Beyond Gen(eric)AI: The Impact of Sales-Aware AI

As we talk about our journey in AI and sales, there's one thing that I thought would be worth just reflecting on, and that is the landscape that we're operating in now has really changed over the last few years. So I talk a lot to analysts like Gartner and Forrester, as I'm sure lots of us do. And you know, we were saying that the barriers between revenue enablement, sales training, sales platforms are starting to come down. And I think that's because in order to really capture the full value of this kind of thing, one needs to bring three elements together. One is the platform and the AI. And there are a lot of organizations speaking yesterday and today about AI, but AI in isolation, although it can do some amazing things and look at sentiment and give you a broad role play around a particular topic, it won't help your salespeople to sell the way you want them to unless it's aligned with your sales methodology. And sales training that gives people the cognitive part of a skill. So you can't just do practice, you have to have that, plus the theory, if you like. And then all of those things require a human view of the change process. So in my mind bringing, you know, there's a proliferation of new platforms out there at the moment, but there aren't many sales training companies who have really been able to grasp these other two areas of change in platform and AI. And that's where I think, you know, we've had an amazing two years really taking all of our experience in sales training and change and because we're a technology company as well, integrating it into our AI. So I'm coming at this with that sort of combined view. And if I may just introduce in Parta very briefly, so we're one of four global leaders in the most recent Gartner sales training magic quadrant. We're pioneers in what we call sustainable revenue growth. So it's less, it is about courses, but it's less about courses and more about that kind of continuous improvement. And there are three parts to our ecosystem. One is agile sales methodology. So we train salespeople not to follow a linear process, but to assess each customer situation objectively, choose the right strategy to maximize your outcome from that, and then execute the right sales motion in a way that balances three dimensions, insight, influence and, and trust. And I'll come back to those in one minute. So we have the methodology. There's a Skills Library of 180 different skills training modules that are instructor led, elearning, virtual and so on, plus CRM canvases. And manager tools and everything that you would expect.

That doesn't work unless it's embedded, as I said, in a change process. So we use an agile change process as well. We assess the learning need, choose the right approach and customize it to the learning, help to apply it and make it stick. And that needs to operate at scale across the whole world, essentially. But the key thing about agility is what we need in your sales team today, the critical skill to deal with today's market. So a year and a half ago it was asking for inflationary price increases. Right now it might well be refilling your pipeline. In the past it has been dealing with the pandemic. The challenges are constantly changing. And so the learning solution, both the AI role plays and the coaching and so on, but also the training needs to be able to adapt super quickly as your internal KPIs move around and as the external market evolves. And the third part of the stool, if you like, the three legged stool that we sit on here, is the technology and the platform. And technology is fantastic. Sales technology can often get in the way, especially when it's a very fragmented sales tech stack, as I think a lot of us ended up with. And so we use sales aware AI. I'll explain that term in a minute, as Gerhard said. And a modular approach to allow us to augment and enhance what our clients already have. So we have our own LXP, but we can also work with other LXPs with your CRM and so on. So those are the three components of our kind of ecosystem, if you like.

And I just want to pick up one element that I spoke about briefly, because it's going to be immediately relevant to why we felt the need to create sales aware AI. My research took every single skill around the buying cycle and there are over 100 potential skills. Nobody needs 100, but there are 100 potential skills you might need at every stage of the customer cycle. And we basically broke them apart and said what is actually going on in each of these? And we found something very interesting, which is that there are three common dimensions that exist in every skill around the customer journey and also in every sales manager skill and even the sales leader skill. And those are insight. So the first is, can you create value for a customer or for your team if you're a manager, through some kind of disruptive insight into their needs, possible solutions and their decision process? I'm paraphrasing, but broadly that's what insight is. But insight on its own just makes you an educator. It doesn't make you a salesperson. So you also need to be able to influence customers to take action through the ethical use because these are powerful tools of behavioral science, decision strategy, psychology, behavioral economics and so on. And if you have those two elements, you start to be a good salesperson. But there's one other thing that you need to get access to be able to have those conversations, and that's trust. And here I'm a big fan of David Meister and his colleagues Charles Green and so on, their trust equation, which essentially says you build trust through credibility, reliability, what we call business intimacy, and low levels of self orientation or in particular customer centricity. So if you have, what we found is if you have a salesperson or a sales team that balance those three dimensions, well, they will outperform a team that is only good at one or two. A lot of organizations have people who are what we call friendly helpers who are kind of only trusted, or consultants who bring insight and are trusted but rarely influence actions. So those are quite common archetypes. So why am I explaining this in a session on AI? Well, the answer is genai. And if we take ChatGPT as a good example, is amazing. It's a really incredible technology, but it's terrible at selling. So if we take ChatGPT and we give it this prompt, write a prospecting email to me, CEO of mparta, to try and get a meeting to discuss recruiting for his sales team. Now I actually did this this morning to make sure I was getting the latest and I used ChatGPT4TUR, which is still seen as the better model at the moment in terms of. I know there are new models coming out that do chained reasoning, but compared to 4.0, it's better at kind of fundamental reasoning. And it came up with this, which is incredible. It can create a good English and put something together that's fundamentally not wrong. But then I scored that against those three dimensions of insight, influence and trust, just as one way to look at how well the AI did this. And so when I did that, and we've done this research over the years, actually it does pretty badly. There's very little insight and certainly no insight that's tailored because you ask it a question and it gives you the answer there's no influence. It's not using any psychological techniques to try and get me to actually speak to this person. And there's very little trust. In fact, it comes across as very self oriented. And see, AI based prospecting, prospecting is just one skill out of Many, many that people need. But it's a good example because we've all been at the receiving end of this kind of email, right? So it says things like I would love to spend time with, I'd like to get to know you and talk about your problems. And that's a very self oriented thing to say. So that's the problem with generic AI is it knows a lot but it's not oriented around what you have to do to sell. So just to take this example of prospecting a tiny bit further, here is some of the science behind good prospecting just grouped into those three categories of insight, influence and trust. So in insight you've got to bring value. You can identify potential objectives and needs that they weren't aware of before. And it should be hyper customized. You can't just expect to kind of give a generic insight and it really land with somebody who you know is getting 20 emails a day or 40 emails a day. So you can talk about needs, you can talk about capabilities. In terms of influence, we can use social proof, we can use reciprocity. So giving people something before you ask for something in return. A key one when you're trying to get a meeting is loss aversion. So frame that call as a way to avoid a loss rather than to achieve a gain. Because people are motivated more by loss than by avoiding a loss than by achieving a gain. And there are other tools like spontaneous trait transference. So if you describe your other clients in positive terms, people associate those terms with you and so on. And then finally in trust, be credible, care about their business and demonstrate low levels of self orientation. So that's just one piece of intellectual property around how you do good prospecting. Now I'm just going to come out of the share and reshare with audio this time.

And what I'm going to show you is just a quick live demo of how I coach AI, which is our sales aware solution will do that task for us. So this is a general AI sales coaching and I'll talk about the different use cases. So we do everything from sales call assessment to eLearning role plays, AI generated role plays to customer conversation simulators where you practice in advance and coaching to help you figure things out and so on around the change process. This is the kind of general coaching mode and I'm going to try that same task here as well. So let's. By the way, this is just running in a browser for simplicity. I can run it inside our LXP which is there, we can put it inside Teams inside Salesforce or Ms. Dynamics or inside another LXP and so on. So let's have a go.

Hi, I'd like your help please, in writing a prospecting email to Richard Barkey, the CEO of Imparta, trying to sell sales recruiting services to him for his team. He's not aware of us, he's not heard of this company before.

Let's start by discussing the content of your email to Richard Barkey.

We want to ensure that it incorporates insight, influence and trust effectively. Firstly, let's talk about insight. What observations can we make about Imparta or the sales industry that Richard might not be fully aware of and what are the implications for his business? Could you share any specific challenges or trends in the sales industry that might resonate with him?

So. So we know Imparter is growing, we think about 20% a year at the moment and that will have its own challenges. One of our insights inside our business is that it's more important to think about recruiting people who will stay. That's often about cultural fit more than anything else, and that's what we're good at. We have. I don't know what the cost of a lost sales hire is, but it's probably quite significant.

Great insights. Let's refine that a bit for your email.

You've observed that Imparta is growing rapidly and with growth often comes the challenge of scaling the team effectively. The cause of potential issues here could be not just finding the right talent, but ensuring they align well with the company culture, which is crucial for retention and performance. The implication we can draw is that focusing on cultural fit during the recruitment process can significantly reduce the costs associated with turnover, especially in high impact roles like sales. This approach not only saves on direct hiring costs, but also protects the investment in training and integration into the team. Now moving to influence, how could we frame this message? To encourage Richard to engage in a conversation, we could use social proof or perhaps reference a common challenge in the industry that that you've successfully navigated with other clients.

So I'm going to stop there because obviously we're limited on time today, but you can see how it's kind of using its knowledge of best practice in sales and taking me through a structured process. Now, if you were going to send out 10,000 of these, you wouldn't have time to do this. But the research shows that you're actually much better off focusing, especially for B2B field sales, focusing on a relatively small number of clients that you actually hyper customize your communication to. You'll actually get a much, much better ultimate result than if you just spray and pray across many others. Now just so if I let this follow through, it will then draft the email and.

I'll show you kind of our scoring for this in a moment. Just while we're in the coaching system, let's imagine because this is capable of switching mode from help me write something to just tell me about a particular sales thing or I'm in a position and I need your help. So let's just try that for a second.

While we're talking. I'm actually in the middle of a negotiation with a very large client at the moment. It's about a $2 million a year deal and procurement are pushing back very strongly against our pricing. They've asked for a 20% price cut and 90 day payment terms, which is a nightmare.

How can I respond to that? What kind of verbal tactics do you think I can even use in that situation?

This sounds like a challenging negotiation.

Let's break it down to help you develop a strong response strategy. First, understanding their interests is crucial. Have you been able to ascertain why they're pushing for such steep discounts and extended payment terms? Understanding their motivations can provide leverage or opportunities to offer alternatives that meet their needs without compromising too much on your end. Next, in terms of verbal tactics, one.

Again, in the interest of time, I'll stop there. But you can see how it's kind of jumped on a dime from talking about.

An email to generate a prospect through to a really in depth negotiation strategy and again pulling on best practice around that particular negotiation area. So if we just run back to the slide super quickly.

When we scored ChatGPT on its own, we got 1, 0 and 1 for insight, inference and trust and the iCoach version. And I'm being fairly harsh here. When we did it earlier, we gave it 7, 8 and 6. The 6 is interesting and I think the point that Gerhard and Stephanie made earlier about, you know, AI is a tool, it doesn't really have motivations, it doesn't really care about you and therefore trust is fundamentally a harder thing for AI to generate, I think. But at the same time it can simulate it, it can simulate self orient lack of self orientation and it can definitely provide insight and influence. So how do we do this? Well, essentially we've got over a million and a half words of intellectual property and research into sales best practice that cover everything from territory management to lead gen through opportunity management, margin management, customer success, account management, core management and the whole of customer experience, sales Coaching sales management and sales leadership. So everything from how do I find, you know, how do I decide which clients to work with out of my 1000 customer territory, right up to how do I build a sales strategy. And that's built into.

A sophisticated system that's built in Microsoft Azure that plugs into an LLM like ChatGPT4T, that's the one we're using at the moment, but we can also use Claude, or we can use client's own LLM. And essentially into that is fed three things. One is our sales IP and it sits inside a secure environment in Azure. And the engine is basically figuring out what's going on in the conversation right now in a very sophisticated version of what's called rag. It's a multipass, multi level rag for those of you who are aware of AI terminology. And then it's pulling in the key elements of IP that are going to help at this moment with this conversation, which helps to reduce hallucinations as well. Then we're also, we've got very detailed role specifications that are tailored and that are dynamic around the different use cases that we have of AI, which I'll explain briefly in a second. And then finally we also can bring in client information in the same way, so you can have something that not only knows how to negotiate, but also knows your pricing policies, your price list and so on. So that's why we haven't trained the model. We're using a RAG solution because training the model, although we can do it, it's quite cumbersome and it's hard to update it quickly, whereas the client information is often changing all the time. And it's that system that human beings are then interacting with. And I loved what Stephanie said, which I think is a Gartner comment as well, which is technology as a teammate, AI as a teammate.

So when you compare that with iCoach, AI, with ChatGPT or a similar thing, whether you're writing an email, tackling a challenge, making a decision, role playing, a conversation, answering questions about sales, the iCoach output is far more tailored and far more structured. And if I just take one example of coaching, rather than just telling you what to do with 10 bullet points that are fairly generic, it will lead you through a coaching process with all of that sales knowledge, but also the knowledge of how to be an expert coach and really help to get people thinking for themselves rather than AI becoming a crutch for their, for the sales team. So moving on very quickly to talk about change, I've talked about the human change process and this is really where the AI use cases come in. And I think there's a lot of point solutions I've seen. So role plays, call assessments. For me the power is bringing it all together and tying it together with the methodology. We've been doing sales training for 25 years. I really see my role now as training salespeople and, and AI how to sell. That's our mission now. So the change process we use is assess the learning that's needed, choose the right approach, do the learning, help people to apply it and make it stick. So in the assessment stage we're prioritizing business outcomes, leading and lagging indicators, mapping those to dominant competencies that will affect those indicators the most and mapping those competencies back to the different roles in your sales team and then assessing Those competencies using AI and 180 degree assessments for both the manager and their team members. Then selecting and customizing the learning, doing launch communications if appropriate, doing the learning, whether that's synchronous or elearning. And in the eLearning we use AI role plays for practice, for skills practice. Application is AI coaching. Call simulators. I'm about to speak to a customer, I want to try that call before I do it for real. Embedded in the CRM are application tools with AI to give me coaching in the CRM, but also to help me plan and fill out my CRM in the best way. Manager and deal coaching. And then the stick phase nudge questions, so little reminders to get people past. Those of you who know about the Ebbinghaus forgetting curve recognition because it's about performance improvement as well. And then finally impact assessment that kicks me off around that cycle again. And I think it's worth mentioning that when we talk to our clients, there's always a tension between the strategic learning needs of the organization and the tactical learning needs that come up day to day. And also between stuff you want to do for the whole team and things that are more at the individual level. So the most common sales training is a course which is a strategic change for the whole team, typically once or twice a year. Maybe it's about driving people to be able to sell to a more senior stakeholder. Stakeholder for example. But you'll also have tactical issues for the whole team that might be responding to a new competitor coming in or a drop in your pipeline at the individual level. You've got tactical stuff happening day to day, which is, you know, I'm trying to win this deal, so give me contextual learning in the CRM to help me deal with that particular situation. And there's also a strategic angle for people which is help me get better at my competencies and advance my career. So get better at the skills I need for this role I'm in now, but also for the role I'd like to move into in the future. Now we call these the five Cs and there is a fifth C which is curiosity. And this is where we're sort of a bit like Netflix or Prime or something. And we're saying because you've enjoyed this, you might find this helpful as well, which just drives that, that kind of learning culture.

So coming back then to the use cases, the change process we talked about, we obviously don't have time to run through all of these, but very quickly just to touch on four. So competency assessment. You'd have a set of skills for competencies for each role with a target level for each role because there can be diminishing returns in a skill that's not really necessary to be a mastery level. And those are fed by detailed competency descriptions.

That are kind of tailored to the language that you would use inside your organization. Now that then feeds embedded in teams a way for you to go assess this call or to do a manager led role play that will assess my skills and provide me with this kind of radar plot in the middle of here's my self assessment against the competency and here's what the manager thinks, but here's what AI thinks and I just want to do a quick demo of that just to show you the difference between this when it's generic and when it actually understands the sales competency. So I've got here a sample sales dialog. It's only a five minute call that I've just pasted into that box at the bottom and I'm going to press go and you'll see I coach now run through an analysis of this whole sales conversation. Now the hard bit with call assessment is actually knowing which competencies you should be looking at in this call. So there's a multi pass process here and the first thing it's doing is going what's going on in this conversation? Because if I'm doing a detailed negotiation, it's a very different set of skills I need than if I'm trying to encourage somebody to move away from the status quo. So right now it's saying that call transcript is the customer is in the need stage of the buying cycle. So they haven't yet decided that they want to move from their current products and services to something different.

And so it knows which competencies are relevant at the need stage and in particular when they recognize there's a problem, but they don't see it as urgent. So here are the competencies. This is an abbreviated assessment based on just what we're trying to do for a demo. But this will give a full blooded, if you like, assessment that feeds into the individual's own repository of what they're good at. And you can see here it's basically going on the left is kind of what the skill is. Then up here.

Just scroll back up, stop.

Then kind of a description of what that skill is. So if we look at building momentum, building momentum involves regularly establishing the costs of not acting, loss aversion and exploring the benefits of new solutions. And it scored me as a two out of four. You attempted to build urgency, but could improve by quantifying the cost of inaction more clearly. So not only a good assessment of the skill, which any AI can do, but it's an assessment of the right skill. And that's a really key point. And it goes a long way beyond sentiment analysis and who's talking the most and did the customer look happy and what point in the conversation did you introduce pricing. It's all about really understanding what skills you need in order to perform well in this scenario. So it then gives me three areas for improvement, tells me how the customer's feeling, and this is the cool bit. It then says, look, I've listened to that conversation that you've had. Would you like to try it again? Because I can now role play that customer. And by the way, it might be more helpful if you want to role play the next conversation in the cycle so that you can practice that in advance. So there's a couple of examples of the AI kind of working. We talked a bit about AI role plays. We believe in learning through experience. So you have an experience, you reflect on what's gone well and badly.

You do abstract conceptualization, which is a fancy way of saying you think about the theory of how you could have done it better and then you experiment. Most E learning is rubbish because it gives you all the theory and then asks you a quiz question. What you can do with AI roleplays is really drive that learning cycle in a much more effective way. Then we have the coaching, which you saw a little bit of earlier, where it's taking me through a structured coaching process, which is an adaptation of the grow model, we call it iGrow. And you can use that to do all sorts of things, achieve a goal, make a decision, practice a conversation and so on. And then finally that all sits inside Salesforce or Ms. Dynamics. So that's a very, very quick run through the change process, but I think it's just worth emphasizing that it's about not having a point solution. It's about not having it sitting in isolation in its own platform, but integrating it into your existing sales tech stack and it's about tying it all together with a real understanding of, of sales methodology. Now, obviously a lot of organizations have their own methodology or have used other training companies in the past. There are ways we can work with that, partly because ChatGPT itself does know about all sorts of different approaches to selling, even if it's not at front of mind. So I appreciate your time today. I'm going to pass back and see if there's any questions in a second. Just to say it's hard to look at the impact of AI on its own away from the sales training, but through attribution, through salespeople saying this is the impact it's had and having that validated by managers. We know that the impact you can have around every single KPI that you might care about as a sales leader is very substantial.

One quick offer which is in our booth, we're very happy to. If you've got between 50 and 100 people you'd like to try this out with, get in touch, please. Thank you very much, guys.

Well, that was amazing.

It is really incredible what AI can do and judging by the comments of audience, it's.

It's mind blowing. You know, perfect practice makes perfect.

Others presentation share. They're everywhere. Fabulous. Excellent question. Any I coach tool able to work real time as the sales call is happening?

Yes. So that's actually. So right now we're integrating into teams and then Zoom and WebEx will follow the ability. So it'll list out the calls you've recorded and you can, with permission from the customer, obviously go. I want to assess that. The next step that's in beta right now, the next step which is coming out in the next certainly by the end of the year is it's kind of. One of our clients describes it as like a Jiminy Cricket sitting on your shoulder listening to the call and going in the, in a dm, a private message in the chat kind of going, look, it sounds like they're not quite seeing the urgency here. How about asking some pain questions? And at that point I think it starts to get, you know, even more powerful. It's I mean Stephanie and I were chatting earlier on, you know, I've been doing this a long time, longer than I care to remember now. But I've never been as excited about the potential for finally realizing, you know, our vision of what sales enablement and sales training can do. It's very, very exciting.

The comment from the audience, I love this. Richard, excellent presentation. And another question, do you have model for companies with 10 or less sellers?

That's a great question. We do. We're just about to launch actually. So I mean we're a full service sales training company that works with, you know, some of the largest and most successful client organizations in the world. But we also reckon that we get called a lot by people with kind of they're on their own or they've got up to 10 people. So we now have, we're about to launch two offers. One is AI only so you just have access to the AI. The second is AI with its, with the elearning modules as well. And then if you've got a bigger team you can have the full, you know, instructor led and all of our change consulting and all that kind of stuff. So it's not out yet but it will be out in the next sort of month I think we're planning to have a place on the website where you can just sign up. So perhaps we will do a press release with you Gerhardt, when we finish that. But yeah, we want to take this out to as many people as possible clearly.

Is it available in Spanish?

So the great thing about ChatGPT is it speaks. The translation quality of ChatGPT is extremely good and so we have a section in there that detects the language you're speaking to it.

And that will actually respond and then speak in Spanish or in Mandarin or in Portuguese or whatever you like, even in American.