Based on its product marketing for agentic AI, Google seems to grasp the need to speak to solving ... [+] business problems.GoogleLast week Google announced a major update to its Customer Experience Suite, which is significant in two ways. First, the update reflects the changing attitudes and success measures in customer service today. Second, AI agents were used heavily to deliver new product capabilities to solve for those shifts in business attitudes.More than that, as I listened to the Google team brief the analyst community, I was really pleased with how the material was presented. Google did not do what we have seen from almost every agentic sales pitch so far, in ways Ill explain below, and it made a big difference in how the message got across. As someone who has been covering AI agents extensively for the past year, I think Google just raised the bar for product marketers across the industry. In fact, I believe that everyone involved in promoting agentic AI needs to pause and reflect on how Googles approach can help drive better agentic adoption.The Problem With All The Other Agentic PitchesHeres the unfortunate pattern of the typical agentic pitch:Establish that AI is great but doesnt go far enough.Possibly, suggest that the current stars of the AI universe (i.e., the vendors that build massive LLM models) want you to buy a one-size-fits-all offering that is not in line with most business needs.Introduce agents as the solution and offer up some sort of definition of agentic AI that reflects either the current state of the market or something very specific to what your company offers.At this point, maybe show a benchmark that nobody understands but in which your company is leading.Make sure you mention RAG and the importance of integrating your own data into AI models, but since thats really complicated, keep it quick.Talk about the cool types of things agents might do and how these will revolutionize work. Observe that the revolution will come through either making humans irrelevant or just making them more proficient at the lousy parts of their jobs. (But dont say cost-efficiency, because workers are super-sensitive to layoffs right now.)Launch a demo of some kind of common workflow (maybe travel reservations?). Preferably, the demo is based on a set of simple sample agents that won't scale in production or wont still be supported in six months due to a new version. From a user perspective, the demo agent seems no different from a chatbot, so now the boss has no idea how this is different from what they are using now.Thank the audience and answer some questions.How Google Did It DifferentlyOn its first slide, Google did quickly introduce agents and how they fit into the overall Google AI strategy. But then it went straight into discussing the challenges of customer service the relevant use case at hand and the potential of AI to not only reduce the costs of customer service (company benefit) but also reduce the barriers in delivering excellent customer experiences (end-customer benefit). The Google presenters described in detail a day in the life of a CSR and a CSR manager, then explained how different features will enhance business outcomes for companies using Googles Customer Experience Suite. To back it up, they even had customer stories from real companies household names like Verizon and Best Buy. They closed with a customer that is actually running this in production! The customer explained the benefits they got from working with Google. And the customer was also able to share both common customer service metrics showing cost savings and net promoter scores as well as new, more granular metrics including the reduction in Tier 2 escalations.No AI benchmarks. Real contextual business metrics. A vision for Googles agentic offering and the role it plays for its customers. Googles approach was refreshingly old-school, reminding us all that AI in its current state is a beneficial augmentation and not a massive infrastructure revolution for most of the companies that will deploy it anytime soon.You, Too, Can Do ThisThe spring tech conferences are upon us. So, lets break down what Google did so well with its presentation. Then you, too, can explain your agentic innovations in a consumable, cohesive way that customers will appreciate.Step 1: Set A Crisp Solution ContextWhen a technology is new, it can be hard to isolate a compelling business problem to target. This situation is even harder for a big business that wants to show breadth in what may turn out to be a massive market. To be fair, breadth is important so sellers and marketers can begin to execute. Sadly, however, focusing on breadth (Agentic changes everything) leads to a lack of focus that dilutes your message, and that honestly makes it harder for sellers and partners to get the point across. This is a case, Id suggest, where less is more.Sure, Google had it a bit easier in this instance because it was talking about a pre-existing product thats been augmented with agentic, rather than some all-new use case. And it was clear that the product team has spent significant time with customers experiencing the exact frictions addressed in the presentation. But even if you dont have an existing product line, going deep on a couple of very clearly articulated business problems targeted at specific personas will get you way further than you think. (As an aside, I am also finding that developing this kind of persona-based research is a lot easier using AI-based research tools with the right set of prompts.)As for any fears of going too narrow, I would say that enterprise IT customers can very often see how a use case can be reapplied to their own situation. If they dont, they will at least respect the rigor of the exercise, which will buy you credibility, and they may be inspired to ask follow-up questions about their specific needs.Step 2: Stop Basking In The AI What And Start Talking About The Agentic WhyI understand that AI is new, sort of magical and technically very different from what has come before. But we are now at a point in this markets evolution when the people who need to care about AI already have a basic understanding of it. This is a clear sign that we are passing through and out of the early-adopter stage for AI.However, agentic is an interface between the real world and AI. Its not AI per se. Heres a great example from Google that exemplifies the difference: While other customer-service AI solution vendors are talking about call deflection and cost reductions thanks to AI agents speeding up processes, Google explained that there are AI agents that can be used to coach and train CSRs so they can be more effective at their jobs. Training is a significant and immediately relevant challenge in the high-turnover world of customer service. Enhanced training features help customers understand why there will be material changes in their metrics, versus just stating that AI accelerates calls.Step 3: Metrics, Experiences And ValueThe natural outcome of steps 1 and 2 will be a clear articulation of what the customer should experience and the value of their journey. In some cases, this will lead to new ways of considering old problems as Google suggests with its vision for the Customer Experience Suite. Or maybe its going to be something simpler like unblocking complex workflows, as weve seen in the case of solutions like Agentforce from Salesforce.Either way, from a marketing perspective the value proposition is not simply Its AI. To overstate something that should be obvious but too often isnt: The value proposition is that there is a business problem that is solved with a measurable impact. And even if you dont have customers willing to testify yet, you do have the ability to lay out the rationale and design points for the agentic product to describe the value it is intended to create. If that resonates in your product marketing, trust that the other questions will come up later in the sales cycle.Stop Doing AI Marketing For Agentic SolutionsIn many ways, what we are seeing amounts to AI marketing teams attempting to apply a technology template to agents which are, fundamentally, business solutions. Yes, AI is in some ways like a server because its a horizontal technology. But agents are not of that class. Agents are a new type of application and need to be marketed as such. Google just made that very clear. Now lets see how long it takes for other agentic providers to get on board.Follow me onTwitter or LinkedIn.Check outmywebsite.Moor Insights & Strategy provides or has provided paid services to technology companies, like all tech industry research and analyst firms. These services include research, analysis, advising, consulting, benchmarking, acquisition matchmaking and video and speaking sponsorships. Of the companies mentioned in this article, Moor Insights & Strategy currently has (or has had) a paid business relationship with Google and Salesforce.Editorial StandardsForbes Accolades