Leadership Insights: Generative AI for CRM Systems

Transcript

Mike Vizard: Hello and welcome to the latest edition of the Digital CxO Leadership Insight Series. I’m your host, Mike Vizard. Today we’re with Amit Ben, who is CEO for OneAI, and we’re talking about how generative AI might be applied to CRM systems. Amit, welcome to show.

Amit Ben: Hey, thanks for having me, Michael.

Mike Vizard: We’ve seen a lot of talk about generative AI and just about every use case you can imagine, but what is really practical from a CRM perspective? What should people be expecting? I know we’ve seen Salesforce talk about a few things, but what can I actually do?

Amit Ben: Right, so I think CRM and sales and service use cases are really relevant and interesting for automation with AI because they’re exhibiting a lot of repetitive use cases, where there’s a lot of value in automating and freeing up people to actually handle unique cases and where there’s a lot of business value behind them. What we can do today with AI to automate the CRM use cases, there are a lot of things you can do, anywhere from processing conversations and feeding CRMs with actual concrete and accurate data in CRM reports. So what’s the summary of these conversations? What product did the customer discuss? What’s the pricing they discuss? What issues they have, what’s the next action items they agreed on?

So all of these details that usually agents will potentially manually populate the CRM with, can now be automatically populated with AI. And that gives you much broader coverage of all of the properties that you can populate the CRM with, and it also gives you higher reliability for that data because the AI is going to always populate with all of the field and always with the relevant information, and it doesn’t give you bias from agent to agent, which might see things a bit differently in different cases. Their training might be a bit different, their experience might be different. But the AI gives you a level set of all of these parameters. And then you can tune the AI as well, if you want specific data to be extracted in your CMR use cases. So that just from a CRM population standpoint.
On the other end of things, everywhere, everything from processing the conversation itself, helping agents to do a better job in identifying issues, deflecting issues, summarizing user transcripts so the agent can be up to speed much faster into the conversation and save their time, that’s also a different perspective, what we can automate the CRM with. So basically give the agent superpowers, if you will, AI superpowers to process all of the customer interactions.

And finally, there’s another suite of capability that gives you a bird’s eye view of all of the operations that happen in the CRM. Traditionally, only very large organizations had the benefit and the ability to deploy such solutions because of their cost and complexity. But today, with the type of AI that OneAI is offering, you can quickly plug in your CRM into the AI data stream and the AI will collect all of the information that go through the CRM pipeline, and then give you an analysis of all of that. So what are the recurrent complaints on which products and which geographies? Which people or what type of customers have what type of complaints? If you have trending issues, if you have competitors that have started to be mentioned in conversation and so on. So anything from the language analytics standpoint, now we can do that at scale, and the AI can read and listen to millions of conversations a month to give you a real coverage.

Mike Vizard: What level of skill do I need to have to accomplish that? We hear a lot about prompt engineering. Does my Salesforce or CRM admin need to have those kinds of skills or will there be templates that do that for me? How easy can this get?

Amit Ben: Yeah. So traditionally, with large language models, you would have to be an expert in prompt engineering. And even then, not every prompt will work with every input and it’s a bit finicky. That’s why OneAI have taken a different approach. What we’ve done is to take those language models and train them to not need a prompt. It does optimize things to a degree where we can make the model be consistent and reliable regardless of the input of the user. And you can just drag and drop these capabilities into your CRM and no prompt engineering is required.
It does mean that you would need to use a capability that we’ve released. Now we’ve created dozens of CRM-targeted capabilities that have been pre-trained and tuned for these use cases. But that means it’s drag and drop and you can immediately use it without any experience.

Mike Vizard: Do I need to standardize on a single CRM to kind of take advantage of these capabilities? Or are you thinking about an approach that spans multiple CRM systems, because not every organization has come up with a standard yet?

Amit Ben: Yeah. So OneAI is an API company. So basically, we provide the API layer that you can integrate your CRM with to augment your CRM. With all of these capabilities. We have integrated with Salesforce and HubSpot and various other solutions, and it’s actually pretty easy to do.

Mike Vizard: What do you think is the issue going forward here? Is it now a question of culture or more of a technical challenge? Or do you think that the sales teams may be running ahead of the organizations because they can access ChatGPT themselves? Where are we on this adventure?

Amit Ben: So you’re asking what is causing the sense of urgency today?

Mike Vizard: Yeah, it’s not just a sense of urgency, but I guess are the salespeople moving faster than the organization themselves? Because the ones that I talked to have already started to use ChatGPT and all kinds of capabilities within their workflows, and are the organizations catching up?

Amit Ben: Yeah. So it’s starting to happen, organizations are catching up. And whenever a technology comes into play where you can augment your Salesforce with, it starts with individual agents. We’re using different tools. But once you put it on an organization level and you streamlined integration and streamlined the tooling, you can get a different level of value add when you get all of your agents on the same platform with the same capabilities, and then you can deliver this value at scale. And if something changes or if you need to update the data or there’s an improvement in tooling, it goes out to all of the agents at once.
I think there’s a realization now that AI has become mature enough that you can rely on it on these use cases. And people have gotten to understand how powerful this is and that they can completely transform their throughput when they utilize these types of capabilities.

Mike Vizard: What do you think the impact will be on the number of salespeople that I need to scale an organization? Can I get by with fewer salespeople? Because the wisdom is always, it takes six months to get a salesperson up and running and then they can only handle so many accounts. So what does this do to the whole metrics that people use to measure their sales teams?

Amit Ben: Yeah. I think it really depends on the type of sales process that you’re running. For example, if we’re talking about an account executive that is doing a very high level sales process, I think it’s going to make it a bit easier for these guys to do their job. With all of the capabilities, they’re going to need to do less prep work, as the AI can do the consumption of the data, and to summarize it for them to give them the key action items for everything. So it’s going to make them more efficient, but I don’t think AI is going to replace high level account executives soon.
And on the other hand, when you see entry level salespeople coming into the play, I think it’s going to give them an unfair advantage compared to the situation that’s been going on up until now, because now they can get the benefit of all of the knowledge, the know-how, and the best practices that’s been accumulated through years. So a new guy comes to the organization, instead of taking six months to ramp up, to understand the product, to understand the services, the needs, the features, so the objection handling, now they can consult with AI that was trained on the organizational data and they can be up to speed much further. And they can consult this AI in real time in many cases. I think we’re going to start seeing solutions that are companions to sales agents on calls or they listen to a sales conversation and they can give the agent hints, from a language model, and suggest answers to the agent based on the organizational data. And that’s going to basically take everybody one or two notches higher in their skill level.

Mike Vizard: Are we entering something that feels like an AI arms race then? Organizations that don’t have CRM infused with AI will find themselves falling behind?

Amit Ben: Yeah, I think it’s not only in CRM, there’s an AI arms race in various industries, where companies that are not going to be leveraging AI in their normal day pipeline, everyday pipeline, they’re going to be left behind. And I think the ability for companies that harness AI to compete is going to be tremendous.

Mike Vizard: What exactly is the benefit of having an approach like yours versus simply relying on the AI engine that the CRM vendor provides exactly? What’s that line of, is one enough or where do I really think about this at a different level of scale?

Amit Ben: Yeah. So the advantage of using OneAI is the ability to have an AI that’s been fine-tuned on your data, and can access your data and is adapted to the organization. And OneAI’s ability to very easily adapt to every use case, to every customer is unique, in our speed and also in our cost structure. Because we’re able to optimize these models in a very unique way and to also consume knowledge in real time when we generate the answers.
And further, the suite of capabilities, which is very broad, is beyond anything that a single provider has. So you get a one-stop shop experience that gives you all of the benefits at once, and the ability to connect all of these value-add capabilities is kind of a one plus one equals three. Because when the AI has access to all of the different data points at the same time, now you’ve got something that can make better decisions and really inform an agent from a 360 degree view.

Mike Vizard: What’s your best advice to organizations then as they look at this new frontier? Should they be just experimenting or is there some other way to think about this?

Amit Ben: I think organizations need to, A: Really seriously think about what they can do and how they can harness these capabilities. One way of thinking about this is figuring out where they’re wasting the most human brain power on things that they can automate, like what are we spending a lot of time in handling or we’re not doing because we don’t have enough manpower? These are opportunities to leverage AI with, either to augment or to offset your efforts.
And B: I strongly urge organizations to take some form of advisory company or organizations are going to help them ingest and harness these capabilities. People that have experience, that you can help them integrate the capabilities so they can move faster, instead of trying to build the knowledge in-house.

Mike Vizard: Hey Amit, thanks for being on the show and sharing your knowledge and expertise.

Amit Ben: Of course. Thank you for having me, Michael.

Mike Vizard: Thank you all for watching the latest episode of the Digital CxO Leadership Insight Series. You can find this one and others on the digitalcxo.com website. We invite you to check them all out. And once again, thanks for spending some time with us.