What happens when AI starts selling to AI?

What happens when AI starts selling to AI?

Episode 377 What happens when AI starts selling to AI?

Jun 2, 202638 mins

Generative AI

Procurement Software

Salesforce Automation

Jump to

Read transcript

Overview

AI is already writing sales emails, updating CRM systems, generating proposals, and responding to RFPs. But the next phase could be even more disruptive: AI agents negotiating with other AI agents before a human ever joins the conversation.

In this episode of Today in Tech, Keith Shaw sits down with QorusDocs CEO Ray Meiring to explore how AI is transforming enterprise sales, procurement, and business relationships. They discuss the rise of agentic AI, the future of RFPs, and why many companies may soon have AI handling research, qualification, and due diligence automatically.

But if AI can manage so much of the process, what happens to the human side of sales? Can an AI build trust? Close a deal? Read a room? Or will empathy, relationships, and human judgment become even more important as automation spreads?

The conversation also examines the risks of overtrusting AI, the future of enterprise buying decisions, and why trade shows, face-to-face meetings, and old-fashioned handshakes may not be disappearing anytime soon.

Register Now

Transcript

00:00

Keith Shaw: AI was supposed to make sales faster, smarter, and more efficient. But for many companies, it's creating new layers of complexity, new trust issues, and a new question: what part of sales should actually stay human?

On this episode of Today in Tech, we're going to look at how AI is shaping the world of sales relationships and trust. Stick around. Hi, everybody. Welcome to Today in Tech. I'm Keith Shaw. Joining me on the show today is Ray Meiring. He is the CEO of QorusDocs.

Welcome to the show, Ray.

00:36

Ray Meiring: Thank you very much, Keith. Thanks for having me on here. Excited to be with you.

00:40

Keith: All right, so let's talk about what you're seeing in the marketplace right now. What is AI actually doing for sales reps?

Everyone says AI is going to save time in sales, but out in the field, what are your sales reps or other sales reps you meet with actually spending their time on?

00:58

Ray: Today we do see some significant time savings taking place right now. Great tools like call recording software can automatically update your CRM records or provide deal updates without having to interview a sales rep for status reports. Those are significant time savings that sales reps are seeing.

Even research and follow-up tasks are being streamlined. What is it translating to? I think it's translating into reps having to cover more deals and more accounts. They have less time to spend on each opportunity, so they're being asked to manage more pipeline activity overall.

01:44

Keith: Are you seeing AI actually reducing workloads, or are we just shifting workloads from entering data into CRM systems like Salesforce into more prompt engineering and AI management work? Or is AI actually adding more work because reps are expected to handle more leads and contacts?

02:10

Ray: I think it's just a different workload. At Foundry, for example, sales reps used to manually update all of their MEDDPICC fields inside the CRM.

Now we have AI agents listening to calls and automatically filling those out, and reps simply review them to make sure the information is accurate. We don't really see reps becoming prompt engineers. If prompts are being created, it's usually handled by sales operations teams behind the scenes.

The reps themselves are mostly focused on talking to more customers, joining more calls, and having more conversations while AI handles many of the administrative tasks in the background.

03:10

Keith: So AI isn't necessarily replacing the sales rep. It's more like a support system or sales administrator. Ray: Exactly.

03:17

03:18

Keith: Are the reps you talk to embracing that? In a lot of companies I've worked with, sales reps can be resistant to technology because they're relationship-driven people. The administrative side always felt like busy work to them. Ray: It's mixed.

03:52

Some reps are adopting AI heavily while others are slower to embrace it. But most sales reps want to focus on relationship-building. If you're taking administrative work off their plate and automating CRM updates, they're usually excited about that.

Where you'd see resistance is if AI started doing demos on their behalf or replacing the relationship aspect of the process. That's not something we really want anyway because the person-to-person connection still matters.

04:53

Keith: Especially as we move into agentic AI workflows. Right now AI mostly exists in chatbot form, but eventually you might tell an agent to handle outreach or even make deals for you. What kinds of friction points are you seeing as AI becomes more autonomous?

05:37

Ray: Agentic AI is definitely going to take over more of the cold outreach process. It's already writing introductory emails and handling scheduling. One of the key differences with agents is that they're proactive and autonomous.

An agent could listen to a customer call and automatically send a follow-up summary with documents, case studies, proposals, or commitments that were discussed. It's still doing work in the background, but I don't yet see buyers getting on calls and negotiating directly with AI agents in enterprise B2B sales.

That may happen in more commoditized B2C environments, but enterprise sales still relies heavily on human trust and relationships.

06:38

Keith: There are obviously different types of sales. Consumer purchases are very different from large enterprise deals. One area you mentioned before the show was the RFP process. How is AI changing enterprise RFP workflows?

07:18

Ray: A few things are happening. First, companies are issuing more RFPs than before. Second, response timelines are getting shorter because there's an assumption that AI can help generate answers more quickly. AI is actually very good at handling large volumes of repetitive questions.

That's one of the natural use cases for it. So we're seeing more RFPs, tighter deadlines, more questions, and heavier reliance on AI to source answers.

08:11

Keith: Are companies still using spreadsheets, PDFs, and Word documents in those workflows, or has a lot of it been automated?

08:21

Ray: They're still using those formats because systems aren't fully compatible yet. But I don't think that's the future. Right now people still have to export data into Excel or Word and manually interpret it, but I think that's going to change significantly over the next five years.

08:44

Keith: You also mentioned that AI is generating some of these RFPs while AI on the other side is responding to them. At some point, why do we still need humans in the middle? And if so, why do we still need them?

09:09

Ray: That's the really interesting part because you've got AI on both sides of this process, and then you've got this human and document-based interface in the middle. That's where we see the biggest changes taking place.

We believe that eventually a buyer-side AI agent will communicate directly with a seller-side AI agent to answer transactional questions. That will become the interface. But keep in mind that there are usually two parts to an RFP.

One part is transactional questions like, "What's your address?" "Do you have ISO certification?" or "Do you support data redundancy?" Those are straightforward yes-or-no questions. The second part is persuasive: "Why are you the best choice?" or "Why should we award you this business?" That's still fundamentally human.

That's where judgment, storytelling, and relationships matter. So I think agents will increasingly handle due diligence while humans focus on building trust and making the case for why their company is the right choice.

10:51

Keith: Someone still has to read and evaluate those proposals though. At the end of the day, a human still has to make a decision, right? I always joked that a lot of enterprise decisions were made on the golf course or over handshakes and personal relationships.

Maybe that's naive today, but I still think humans are going to stay involved.

11:40

Ray: I actually think we're going to see more handshake deals and more relationship-driven selling. AI will strip away the transactional pieces, but people still buy from people, especially in enterprise sales. We're already seeing this at trade shows and conferences.

People are showing up having already done their online research. They're not there just to gather information anymore. They're there to build relationships and determine whether they trust the vendor.

12:57

Keith: That's interesting because when I used to cover trade shows, it felt like customers were there mainly to do research. You're saying people are now going more to make connections and potentially close deals. Ray: Exactly.

13:14

We recently attended APMP, the Association of Proposal Management Professionals conference, and most people who came by our booth had already researched us beforehand. They knew who they wanted to talk to before they arrived. They came prepared and were really there to build connections.

13:58

Keith: Getting back to RFPs for a second, do you think the idea of an RFP eventually goes away? Maybe we don't even call them RFPs anymore. You just tell an AI agent what you want to buy and suddenly you get 20 responses back the same day.

14:24

Ray: I'm not sure what we'll call the process in the future, but I think the process itself evolves. AI agents will handle the initial research, qualification, and due diligence. Then once a shortlist is created, humans step in for relationship-building and trust-driven conversations.

Today all of that exists inside one giant RFP document. In the future I think the process becomes more segmented.

15:49

Keith: You also mentioned before the show that companies are shifting from storing answers to storing facts. Companies now have years and years of deal data, and AI can finally analyze all of it. Explain that a little more.

16:30

Ray: Most organizations have massive amounts of institutional knowledge buried inside years of RFP responses. Before AI, someone had to remember the right keyword and know where the document was stored in order to find that information.

Now AI can analyze huge amounts of historical data and say, "You answered this same question in 2020. Here's the answer you used." But facts change over time. In the past, answers were tied to Word documents, PowerPoint presentations, or spreadsheets.

We think the better approach is to store the facts independently from the format itself. That way AI can reference the latest verified fact first, regardless of whether the final output ends up in Word, Excel, or another system.

19:12

Keith: Does that require companies to open up all of their data to AI systems? That's something that worries a lot of organizations. AI works best when it has access to everything, but a lot of people aren't comfortable giving AI unrestricted access to all of their company data.

19:42

Ray: There's definitely a tipping point where too much unstructured data actually becomes a problem. If you throw every historical RFP at AI, you'll end up with conflicting answers and confusion.

That's why we encourage companies to narrow the scope and focus on what we call "gold standard" responses — the best RFP responses that actually won deals. But eventually, I do think AI systems will gain broader access to enterprise data through technologies like MCP servers and other integrations.

Over time we'll have to build trust around that reality.

21:09

Keith: I want to shift to the human side of this discussion. Before the show we talked about how AI fits differently depending on the complexity and risk of the transaction.

For example, if I'm buying a pair o

[truncated for AI cost control]