According to a recent Plug and Play Enterprise AI survey, 74% of companies now have at least one AI deployment live in production — past the pilot stage — but only about half of those companies can point to a measurable return on it. That gap isn’t only a CX problem, but it shows up clearly in the CX and Contact Center world, where a live production deployment means real customers are already on the other end of it. That’s what eGain, Liminal, NICE, RingCentral, and Sinch each spoke to in their own way at the Velocity CX/AI Event in Dallas — working through what makes an AI conversation useful to a customer instead of just easier to sell. Three questions shape how you close that gap for your own clients, serve them better, and grow your business at the same time.
The Signal: What Signal Is the Client Already Showing You?
Every AI platform can describe what it does. Fewer conversations start with what’s going wrong inside a customer’s business: the missed calls, the NPS scores slipping, the complaints piling up. RingCentral describes these as signals rather than pain points, because a client’s own customer experience data usually surfaces the issue before they can explain it themselves. Asking about the signal before naming a solution is what a trusted technology expert does, and it leads to a recommendation that actually fits the client instead of whatever’s easiest to pitch.
The ROI Gap: Is a Governance Gap or a Knowledge Gap Holding the Return Back?
Once that signal points to an account already running AI in production — and per the numbers up top, most now are — the next question is why it isn’t paying off. Liminal and eGain each point to a different answer — not the only possible causes, but two that come up often enough to be worth ruling out first.
Liminal’s approach centers on governance: most companies already have employees relying on public AI tools with no visibility into what data is leaving the building. That’s the shadow AI problem, and it puts a CIO or CISO in a bind — shut down access and lose the productivity gains everyone else is capturing, or leave it open and inherit the compliance risk. Either way, leadership can’t point to a clean ROI story when they can’t see how AI is being used across the business. The result looks like a governance gap holding back the return, not a technology problem.
eGain’s approach centers on something more basic: a bot or agent is only as good as the knowledge it can pull from, and scattered, inconsistent content produces a confident answer that’s wrong — which shows up to the customer as exactly the kind of inconsistent experience that erodes trust.
For the client, that’s the difference between a deployment that’s stuck and one that’s actively wrong: governance keeps it boxed in, bad knowledge makes it misfire. For you, it’s a sharper, more specific opening than any generic AI pitch.
The Overlay: Does the Solution Require Ripping Out What’s Already in Place?
Once the blocker is diagnosed, the next question is whether fixing it means starting over. For eGain and NICE, the strongest AI opportunities right now layer on top of what a client has instead of asking them to replace it — just from different sides of the Contact Center. eGain’s knowledge hub connects into a CRM or contact center a client already runs. NICE’s Cognigy platform sits over an existing contact center rather than swapping it out, so a client keeps the tools their team knows. Sinch follows the same logic at the infrastructure layer: its Voice Relay product adds voice AI capability to a client’s existing agents and call center operations through Sinch’s API, rather than requiring a new platform underneath it. For the client, that means no disruption to the systems their team already relies on, so the customer experience doesn’t dip while the AI layer goes in. For you, it means an easier yes.
Pick an account this week. Start with the signal — always. If AI is already part of the picture, rule out the governance or knowledge gap before you assume the technology itself is the problem. If it isn’t there yet, that same signal is your opening to shape the deployment around what the client already has from day one. Either way, confirm the fit before you name a solution. That’s the conversation that leads to a better customer experience — and, for you, a clearer path to the next deal, whichever of these five providers ends up being the right fit.