMPG discussed and demonstrated using the Topic definition capabilities in the CCAI to identify key topics and then zoom in to which areas need to be changed. The focus is on eliminating calls to a live agent or improving the information for the agent to resolve the issues. KPMG is developing a methodology for the process of managing the data sets and the topic analysis. Finally, Chatbase went through using the topic and analysis tools of the CCI to eliminate the very time-consuming task of building the potential response areas for a chatbot. Chatbase claims up to an order of magnitude reduction in bot generation time and effort by using the CCAI to structure the bot.
Overall, the complexity and challenges of machine learning and AI are rapidly moving outside of the ability of traditional communications, collaboration and contact center companies to develop internally. The emergence of API available sophisticated cloud ML/AI solutions from vendors like Google, Amazon, IBM, and Microsoft will make it difficult for the communications companies to develop their own ML/AI technology. A major question is how the ML/AI tools partners of the CC vendors will align with their customer choices for other activities. Or perhaps ML/AI as an analysis/computational engine is more like an operating system where the choice will be matched to the application’s solution.
The second major announcement, one that almost was under the radar during the keynote, was the announcement of Google Voice as a subscription enterprise offer. In a session innocuously titled “Exploring Telephony Solutions with G Suite,” Smita Hashim and Ajay Surie of Google detailed the strategy for Google Voice as a subscription service. After talking about how Google replaced existing PBXs (complete with pictures of old equipment racks), they proceeded to detail how a G Suite user could easily deploy Google Voice to replace an existing PBX. They positioned Google Voice, Google Chat, and Google Meet as the three components of the overall Google communications and collaboration solution. In addition, they presented a roadmap that indicated Spring 2019 general availability for Google Voice. The roadmap includes desk phones (they indicated discussions are underway with Polycom and others about compatible phones) as well as features like auto-attendant and group ring (in a later release). However, when asked about support for SIP trunking, a key capability for integration to other communications systems as well as managing PSTN access costs through third-party providers, they indicated there are no public plans to provide this level of interoperabity at this time.
The Google Voice announcement is a bit of a Gordian knot. Microsoft has had marginal success in winning large telephony deployments that were not associated with Knowledge Workers. Similarly, Google’s strength is in G Suite and Google Apps. However, these are deployed in a relatively small percentage of personal productivity applications, where Microsoft Office enjoys 80 percent adoption. However, for organizations adopting G Suite that have predominately Knowledge Worker populations, this is a potentially interesting offer. Google is making progress in winning organizations, especially in higher education, health care and government. Across those markets, the option of Google Voice for the G Suite users may be a viable option in the near term.
After it was all done, Google Next emphasized how dramatically the communications market is changing. From the multiple vendors partnering on ML/AI for contact center to Google Voice for enterprise announcements, Google is ramping up a focus on enterprise. As Google starts from the cloud and G Suite as the launch point for this entry, it is clear that adjacencies continue to be a major factor in the communications market. For end user customers, Google is creating the same challenges that Microsoft has created. Both Microsoft and Google tools are focused on the Collaborating Knowledge Workers (read this PKE Consulting White Paper on worker types).
This means for many of he organizations that are adopting these solutions, there is already an in place telephony system that probably will remain, at least for some time. With over 66% of North American Business Communications endpoints at locations that are not assigned to a Knowledge Worker, the lack of broad telephony functionality support limits both Microsoft Teams and Google Voice as a complete solution today for many organizations. For companies deploying G Suite and Google Voice for part of their user base, the key question may be how to integrate to the solution supporting the rest of the organization. For vendors, the challenges of living in a frenemy world continue and the environment gets ever more complicated. The good news for G Suite users is that they no longer have to be envious the telephony capabilities of their Office 365 brethren. Can you hear me now?