Because conversation analytics platforms are becoming more prevalent across the AI/ML landscape, there is a growing opportunity to explain how Symbl.ai is leading the charge in terms of deriving meaningful conversation insights. One of the best ways to illustrate this is by showcasing the work we are doing to satisfy the needs of our enterprise customers.
Enterprises typically have far more lofty goals than your typical mid-market or small business concerns; think far less task oriented and more predictive analyses and optimizations of outcome scenarios. This post, the first in a new series, will explore how Symbl.ai is tackling enterprise pain points in a way that differentiates itself from other voice intelligence platforms.
The problem: Conversation isolation
The biggest problem conversation platforms struggle with today is conversation isolation.
What does that mean? In brief, conversation isolation occurs when all of the insights—dynamic topics, entity detection, etc.—are siloed to that single Zoom session, telephony call, etc. That data never gets accessed or used outside the confines of the session or conversation, and you can’t establish patterns between all the conversations that are taking place within your business. Simply put, we are talking about aggregating ALL conversations and applying predictive analytics to them.
The first obvious question everyone should ask is why that even matters. Why do we care about establishing patterns in conversations over time? If you are a small- to medium-sized business, you might not actually care.
You may be able to satisfy all of your business’ needs—enhancing call center support, sales enablement, etc.—with any of the platforms out there today. There are definite and proven advantages to working with smaller teams; among those advantages are heightened communication and collaboration.
Scaling the contact points between your teams becomes impossible if you are operating at enterprise scale with the aim of optimizing customer experience, revenue growth, and communication channels. Those businesses that do this well are often the leaders in their industry. What if I were to tell you that there was a hack such that your business can put the dots close enough together that patterns to your business’ conversation will start to emerge? That’s what the Symbl.ai platform is all about for these enterprise use cases.
No more pie in the sky—give us the use cases!
The applications of conversation analytics in a given enterprise are endless, but here are a couple of use cases that are very relatable problems in this space.
Use Case No. 1: Enhancing support in the call center
Let’s say we are looking at a typical day for a cloud-based service provider. Think everything from video conferencing platforms such as Zoom or, in the use case we are going to examine today, an internet service provider. You notice that support calls are increasing this morning beyond the normal “I need to reset my password” requests.
If you employ real-time conversation analytics, the platform can pick up on keywords (we call them Trackers) such as “outage,” “offline,” and “no connection,” and then start to analyze these customers’ profiles.
Because call volume is increasing, the aggregation of these Trackers might reveal that the designated customer calls all originate in the Los Angeles area. Based on these Trackers and the information provided, we could also proactively signal a higher-level engineer to check out dashboards for connectivity in the Los Angeles area. Even better, perhaps this signal actually kicks off the automation to check services in Los Angeles.
Suppose that automation finds something wrong with the network, and based on the automated health checks the failing test could determine that human intervention is necessary to fix the problem. The platform might be wise to inquire whether this has happened before and find the engineer(s) who did the root cause analysis in order to assign them to this issue.
A significant amount of day-to-day business activities are spent reacting to events, issues, and requests. Given this paradigm, your business can take a proactive approach via automation based on real-time conversations and optimize outcomes by intelligently allocating workloads and resources.
Use Case No. 2: Understanding your customers’ needs before they do
A far more advanced capability is trying to predict the future or, better yet, change the decision tree of an individual’s choices such that the final outcomes are favorable to the problem at hand. This sounds extremely hand-wavy, but in reality this is happening all the time whether you realize it or not. It affects everything from what your next purchase will be all the way down to how you might vote in the next election.
At some point or another, everyone has noticed that what they were checking out on the internet 30 minutes prior suddenly appears on their social media feed as product suggestions. This is 100% NOT a coincidence, but it is a form of manipulating the decision tree such that the paths are redirecting you to a specific desired outcome.
This capability extends far beyond what people you follow, the videos you watch, the friends you have on various social networks, etc., but all of those things collectively as a whole contribute to altering your decision tree. This is exactly how playing an AI computer opponent in chess works.
That same concept also applies to the conversations being had across your organization in the form of Zoom meetings, Slack messages, emails, and more. Topics and Sentiments discussed between sales teams, customers, product managers, and engineers play an essential role in developing better solutions that solve real-world problems.
To make this more tangible, great sales organizations constantly discuss new use cases with their customers. Typically, these conversations are ongoing because your customers’ needs continuously evolve as they grow.
With Symbl.ai businesses can easily compile all these customer conversations and aggregate that information using Trackers and Action Items to recognize patterns across all conversations. The resulting data could prioritize requirements for your product’s next iteration or perhaps an entirely new solution based on customer gaps. Without realizing it, your business has just altered its decision tree for the product roadmap in favor of one that will have a greater impact on your customers.
Where do we go from here?
This blog post is meant to address the “Why?” of conversation analytics for enterprise customers.
For the most part, these large-scale systems aren’t being used to solve “traditional” tasks. Instead, they are being used to alter the decision-making process and give the business tasks that offer a higher return on investment. That is a powerful and game-changing realization.
In the next blog post, we’ll get more tangible regarding the question of how any enterprise might build one of these systems and what that architecture might look like. Stay tuned for the next blog in this series if you’re a system architect, software engineer, solutions engineer, or someone just interested in problem solving with cool tech!
