No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Speech Tech 2020: More About Innovation Than Disruption

SpeechTech_012820-AdobeStock_184999496.jpeg

Illustration of chatbot on smartphone
Image: sdecoret - stock.adobe.com
Enterprise Connect 2020 is fast-approaching, and aside from being a window on a new decade for communications technology, it’s also the 30th anniversary for this long-running event. While these numbers are noteworthy markers, the bigger story is the exciting promise of technologies that are coming of age now, especially cloud platforms, AI, and 5G.
 
In my own small way, I’ll be weighing in on some of this during my session at Enterprise Connect. I’m happy to be returning for a third iteration of the state of speech technology in the enterprise, and if that’s of interest, all the details for my session are here. This post will highlight the key themes I’ll be talking about, and to complement that, last week I spoke with No Jitter’s Beth Schultz, and our No Jitter on Air podcast serves as an extended preview of my session. Click to listen here, or on the player below.
 

Key Theme #1 — Disruption Versus Innovation
When talking about speech technology, it’s impossible not to touch on AI, which is hitting its stride across the board in 2020. When AI gets attached to anything for the first time, it’s disruptive since so much is unknown, and the impact will not likely be small. Nobody is looking to AI for incremental gains — it’s all about being transformative, and that’s why it’s disruptive at first.
 
That has certainly been the case with speech, which was a pretty mature space until AI came along. The leaders were well-established, the use cases clear, and the limits largely known. AI, of course, is an umbrella for many technologies and data science disciplines, and when driven by the cloud giants — namely Amazon and Google — new possibilities and applications opened for speech.
 
That’s how the market looked when I started doing these updates in 2018, and it certainly seemed disruptive. AI was going to turn the workplace into a surveillance zone, and contact center agents were going to be automated out of their jobs. These scenarios haven’t happened — at least so far — but a lot of progress has been made with speech technology in a few short years. In some areas, AI still has a way to go to make big inroads, but there are solid proof points to show where it’s bringing new business value with speech.
 
In 2020, the state of speech technology in the enterprise is more about innovation than disruption, and that’s a good thing. Last year I talked about how Alexa for Business marked the first visible entry with AI-driven speech, and it’s an important first step in normalizing this mode in the workplace. The major collaboration vendors are now all talking up their flavors of AI — some of which are speech-related — and while it’s still early days, there’s no turning back. AI already seems mainstream in the consumer world (think smart home, smart car, wearable tech), and the enterprise space is the next frontier. During my talk I’ll showcase some examples of speech innovation, both in the workplace and the contact center.
 
Key Theme #2 — Where the Traction Is
Even though new technologies like AI can diffuse faster today than in the past, things are moving at different speeds in the two spaces we pay most attention to when it comes to speech — the workplace and the contact center. Things certainly seem to be moving quickly in the consumer world, and from that, it’s not surprising to see more applications of AI-driven speech in the contact center. Consumer expectations related to AI in their everyday lives is carrying over to customer service expectations, and that’s why the contact center has been more aggressive in finding ways to deploy AI. The stakes around CX are getting higher now, and every contact center vendor is touting its AI capabilities to enable agents to level the playing field with today’s more demanding customers.
 
My session is focused mostly on the workplace, but I will touch on how contact centers are using conversational AI for customer engagement. Since the use cases are more pressing in the contact center, AI is gaining faster adoption there, and during the conference, there will be deeper dives on this in the Contact Center & Customer Experience track, especially here and here.
 
The workplace environment is embracing AI-driven speech more cautiously for exactly the opposite reason, so there isn’t as much to update on the innovation front from last year. Aside from Alexa for Business in the conference room, the main breakthrough for AI-driven speech would be real-time transcription. I covered this in last year’s talk, and the use case remains strong, but not much has changed since last year. That said, it’s a great showcase for AI’s value in the workplace, and it helps build a foundation of trust for the next wave of innovation.
 
Concepts like “cognitive collaboration” and “social graph” are very much AI-driven, but the use cases aren’t as universal as transcription — at least for now — and in 2020 I expect the vendors will keep fine-tuning these offers. It’s also important to note that speech is a secondary component of these applications, and I need to stick to the focus of my talk. While most innovation around speech is AI-driven, not all forms of AI are about speech. The business case for AI in the enterprise goes well beyond voice, so my focus is really just a subset of a broader trend.
 
Coming back to voice, I will cite two examples where AI-driven innovation is emerging in the workplace. One will be voice biometrics, and the other natural language generation (NLG). The former has fairly specific use cases, and while we’ve seen some applications for authenticating workers to join a meeting, I’ll talk about other examples where usage is growing. NLG is a separate space altogether, and while it’s been overshadowed by its cousins, natural language understanding and natural language processing, I’ll talk about how it’s about to become part of the enterprise speech landscape in 2020.
 
Key Theme #3 – How Decision-Makers Should Be Thinking about Enterprise Speech Tech
There are a few distinct messages I’ll focus on as takeaways for investing in new speech technologies, especially those heavily driven by AI. The first is to have a clear sense as to how much of this is going to be tied to broader AI priorities, as the business case for AI will only be partly built around speech. Building a business case around speech technologies needs to be more than an AI story, especially since associations with AI still carry high expectations that may not be realistic in your organization.
 
Another message will be a familiar one, namely to focus on specific use cases for which the benefits are clear, such as voice-enabling meetings or real-time transcription. This is especially important for the workplace, where the use cases aren’t as clear-cut as in the contact center. While it’s not hard to visualize how AI-driven speech applications can improve productivity or enable better collaboration outcomes, these solutions are still early stage. Some of this challenge lies with the vendors, but some also lies with IT decision-makers, and I’ll speak further about how buyers need to think about where these new capabilities really fit and what expectations need to be set around them.
 
This leads to another takeaway that hasn’t changed since I started doing these updates. Namely, how anything that is AI-driven will be iterative, especially speech, where the standards for accuracy are very high. Unlike a PBX, which works pretty much perfectly right out of the box, AI-driven speech applications will be good out of the box, but far from perfect. Again, unlike the PBX, which is effectively a static technology, AI applications “learn” and get better over time. As such, decision-makers need to recognize this is a journey, and it’s best to start with small pilots, and scale them as a comfort level is reached with the learning curve for improvement.
 
Conclusion
This post alone could take all my speaking time to cover at Enterprise Connect, but I’ll have quite a bit more to add, not just as an independent analyst, but also by bringing the perspectives that come with wearing my business communications expert hat as part of BCStrategies. This is one of the first sessions kicking off Enterprise Connect on Monday, March 30, at 8:00 a.m., so get there early. Also, if you haven’t registered, you can use this link to save up to $400, and I hope to see you there.

This post is written on behalf of BCStrategies, an industry resource for enterprises, vendors, system integrators, and anyone interested in the growing business communications arena. A supplier of objective information on business communications, BCStrategies is supported by an alliance of leading communication industry advisors, analysts, and consultants who have worked in the various segments of the dynamic business communications market.