When Decisions Multiply: The New Shape of Revenue Acceleration

So let me start here by saying hello and what I want to talk about today. I want to talk about revenue acceleration and what does it really mean? Like, how practically do you get there? Because what we're seeing out in the market is AI at this point is not new, right? It's a couple of years in from GenAI to Agentic AI. It's been integrated into most of the common seller enablement platforms or CRMs that you have. There are many AI-native applications out there, and many of you are already exploring with them, and you probably have some sort of enterprise license to a ChatGPT or Gemini or Claude. And the question is, is, you know, as we've improved experiences for sellers in some ways and we've made some things easier, have the deals moved? You know, we've rolled out the Copilot, we've got call summaries, we've got, you know, AI-assisted forecasting, but is our Is our conversion rate faster or better? Is our velocity faster? Is our pipeline bigger? Or is our cycle time of quote to cash faster? We're finding often not, it's not moving with the investments that we're making. And so we, we have a program at Deloitte called Future of Sales where we do research into what organizations are growing faster. We call them leaders at a lower cost or cost of sales. So growing faster, but doing so more efficiently. And as we look at those companies within cohorts of groups, we say like, what is it about them that they're doing different? And that then drives a lot of our strategy in terms of what we recommend. And I'm going to kind of walk you through some of that today. So one of the things that we're seeing is that deals are slowing where decisions are piling up, and particularly across people. So as much as we talk about AI, our world is still anchored in humanity. It's still anchored in people partnering together, you know, whether it's at your organization or your customers. And on the left here, we talk about the buying committee. And so it's very typical. What we're finding is that you're often engaging, or your seller is often engaging with a couple of people at the organization you're selling to. But the buying groups have continued to expand. AI is now embedded. AI and digital capabilities and services are often embedded into solutions. This is not only when you're selling tech, this happens when you're selling medical devices that come with software and consumables. It could get, it's true when you're selling, you know, financial products as well. The number of people on the other side at the customer's organization that is influencing deals growing, and you often don't know who they are. You're not engaging with them. And you're not in that decision-making process. The second thing we're seeing is that you, we still have a number of systems that we have to interact with. So, you know, whether it's how you manage your, your prospecting to how you manage your deal funnel to how you do your configuration and quoting to how you build your BOMs, if you sell widgets of some kind that are complicated, to contracting processes, you know, to then how do you get an order? How do we manage subscriptions? Understanding our install base. Not to mention where do we keep our sales enablement platforms and our marketing content, you know, and then we have our collaboration systems. We can go on and on. There, there, there is no single pane of glass for most of us in the work we do. And the transactions from, you know, engaging a customer to that customer signing up often traverse these processes and people and we get stuck in the middle. And so the, The 5 to 7 days from a verbal to a proposal can be even larger when you're selling a large complex deal. We have clients that take 30 days to get a proposal and a quote done because there's multiple business units, there's solutions, there's hardware, there's software, there's services, and sometimes there's other partners involved. So what do we do? We accelerate the deal. So rather than thinking about accelerating the seller, How do we, how do we flip that and say, how do we introduce AI that doesn't make the seller more efficient only, but how do we actually make the buying experience better for the customer? How do we make the deal progress faster? And there's 3 things we're finding that are really delivering value right now. One is how do we engage the committee? How do we engage those 10 people that are involved and make sure that we are selling to them? We'll talk about that in more detail. Two, how do we progress the deal? So understanding the insights from the deal and all of the signals we're getting to what do we do next to move the deal forward? How do we move it along? And lastly, how do we accelerate the quote-to-cash cycle? Because the configuration-to-revenue cycle continues to be an area of pain and we're seeing it across industries and sectors. And ultimately, we want to manage the velocity. We want to manage the velocity of reaching consensus. We want to manage the velocity of moving the deal through its stages, and we want to manage the velocity of getting to an order, right? So let's take, take each of those and kind of talk through them. So the first one here, you know, your calls and emails, all your interactions you're having with a customer often name the committee, but we're not capturing it in any way. And what we mean by that is you're on the phone with a prospect, maybe you're talking to a customer about an expansion or existing prospect, and they'll reference other people in the organization and the roles that they have. And your, your top sales reps who are really good at relationships are going to write those names down and then they're going to go research who knows that person. Do we know them? Does somebody in my organization know them? Is this person friendly? Will they make an introduction to them? What role do they have? What school do they go to? And they'll figure out how to engage them. The rest of your sales organization may not have even heard the signal, right? The— so, so there's 3 things to, to talk about here. One is who's really in that committee? Do we even know who they are? How can we map who these people are, what the influence is, and how we engage with them, right? Some of that is known to us because they're either engaging with us or they're talking about people, and, and others we can go get signals from by getting third-party data, and I'll, we'll talk about how we do that. We'll show you an example. Second is who's quiet? So are, who are we engaging with? What is the sentiment of those conversations? Are we moving the deals along fast enough, making sure that we're really matching the engagement cycle with our customers at the velocity we need to? And lastly, making sure who we're actually, who's selling for you at an organization. If there are 10 people in a buying committee, there's usually a champion who's saying, I've spent time with this vendor, I understand the solution, I think it's going to create value, let me show you how. Are we arming that person, right, who's our seller inside our customer with the information that they need, the help that they need to do that? That's often not proactive enough. The second play is one that is now really accelerated by AI, right? And so we call it— you can call it the deal copilot or the deal assistant, or how do we measure deal progression? But number one, it's capturing data. How do we capture our signals? So we've got signals from calls, meetings, emails, text messages, and also third-party data, right? Intent signals, you know, what they post on LinkedIn, who they're hiring, role changes, those matter. Let's capture all of that. Now, when we get, when we think about capturing that, my clients tend to have different points of view on this. Some of them are capturing the data centrally and they've introduced capabilities like Gong where they're really looking at that data as, you know, enterprise data that will create enterprise value. Others are helping to create seller assistance where sellers are at least capturing that data in their own ecosystems. All my calls, emails, files, context is available to me as a seller and making that seller as productive as possible and then helping them get it into CRM. And there's a culture thing there that's different, but you have to capture the data once you've captured it. Now it's what's the health of that deal? What's the risk associated with the deal? What's my next best move and how do we actually go get that done? So we've moved from CRM being something that we use to enter data, to capture information that the seller wants to share with us so that somebody has a forecast, to how can we actually do the work for the seller in terms of running the pipeline analysis, figuring out the next step, drafting the emails, creating the tasks, collect— creating content that we're going to sell to them, drafting an outreach campaign or account-based marketing campaign. AI can do this work for us and then it can pull the right people together in your organization, whether that's your marketing team, your sales team, your engineering team to, to act on that data. Number 3 is our quote-to-cash process. So this is where deals are stalling more than ever, right? So we have an opportunity to leverage AI and collaboration here also. To really get our timelines down. And then on average, we're seeing, you know, from 5 to 7 days down to 24 hours, or from 24 hours down to 4 hours, or from 30 days down to 4 days. And then, and those are just different sizes of companies and different complexities of deals. But when we implement this, we're starting to see real value back to the customer and back to your EPS. So if you look at configuration, for example, We don't have to configure from scratch. We know so much about the deal if we actually capture all the signals we talked about. We know their pain points. We know the kind of solution where we are communicating with them. We know their requirements to some degree at least. And we know companies like them typically need what kind of solution. We can draft a configuration from that information without a seller having to do anything. When we price, common challenge my clients are having is that the sellers will always say, My client is getting a better deal from somewhere else. The price that the system is giving me is not the right price. I need a bigger discount. So one of the things we can do now is explain to sellers why a certain price is the right price or what different pricing ranges are based on what information. So what are the typical discounts that we're actually seeing approved and sold at what velocity by, by what kind of company? What is the market data telling us the street price is? We can provide a lot of contextual information in natural language back to sales reps during a pricing process to not just to say this is the price, but look, this is a good price, a better price, the best price, and why and how you can communicate this to your customer. And we're finding, you know, a couple points of margin improvement is having massive impact, especially when you're in a cycle where deals are stalling or your pipeline is that big, that, that couple percent on margin matters a lot. Then you get into finance approvals, contracting. And so if you're not using AI around contract management, redlining, your legal process, automation around all of that, there's a real gap to be filled there. And you can go all the way from your quote and the requirements associated with it to the price to the contract in a, in a singular motion. And whether you have one single system that does all that naturally for you, or what most of our— my clients have is not a single system. Now you can orchestrate that, you know, with AI and do it across your systems and get value from it without having to reimplement all of your tools. So one of the questions that comes up as we talk about that is like, okay, and we run these feature sales roundtables with CROs and question when we talk about this comes up is like, how much do I let AI actually do? You know, what are we comfortable letting it do? And that also becomes a cultural question. There's not a single answer, but, but the reality is that there's no single answer for every workflow. It's different by workflow and governance, regardless of whether AI is assisting a human, which is the left column there, or AI is executing and running a process, but it's deferring to humans for exceptions or for anything that requires risk. Which is sort of semi-autonomous, or AI is really running the process. Governance sits in all of them. They just sit at different levels because in a fully autonomous system where AI is running, say, a prospecting process, you're building governance at the system level. You're defining what that agent does, what it's allowed to do, what data it uses, who it can reach out to, and when it should stop, right? On a human assist, it's the other end of it where the AI is not doing anything on its, on its own. It's providing information to help a human do it. And semi-autonomous is somewhere in the middle. And we're finding is there's common things, right? So it's— you're talking about next best action and meeting briefs and post-call summaries. These are assistive capabilities to help your seller, right? On the very right end of it, you got volume renewals, for example, or SMB outbound prospecting or order processing where an AI agent can actually run the process for you, right? So it's helpful to think about the risk to, you know, to the risk associated with the workflow and how you actually design where humans are in the loop. So how do you build it? So if we think about 3 layers of the architecture, at layer 1 at the top is the engagement layer. It's how do you actually engage with your AI? And we'll talk, you know, I'll show you an example and we'll talk a little bit more about it. But at the end of the day, where does our work get done? Our work gets done in emails and phone calls, in meetings, in personal conversations. That's often the best work we do, right? And so the tools to do that work are Outlook and Gmail and Teams and Zoom, you know, and Webex, Slack, right? These are the places, you know, and our phones, right? Those are the places where we do a lot of our best work. So what we are doing these days is bringing AI to those interfaces, right? Rather than say you have to go to your structured workflow and your applications stack to go do all the work. How can we actually bring the work to you in Slack, as an example? And if you're following along this week and you've seen Claude Tag, it's a great example of right in Slack being able to call your agents to go do work for you as a digital worker. That's the future of how we're going to interact with AI. The, the middle layer in order to make that possible is the orchestration layer. And, you know, the way that we do that is we build an open-source orchestration platform. We have an open-source sales-specific orchestration platform that anytime we deploy these solutions, we give to our clients and, and it's open source. We give them the IP, you deploy it on your own cloud, but now you can use any LLM, you, it stitches your data together, you hold your context. And so regardless of which cloud you go to, regardless of which LLM you go to, you're retaining the, what matters most, which is the inside of your organization, which is really important that you create a recursive loop that learns and self-learns and captures your own information that then allows your digital workers to do work to help your humans. At the very bottom, you have your systems of record. Those are not going away. So you still have, you know, your Salesforce, your CRM, Sales Cloud, you might have your Salesforce CPQ or ARM product, you might have their billing product, and you might have SAP, Oracle ERP. Those are continuing to be systems of record. They also have agentic platforms, and you're probably going to build some agents on those platforms, whether on AgentForce or Oracle AI Studio, that solve problems for data that's sitting on those platforms or data you've brought into those platforms as well. And so, so that, that's kind of the layer cake that we see as we go through this. So where do you start? Number one is build your, build your adoption muscle. Make sure that you have Assistive AI capabilities deployed to your users. People need to know how to use AI and how to use it well, and that comes with practice. So start to get it deployed out to users. I think most of us are already there. If not, that becomes very important to do. It could be Claude or Cowork or Copilot or GPT, but you, you should deploy AI tools out there and connect them to, to some of your data that's low risk. 2, we would recommend building that orchestration layer and data context. So own your data. Make sure that it's in a place that it's portable for you and that you start to build context. AI is no different than any other worker. When you hire a human, that human is productive because you give that human tools like a laptop and access to the building and a desk. And you give it— you give the human context, like what work do we do? What's the definition of good? Who should you work with? What can you do and not on your own? And when do you need approval? We, and with tools, context, and their own knowledge, humans are, are, are amazing at the work they do. AI is actually not that different. It comes with this capability, but if you don't give it tools in the form of, you know, APIs and MCPs, right? And, and, and actions that it can go take in your environment. And if you don't give it context, like what is your work, it doesn't do very well. So we would build that. And then lastly, get your, your application's API and MCP ready. So where do you have your data? What applications do you have? Can they write— can they read and write back to whatever your, your AI orchestration layer? If not, it's worth building that or migrating so that you have that. And then go work on your processes. And we— where do, where do your deals actually slow down? What KPIs are impacted? And what data do you have available? That leads us to where should we start? And then you work in 12-week sprints where you launch capabilities, you deploy them out to markets, you run A/B tests, and you measure the results. And these are not big programs. These are small pods of like forward deploy— we'll have a manager and a couple forward deployed engineers that will do this as an example. And so it's weeks and not quarters where we used to, you know, do a project to go build a plan and size it and then do design and then do development and then deploy it. Now we're talking about working in 12-week cycles, and it really accelerates the time to value, and you learn much faster as you go through this cycle. When you're done, your Mondays should start to look different for your users, right? Your sellers should have reclaimed time. They should be able to walk into the day understanding their priorities, which accounts are heating up, you know, which ones are at risk, which ones do we advance and how do we do that? We should be getting real-time coaching now before and after our calls. We should have, you know, quoting done by our agents and CRM should stay current without the sellers having to enter any data in there. Now the era of CRM for managers is gone. It's now AI leading sellers to actually better outcomes. Managers should be understanding their, their risks and pipelines and where to focus, like which sellers to focus on, which dealers, deals to focus on. They shouldn't have to figure that out. It should be given to them. They should also know how to coach their teams much better because we know a lot more about their teams and how they're performing. As well as be able to understand who's doing well, like what, who's, who's doing something that we should go amplify, right, as well. And then, and then lastly, managers should no longer be in the forecasting process the same way they have been in the past where they were crafting the forecast and trying to understand these signals and trying to do their best to give a forecast that's, you know, accurate but not too high. We want to take sales out of that process by making it more, more agentic. And what we want from our humans is to give us insights, like to review it and to say why they agree or disagree. And lastly, don't forget about your RevOps team. So how you manage, you know, your accounts and your territories, your assignments, and the continuing shift of your market on a much more continuous basis is an area for AI to lead. And then as you deploy all this AI, your RevOps teams have new jobs. They're now managing, building, managing, and monitoring all of this AI. They have to make sure all of the data is actually governed. So data governance and the unglamorous work of data governance becomes very, very important to being able to execute on all of this. And leaders, where I started in our future of sales research, is leaders are winning because of it. They are 50% more likely to be scaling AI, 8 times more likely to be getting value. It says 5 here, it's actually 8. Twice as— their sellers are twice as productive for the same kind of companies and their cost of sales is significantly lower, like 15% of cost of sales versus 24% for companies that are actually growing faster as well. So the results are in the— the proof is in the pudding, if you will. So lastly, I'd close here with, you know, when the— where the deals flows. Fix it in the order where the deal flows. Like, understand where is your deal slowing down, where is there opportunity, and start there. And I just want to give you a 2-minute version of an example of what this looks like. And so this is a composable front end, meaning a sales, you know, application for sellers that's built right on top of the orchestration layer that we talked about. And underneath that is Salesforce CRM data and ZoomInfo data. Other data sources across a company. And so when the seller drops into their day, they know right away, okay, what are my priorities? And, you know, I've got a deal at risk here. I need to go find a decision maker. This was an important one installed. What, you know, what approvals are needed from me? And they can click in and kind of go through this. They, as they, as they work through it, they can actually have the AI do work for you. And so in this case, you know, can be done by your organization rather than by the seller themselves. But then they can go run their pipeline and say, okay, go work my pipeline. And as it works your pipeline, you know, it can actually go take actions for you, like whether it's tracking activities and logging them, whether it's queuing up things for you to go do. And so you're waking up now to actions that your Copilot's actually running for you. And then you can start to look at, okay, how, you know, how do I actually manage my deal cycle? So for example, if I went back to my briefing and I said, okay, let's go tackle this activity. So now it's going to draft a re-engagement email for that FY26 stocking order. It's going to go pull context. It will create an email. And then I can go, you know, take a look at that email draft. So Lastly, we talked about relationships, and so I want to show you a little bit about that. This is an example here of a relationship dashboard where you've got an understanding of your actual decision maker map. The, the people are hypothetical, but, you know, it's pulling from, it's pulling from data sources that we've masked. So you've got, you know, your decision maker map. Who's your economic buyer? Who's your champion? Who's your influencer? Who are the key people in our organizations? Where— what's mapped and what's not? Who do we need to reach out to? Let's plan an outreach, which will take you back to the, the agent, which will go run a plan for you based on research of who this person is and how we know them. And then it'll also prompt you with like, who can we reach out to if we need Marcus Webb? Your colleague actually knows Marcus Webb. We, we ought to reach out to them. They can draft an email and reach out to them directly. So it's a different way of working where you've really got your AI doing your work for you to drive your deals forward rather than CRM being something where you've got to enter data so your manager can create a forecast to run a forecast meeting.

Thank you. That is incredible. I got a question. This is a real behavioral change for an organization to undertake. What are the cultural issues that you see organizations going through adopting this process?

It is a great question. There's two things that we're seeing. Your top performers are saying, I don't need this. Like I'm really good at what I do and actually I don't want my secret sauce to be stolen by some AI and replicated. I don't need this. So for them, it's really about helping them understand how it can be a force multiplier against what they already do well, right? Nobody's going to take away what they do well. The sellers are not being replaced anytime soon. So demonstrating wins to them matters a lot. Celebrating when great performers are even greater because of it and showcasing that works a lot. The middle performers is a little bit of a different story. They may over-rely on it. And so if you've heard the term AI slop, a lot of AI slop happening with the middle performers where they're just sloppy emails being sent. Some of you are probably receiving them, you know, proposal decks that are not thought through. So there is a new muscle to how to use AI and how to do it well. At the end of the day, it can help you think and organize your thoughts. It can draft something for you, but you have to own every word of what comes out of that. You And so that also, Jeff, requires training. It requires real engagement. And I'll tell you at our own, you know, at Deloitte, same issue. We have 300,000 people. We've gotta get trained up on this. And there's a range of people that are, you know, range of capabilities. One of the things that's worked really well is informal sessions. So we have an AI water cooler chat and every Friday we get on and people showcase things that they're doing and how they're working. And those ideas spread. So the, the, it's a very human thing, right? Like we learn from other people better than we learn from formal trainings that you roll out. And so there's a lot to be just, it's funny, but if we, if we're better humans, we'll be better at AI.

I like what you're saying there. AI slop. What, because again, you can let it do everything and it's going to run away from you and you're going to be in trouble.

Yeah. Gerhard, jump in.

Do you think that some companies like midsize companies are using AI to replace consulting consultants?

It's a great question. Existential for our business. Yeah, to some extent, I think if AI answers a question that you needed a consultant for, and it does it well and it's cheaper, use the AI, right? Like I said, the obvious thing to do. I, I think we're in the business of value creation and transformation and value creation and transforming a business will always be true. But the way we do it and the work that we do is fundamentally different than the work we did 2, 3 years ago. So I think consultants that change how they work will continue to survive, and the consultants that provided PowerPoint decks of information, you know, may be in a little bit of trouble.

So what is the single most biggest challenge that you see when you work with clients? You know, we, we talked about the complexity of making decisions that on one hand you have so much information, on the other hand, you have more people involved in the decision. So how does AI deliver more clarity, reduce complexity, and accelerate decision-making?

I, I, I think AI is not fixing that yet. I think in the organizations that I work with, there are, everybody's racing to go get value from AI. And they're making decisions that are sometimes in conflict of each other, right? They're— they'll have applications that do the same thing at different parts of the organization, not thinking about the fact that what's powerful for us now is data. So I think where it's working well, the companies are hyperfocused on the most boring things. They're hyperfocused on data governance, they're hyperfocused on data ownership, they're hyperfocused on end-to-end processes and bringing humans together across them. They're actually getting the biggest benefits of AI.

That's awesome.

Excellent conversation. Boy, thank you so much for sharing. And, and again, I, I— behavioral change at its utmost. So hey, we got what, 9 minutes, uh, a break here, and then coming up here in a bit, David Tallon, founder of Tech Contracts Academy, is going to walk us through. Abe, this has been absolutely fantastic. How do people get in touch with you?

LinkedIn, Abe Awasti is the easiest way. And I'll just throw up my email here. And, you know, I would love to interact with anybody on this topic. I have a new playbook that we can share all around AI and how to, you know, what the AI empowered buyer means for sales and how sales leaders can deal with it. So I'm happy to share that with anybody who wants to reach out.