So What? Marketing Analytics and Insights Live
airs every Thursday at 1 pm EST.
You can watch on YouTube Live. Be sure to subscribe and follow so you never miss an episode!
In this episode, CEO Katie Robbert, and Head of Business Development John Wall dig into how deep research is reshaping what’s possible when you’re trying to close a deal. They walk through pulling up hidden prospect data before you even start drafting a proposal, and how weaving deep research and the sales process together gives you real clarity on budget ranges and who’s actually got the authority to say yes. Rebuild your sales process around this approach and you can cut weeks off pitch creation, while still landing the exact answer every client is asking for. It’s the kind of shift that helps your team pick up on subtle buying signals and outmaneuver bigger competitors who are still working off guesswork.
Watch the video here:
Can’t see anything? Watch it on YouTube here.
In this episode you’ll learn:
- The best practices of deep research
- How deep research is changing the sales process
- Next steps you can take to enhance your sales pitches
Transcript:
What follows is an AI-generated transcript. The transcript may contain errors and is not a substitute for listening to the episode.
Katie Robbert – 00:18
Well, hey, everyone. Happy Thursday. Welcome to So What?, the Marketing Analytics and Insights live show. It is me and John, and that is it. We are holding down the fort this week. Yay. What could go wrong?
John Wall – 00:30
Yeah, right.
Katie Robbert – 00:33
This week, John and I are talking about deep research and the AI sales process. AI has changed a lot of things, specifically the sales process, because there’s a lot of opportunity now to really dig deeper. I’m seeing a lot of companies really use the features that come along with having a generative AI system to do more advanced modeling and more advanced analytics.
Last week or the week before—I don’t remember—I was working on filling in our CRM to really try to understand buying signals. That is something that is important, but based on sheer bandwidth, resources, and ability to get to the data, I couldn’t do it until now.
So, John, before we get into what’s changed, I want to get your perspective. You’ve been doing sales for quite a while. You know sales inside and out. Give us the basics of what you’re seeing in terms of the challenges with sales today and where AI has complicated things.
John Wall – 01:45
Things are always in complete disarray and a disaster when you’re looking at sales because so much has changed. The biggest shock—I mean, we’re still feeling the impact from just online sales—is the fact that people can go online, check reviews, get to 98% of purchase, or even disintermediate sales entirely, just go ahead and close the deal and move on.
Then it gets more interesting. In the stuff that we do, there are all kinds of levels of where it happens. For us, just for our case specifically, we’re doing a lot of product-market fit. There’s a lot of product marketing going on in our sales cycle because pretty much every deal has some custom thing that makes it different from the previous customer. We’re absolutely not selling widgets. We can’t automatically predict all of our sales based on who’s coming in the door and know how the sales process is always going to go because there is always a wrinkle every time.
AI has a whole bunch of things there, too. Automating a bunch of this process is interesting, yet fraught with all kinds of risk. For us in our situation, it’s not even really applicable in a lot of situations because AI would just run you down some hole and promise stuff that we can’t do or can’t afford. It would just go crazy.
There’s a really fun and interesting section where research tools are better than they’ve ever been. That’s the part that interests me the most: having AI be able to see and understand more. When you get down to it, for us, it’s a poker game. That’s one way to think of it—game theory—in as far as we only have a few pieces of data and we understand where we’re trying to go, but we really can’t see everything that’s going on in an account.
We always fall back on BANT: budget, authority, need, and timeline. If someone has all four of those, then we consider them a real prospect that could buy. Even at that point, we still need to figure out what’s going on.
Having AI do additional digging—it’s not as good as card counting in poker or blackjack, where you can guaranteed increase your odds of success, but you can at least do a much better job and get to more data. There’s also the risk that data may be wrong or fake. We’ve been dealing with this. It’s even funny, the things that we’ve run into where it led us down a path, and then we get there and we’re like, “Oh no, wait, this is totally the wrong neighborhood. This has nothing to do with us.”
My answer is just rambling because we still are feeling where this is going and what’s going on. But there’s been a bunch of stuff, and you’ve played with a bunch of stuff as far as being able to unlock things that used to be behind paywalls. We really couldn’t do much of anything with any of our data, and now you’re able to do some interesting stuff. So, we’ve got a bunch of stuff to talk about.
Katie Robbert – 04:44
One of the examples that we’ve been working on more recently is cleaning up our CRM data. John, this is something you and I have been working on. My goal in doing that was to understand at a high level what’s working—what are we doing as a company that brings people in the door where they say, “Yes, I want to buy from you”?
Up until recently, we’ve had a good sense of direction, but it was really more that we think this is what it is, but we’re not 100% confident. If somebody came to us and said, “What is it that you’re doing that’s working? I need to know,” we couldn’t say for sure what that was. We could say, “It might be this. It might be this.” We were close.
Using generative AI, because generative AI is really good at pattern matching and finding gaps in data, you—the human—still have to be the person in charge making those judgment calls. I was able to start to pull together more pieces of data that lived in a few different places in our CRM that I couldn’t look at before. That was helpful to start to narrow down what’s working and what’s not.
When I brought it to you, John, the human whose job is the sales process, you looked at it and said, “There are some issues with this.” I said, “Lay it on me, tell me what it is,” because AI can only take you so far. It doesn’t know what you, the human, know.
One of the limitations with what I was able to do is that it took us pretty far. It probably got us about 90% of the way there, whereas before we were maybe at 60%. But it can only see what lives inside that system that I was working with, and there were pieces that live before that system.
When we talk about sales, we’re talking about your sales funnel, which is your top of the funnel—this is awareness. Then you have the middle of the funnel, which is consideration and evaluation. Then you have the bottom of the funnel, which is where people purchase things. We were struggling at the top of the funnel to make sense of how people are coming in the door.
We have data in a bunch of different systems that tell us that, but nothing that really connected our Google Analytics to our CRM. Google Analytics feeds into the CRM, but once it’s in there, does it connect back to the deals, or does it still live in the contact? It was all over the place. I’m sure a lot of people have a lot of the same challenges with having a larger tech stack that maybe wasn’t set up the way that you intended to use it. That’s where we fall.
What we found out was that it looks like certain things that we do are what bring people in the door. When I presented that to you, John, you said, “That’s cool. Close, but not quite right.”
The data that we cleaned up alluded to doing events not really leading to people buying things. It’s just great for awareness, but nobody buys anything. You rightly pointed out that when we get paid to do events—like if someone buys a custom workshop—that’s actually revenue and money in the door. Why isn’t that factored in? That’s not something that AI is going to go, “I bet this is what this means.” That’s where that human judgment comes in.
It gave us the opportunity to look at those pieces. In terms of AI changing the sales process, it’s getting us closer to the answers that we’re all struggling to get, but there still has to be human judgment in it.
John Wall – 08:24
Absolutely. There are a lot of cases where these arguments are still the same problems you get with attribution models. We say this came from the show, but then you can dig a little bit deeper and find out, “Yeah, but we were invited to that show because somebody met somebody at this other thing.” There’s never a bottom to the well; you can keep digging.
Another thing that’s interesting that you can always see and act on is campaigns where you’re 99% certain that you have never seen any business from them. We have some free resources that people will download, and we know that if they’re downloading those, they don’t have the authority to buy. If they’re looking to get their LinkedIn profile updated, obviously things are not going well where they are today. We know that they don’t have decision-making authority. That breaks BANT, and they’re off the list.
There were a bunch of other things that came out of that. The big one, though, is that you definitely have to give a shout-out for having AI clean up the data. There’s so much stuff in there because we have all these automated programs running. There are thousands of leads a year that go into the system, and nobody has ever taken the time to connect the dots. Like, “Okay, we’ve got some new people in from this Fortune 50 company. How are they related to the other five divisions that we’ve talked to over the last five years?” Doing that kind of stuff is really important.
As you’re thinking about applying AI research to your process, think about applying it to each of the individual BANT steps. To determine authority, have queries where you’re asking, “Okay, here’s my person. Where do they fit in on the org chart? Who are their superiors? What divisions are they in?” Basically: can they buy this thing? To get some insight into that is key.
We’re already running ahead. I know we should have hit the 5P Framework before we started talking about doing day-to-day tactics and where to go, so I’ll let you run with the deck here.
Katie Robbert – 10:42
As the subject matter expert on our team, I want to hear from you. What I’ve done is put together some thoughts on how it’s changed our sales process, and then I have some real examples. I always enjoy hearing your perspective on it and how closely it aligns with the work that I’m trying to do.
I always say I’m not a salesperson, but we’ve had this company for almost nine years. I’ve done sales; I know what sales entail. I need to stop saying that I don’t know how to do sales. I just don’t have all the vocabulary. I wouldn’t be able to come up with BANT off the top of my head or some of the other frameworks.
BANT is one of the best examples because one of the things that we surfaced, and that you rightly called out in the data that we cleaned up, is that our product is ever-evolving. We’re not just selling the same widget over and over again where we can say predictably, “If we just do this campaign, this is what happens,” or seasonally, “Everybody wants this thing in October.” Our services have to evolve with how the industry and the markets are evolving. That’s one of our biggest challenges.
That’s where I really wanted to dig in. Is there a better way for us to be scoping when someone says to us, “Yes, I want to hire you, but first put a proposal in front of me”? We have a decent win rate, but we also have a loss rate. We’ve been digging into why we’re losing, and a lot of people are now turning to RFPs—requests for proposals—versus a, “Yes, I want to go ahead and just hire you.”
In talking about that internally, the assumption is that more companies are turning to RFPs because they’re safer. You can get more bids for the same project, and you can weigh them against each other. The decision is made generally by committee versus one individual who would be on the line for saying yes to something that maybe didn’t work out.
Now, if you are on our side of the table, an RFP is a pain in the butt because you know very little about what they want. They’ll give you a list of questions and things that you have to fill in, but very rarely are they telling you, “This is how much money we want to spend, and this is who’s making this decision.” That BANT framework goes out the door when you have an RFP on the table—until now.
Where we started with using generative AI for our sales process—and admittedly, I did not do a great job; I kind of threw it together, and we’ve been using it, and it’s been okay—I put together a scope of work skill in Claude and said, “This is our template. This is how we fill it out. This is what we need based on historical conversations, our transcripts, and our internal notes. Take this information and put it into a contract.” That’s been serving us well-ish.
More recently, the writing got cheap and the proof got expensive, so it didn’t differentiate us. Anybody can put together a contract. This is where AI fails us and falls down. It says, “I’m no different than anyone else. You’re not telling me what I need to know.” I was like, “Okay, we can do better.”
As I mentioned, we had AI writing our scopes. What we’ve been noticing over the past few weeks is that the quality has been degrading really quickly. John, in one of the last scopes that you put together, I saw the amount of heavy editing that you had to do just to make it make sense. That, to me, was a huge signal to say this isn’t working. AI is not doing what it’s supposed to do, because this is supposed to be a repeatable process, and it’s not anymore. That is a huge red flag in terms of whether AI should be doing it. If it’s not doing what it’s supposed to do, something’s broken.
It was creating things like work streams and waves, which meant nothing to the client, nor did they mean anything to us because they’re not measurable units of time. The tactics and deliverables never pointed at each other. In human-created scopes of work, the tactics and deliverables are always a one-to-one match. This tactic has this deliverable, and so on and so forth. There were no quantities. The timelines were made up. “Ongoing support” meant whatever the client decided it meant. We were doing ourselves a disservice by not putting those guardrails into what was coming out of AI.
Then there was a padded addendum. One of the things you probably noticed, John, is that all of a sudden with the scopes of work that we were creating, there’s this whole other addendum with all this other information in there that AI decided was important, but we never put into the template. The process wasn’t working, which is where we get into the 5P Framework.
It was never a writing problem. Running it through the 5P Framework by Trust Insights, what was broken was not the platform; it was the process. We didn’t give it enough contextual information to say, “This is what you have to do.”
The 5Ps being Purpose, People, Process, Platform, Performance: what the heck are you doing? Who’s it for? What are you trying to do with the scope of work? Ideally, win. You want to win the business, but that’s not enough because that’s not enough context for AI to say, “Okay, now I know how to write a really good scope of work.”
Who’s the audience? We think we know who the buyer is through authority, but that may not be who’s making the ask. We need to figure that out, and they’re not necessarily going to tell us, especially if it’s an RFP. Then you have your process, platform, and performance: did we win the business? That’s a really good way to measure success.
Here are the six things we changed. This is where we get into how deep research changes the sales process.
First, I changed it to ask me first: Is it an RFP or is it a scope? Is it a new client or is it a renewal? Do we have research, or do I go get it?
As a quick plug, we have a new Deep Research course that I’m using the plugin for to modify this skill. You can get that at TrustInsights.ai/deepresearchcourse. It’s up for sale now.
The reason we want to do deep research before we start writing these contracts is that we make assumptions that AI knows more than it does. It doesn’t. I’m actually working on a workshop right now, trying to figure out who I want to put as stand-ins for “this is what you think AI knows” and “this is what it actually knows.”
I don’t want to give away any spoilers in case anybody’s attending, but I’ll show it to you afterwards, John. I think it’s pretty funny. Basically, you think it’s this really smart person, but actually it’s a hot mess of a person because AI is not doing the thinking for you. You have to provide that context.
Especially when you’re doing RFPs, there’s so much you don’t know. This is a great opportunity for deep research. Even if it’s not an RFP and it’s a scope of work where someone says, “Hey, can you just throw me a pitch?”, you want to know things like: What is the annual revenue of this company if it’s publicly traded? Where are they investing their money? These are all things that deep research is really good for.
The first step that we changed is: Ask me first what kind of thing this is—RFP, scope, and so on.
Then you do the deep research. Go read about them: budget cycles, approval limits, who already does the work inside their building, whether they have worked with other vendors before, and what publicly available information exists that will help give you that competitive advantage when you put together a scope of work. It is such a tricky market right now, especially for consultancies, to demonstrate why you should hire them.
Then, list every ask. Each one traces to a person who said it with a yes-or-no test. If you’re on a call with a prospect and you have that transcript, it is a huge missed opportunity if you are not repeating their words back to them in the language that they’re using, because it indicates you’re really listening. If they say, “I want 100 green painted turtles,” and you give them a proposal for 52 pink painted turtles, you missed the mark. It’s so common, especially when we let AI just run with it and do it—like, “Here’s the transcript, go ahead and do it”—and we don’t check it. You have to make sure that it’s exactly what they’re asking for.
With an RFP, that is so hard to know specifically, but there are going to be other indicators. In our own proprietary skills, I have some of those buying signals. That’s going to look different for every company. What your buying signals look like should go into your skills and your deep research.
Then you have to connect it. If you’re saying, “I’m going to do this,” there should be some sort of measurable deliverable that says, “And this is what you get.”
Then stop and ask. One of the challenges internally is not having enough information or context. Sometimes someone on our team who’s maybe not on this live stream today has additional context because they’ve had conversations and maybe forgot to let the rest of us know that those conversations happened. This is a great opportunity to stop, pause, and say, “Hey, other person on our team who’s not on the live stream today, did you already have a conversation with this person? I feel like maybe you did and forgot to tell us about it. Or did you have a conversation where you discussed budget? Maybe we should be aware of that information.”
Something that feels like it should be straightforward, but when you’re moving so quickly, you’re a small team, and you are juggling a bunch of things, this is the stuff that’s hard to remember to do. Now it’s built into the skill.
The last piece is to check the math. It’s not that the skill is doing the math for you; it’s just doing a comparison against discrepancies. Did we say the fees are this, but then the total is something different? It’s just checking things.
John, I flagged something like that for you the other day. When I was doing the test, I said, “Hey, there’s a $6,000 difference here,” and you were like, “Oh no, this explains it.”
That’s what we changed. But the big piece that we changed was the deep research: the “go read about them.” In doing that stuff, it’s delivering up really interesting nuggets of information that we may not have known.
For example, a public agency publishes its purchasing policy. Unless we did the deep research, we weren’t going to know that. Their purchasing policy was that above $175,000, it goes to the board; below that, they need two separate quotes. We took that information and asked, “Do we want it to go to the board, or do we want them to be able to approve smaller projects?” That’s something we would not have known if we had not taken the time to do deep research.
Another example: a prospect asked for help in a category where their own team had shipped a product earlier that year. We were going in thinking this was net new, but in doing that little bit of deep research, we realized, “Oh no, they already have this.” The people evaluating us were the people running the thing they had already built. I know that sounds a little convoluted, but I can’t give away proprietary information or anything under NDA.
In running another test, what we found was that the same company that had already built the thing they were hiring us to help them with had just posted a job for this exact work with their salary band included. That gives us even more information about what they’re willing to spend to get this work done.
Again, these are all really straightforward things that, given more time, resources, and bandwidth, we could find out. But we’re a small company. A lot of people are under pressure to keep delivering, get as many proposals out the door, fill the pipeline, and churn them out. Relying on deep research is going to be your friend in these situations.
John, I want to pause there and get your thoughts on this new version of the process.
John Wall – 23:42
All of this stuff is market stuff that we’ve actually seen and played around with. I want to give some credit towards this, especially for RFPs: this has completely changed the game.
Just the fact that we can run all the stuff that we have through the model and have it come back with a first cut of our scope that’s formatted out and has tagged all of the items in the RFP is huge. That’s the worst part. In fact, until earlier this year, we would flat out refuse all RFPs because it was just a waste of time and a ton of work, and we knew that the hit rate of wins wasn’t there.
Now that we can go through and see 50 or 70 items in the RFP, the model can put together the paperwork and make sure that all of those have been touched. It really changes the game as far as what you can do and what you can answer. Even if most of the stuff is wrong, you have a document in our format, correctly laid out, and you’re just making edits. You’re not having to build it from scratch, which is huge.
It doesn’t eliminate all the work. You still have to go back, match everything up, and give it the sanity check: “Did the thing go off on some kind of crazy tangent and promise something that’s impossible or mess up the rate?” Having RFPs and scopes that are correct, on the mark, and good business is one whole thing.
This is the interesting stuff—the poker stuff we were talking about where you can find things out in the whole sales environment. The budget thing is huge: knowing that procurement is going to be implementing a whole bunch of stuff over $175,000. Spoiler: never go over $175,000. You need to have something that’s $120,000 or $150,000 so you don’t trip those alarms that get more people involved.
The twist one is interesting, too, because there were a couple of learnings from that. We found that the company is already providing products that are similar to the work that we do. It’s not a direct competitor; in fact, it’s adjacent. Again, we can’t go deep into it, but one thing I’ve considered is that this could easily be an acquisition or a consulting thing. The work that we’re doing meshes with the existing product they have, so there’s a different kind of opportunity there.
Knowing that as we go into negotiations and try to figure out if what we have fills their need changes the whole dynamic of certain conversations, and it anchors stuff at different points for us. It’s not just, “How much money can we make in six months?” It’s, “Well, if this works out, there could be a two- or three-year opportunity that makes the six-month thing look like chump change.”
Katie Robbert – 26:29
It saves us from wasting our time because, with the twist, we would have proposed one thing and they would have said, “Nope, we’re good,” because it would have been the wrong thing. But we didn’t know that.
This is the thing that I personally dislike about sales: everybody’s kind of being cagey. Nobody wants to give everything away until you get the deal closed. The person who’s asking for the proposal is thinking, “I’m going to tell you just enough and see what you do with it.” The person who’s writing the proposal is thinking, “I don’t want to give everything away because what if I undervalue it?” To me, it’s so frustrating. Could we just have an honest conversation about what’s happening? But at least in my experience, John, that’s not how sales works.
John Wall – 27:21
No, absolutely. That’s the biggest problem with all of this. I call it black swan hunting. Every deal that you go after is a black swan—you have no idea.
Even with the most simple thing, the biggest thing that derails so many deals is: Does the person you’re talking to just want the basic facts of what they get, or do they want you to explain where all of those things come from? Are they the kind of person who wants to dig in, fully understand what they’re buying, and be educated, or are they the kind of person who says, “No, I need to sign this within a week so I can get this off my table and never look at it again”? Those are two entirely different pitches.
We didn’t want to go too far into the cutting-edge future stuff that we haven’t tested and are still playing with, but there is something there. You can make offers to people, frame them different ways, and do some measurement as far as what kind of decision-maker or person this is, just based on a free white paper or a webinar that we offered. Once you get to the point of a scope, that could give you indicators as to how you should handle that kind of stuff and what kind of person this is. Unfortunately, you start getting into weird surveillance stuff with that, which causes problems.
There is definitely something to be said for the fact that all of this boils down to: Can they trust you, and do you trust them? The only way to do that is to work in increments of promise and deliver. It’s on both sides, too. A lot of sales tactics you’ll see do stuff like this: for the first meeting, they’ll say, “Hey, can you send me something about your product?” or “Can you tell me the two other people who are in on this meeting?” You’re constantly promising and delivering, and that allows you to ladder up to make bigger asks and deliver more product as stair steps to building trust.
Katie Robbert – 29:18
On that note, as you’re talking about building trust, I’ll pull up an example from the actual Deep Research course. Again, TrustInsights.ai/deepresearchcourse.
This is one of the real examples from the course that we just created. We’re using a previous RFP that Chris and I did way back in the day. Obviously, I’ve blurred out anything that could get us in trouble, but we were pitching T-Mobile.
When you’re pitching a company, we call it the dog and pony show because you’re putting your best people forward, trying to shock and dazzle them. At the time, if I recall correctly, the proposal itself—the response to the request for a proposal—was pretty good, but lightweight enough that the team brought in a live whiteboard sketch artist to do this while they were pitching T-Mobile in person. That was a bit of a tactic—a “look at this, I’m distracting you over here because we’re lightweight on the rest of it.” That’s not shade necessarily; it was just the reality of it, because the agency at the time was moving so quickly and we had to do a bunch of proposals to try to get money in the door. This was something different—a differentiator.
We took this example and said, “With today’s tools, let me close out of this thing. Now with today’s tools, let me do a little bit of deep research and see what that would look like today.”
Before we got into looking at resumes of whiteboard artists, we said—again, this is totally fictional at this point; we’re not pitching T-Mobile—through the Deep Research course, we show this example where we said, “If we did deep research first before we started putting together our proposal, what are we going to learn?”
All the steps are in the course, but what we came up with is a whole 17-page research report on T-Mobile US corporate challenges and the Trust Insights capability map, which changes the tone of the entire proposal.
What we have in here are the executive summary, the T-Mobile US corporate snapshot, business challenges, competitive challenges, staffing challenges, technology challenges, and then, for us, everything that Trust Insights does to solve those problems. That makes for a much stronger proposal versus something that is, “Here are all the logos we work with, here are the bios of our team, and here are your points.” That stuff is important, but it doesn’t dig deep enough to really differentiate you.
If you can use deep research to say, “This is what we do, these are the problems you have, and this is how we solve them,” that’s what closes business.
I’ll go through this at a high level. T-Mobile has many AI assets, but limited evidence of one enterprise operating system. The Trust Insights angle: AI strategy and consulting, plus fractional leadership. The employee adoption need is visible and immediate. The Trust Insights angle: AI enablement package, our TRIPS framework. These are things that are table stakes for us that we do every day, but until we break it down into those pieces and say, “Here’s your problem, here’s our solution,” T-Mobile doesn’t care. They’re not going to go through our website and try to find the solutions to their problems. They’re telling us, “You tell us the solutions to the problems.”
This research report really goes through in great detail everything that they’re struggling with right now, but also everything they’re doing well, who their leadership is as of September 2026, and their strategic narrative. This is all publicly available information.
That’s one of the things I want to note: we didn’t go digging behind password-protected gates, and we didn’t hack into anything. This is all publicly available information that, prior to using a deep research prompt, someone would have had to take the time to go find, make sense of, and match up against what Trust Insights does. This is the part that we feel more confident automating using those deep research prompting guardrails to say, “This is what we’re actually looking at.”
They have their business challenges: providing organic growth and customer economics through a changing KPI regime. That sounds like a big problem with immediate urgency. Then we get into what Trust Insights can do.
What that looks like in practice is that now we can take that deep research and their RFP with “here’s what your response proposal has to include,” and we can start to build out something that’s more meaningful, more useful, and more targeted. Again, these are all samples in the Deep Research course, which you can get at TrustInsights.ai/deepresearchcourse.
You’ll start to see, based on what they pulled out: Why this conversation? Why Trust Insights? What we’ll cover. We put together an agenda of the highlights: T-Mobile is already telling the market it needs this; the evidence in their own words.
As we talked about earlier, if you’re not using the client’s own words, you’re missing an opportunity. Even for private companies, if they have any sort of blog posts, social media presence, or content, you can mirror what they’re saying and reflect back to them what they’re highlighting as their own problems. This goes through the research, and based on what they said they wanted, here’s how we can help—more customized to them and personalized based on what we do.
I’ve said a lot. John, what are your thoughts on this approach?
John Wall – 35:29
This is an example of exactly what we’re talking about. You had a bunch of people sitting around getting paid in the conference room to talk to the person drawing this custom infographic for you, whereas now you just throw that paperwork in and you get a 13-page document done. You’re three weeks ahead of the process, and you’re doing it with one person instead of a dozen packed around the conference table. This is right on the mark.
There are additional levels of: “Okay, you’ve done a great job of getting the RFP to match. Now where do you go from there? What are the next meetings like, and what else do you do?” This is cutting thousands of dollars out of the process and getting you weeks ahead.
Especially when you see plenty of RFPs that say, “We need our answer in two weeks or a month,” you are under the gun. With this, you can get rid of everybody doing the war room Sunday night trying to get the thing done for Monday morning. You can now have your first cut within a day, and you’re moving a lot faster.
Katie Robbert – 36:36
That’s the important point here: it’s not replacing the human judgment of whether we should or should not pitch this. It’s making putting the response proposal together more efficient.
You and I have both worked on responses to proposals, and it’s very time-consuming. Everybody’s given their list of things they have to go off and do and come back with. Now these deep research prompts—or whatever you’re using—can do all of that for you, and you can focus on: Is this what we want to propose? Does this make sense? Does it align with what our business needs are? Does it align with what their business needs are? Who do we have to have in the room? Who’s the best on our team? What are we pricing this at?
AI is going to give you a lot of those suggestions, but you have to be the one to make the decision. You, the human, need to be the one to take all of that information and say what makes the most sense.
Deep research prior to sales is going to get you a lot farther. Can you use deep research to understand what your competitive advantage is over other firms that are likely to also be responding to these RFPs? Yes, absolutely.
You may not know who else is responding, but given what is being asked, you can come up with a list or stand-ins of what another firm might likely be, and really dig deep into your critical thinking: What would help us stand out? What do we do that nobody else does? Or what do we do that’s too similar to everybody else so we leave that part out or say why you choose us over the other guy?
John Wall – 38:27
Now that you’ve got the thing built, you can just say, “Have X vendor answer this RFP as well,” and have two or three of them do it. You can also look at these three or four RFPs and ask, “Who do we target as far as these other competitors, and where are they weak?”
That way, at some point later on, you’re going to be able to talk intelligently about the competitors and what they can or can’t do, or what to stay away from because you know that they’ve got you beat in a certain segment. That kind of insight is all trust-building, because at some point you’re going to have a chance to demonstrate that knowledge. It’s going to make it look like you know more about the industry and what’s going on than everybody else in the process.
Katie Robbert – 39:07
As of today, I’m still working on our revised skills for contracts and RFPs, but they’re pretty close. I’m doing a lot of testing to make sure that it’s giving us the kind of outcomes we want, so that before I hand it off to the rest of the team, it’s very clear what it does and what it doesn’t do. That is basic software development best practices: if you’re building skills, make sure you’re testing them, especially when you’re talking about skills that are helping you win business. You want to make sure that you’re thoroughly testing to make sure it’s not just making things up.
John Wall – 39:49
One last thing to check is: don’t be afraid to run your prospect through and ask how they adopt new technology. Try to get a feel for when they’ve adopted past technologies—how long did they have them, how long did it take them to get into online advertising, or how long did it take them to get into social networks? Any kind of questions like that give you a feel for the big problem that you are trying to short-circuit: Are they going to come back and say they’re doing nothing? We see this all the time. People are like, “Hey, we’re afraid to adopt this new technology; we’re going to bail.”
Trying to get some indication as to how tech-averse and risk-averse they are can help you, save you a lot of time, and especially tell you if you’re wasting your time with certain prospects.
Katie Robbert – 40:39
A few episodes ago on the live stream—which you can get past episodes of at TrustInsights.ai/youtube—Chris showed how to use AI to understand the tech stack. If someone has a website, their tech stack is basically public knowledge. If everything’s connected through their website, you can use deep research to figure out, “This is my prospect. They want us to evaluate things, or we want to know how much software they have. How heavy is it?” Include that in your deep research to say, “Help me understand what’s in the tech stack.”
There are publicly available tools like BuiltWith or others that can tell you what’s going on based on their website. Let the deep research do that for you and bring back all the tools that can publicly be seen in the tech stack. You can go, “I’ve never heard of six of 10 of those, so maybe we’re not the right fit,” or, “We’re going to do our own deep research on all of those tools and get up to speed really fast.” Use deep research to build your competitive advantage and stay one step ahead of your competitors.
John Wall – 41:50
Knowing their tool stack can open up all kinds of doorways for you as far as understanding what to talk about and what to stay away from.
Katie Robbert – 41:58
I’m not going to get into deep technical demos today; we really just wanted to talk about why deep research is beneficial to the sales process.
The other thing worth noting, because people often ask why they should start with deep research: large language models are essentially a Google search, but a lot of models in software providers haven’t been updated with their knowledge since January 2025. You’re running on older information. The more context you can give the model about what’s happening today, the better off you are.
When you’re talking about things like sales, you want to have the latest and greatest information so that you’re demonstrating to your prospect why they should become a customer—that they’re not just one of many, but that there’s a bit more personalization. That’s what companies are looking for: “I need you to focus on me and only me. I don’t care that you have 100 other clients right now; it’s about me.”
John Wall – 43:00
You actually hit the biggest realization that I’ve had over the past year: deep research has replaced search as a replacer. You’re just not getting the same quality of results from any regular Google query you’re going to do. It’s going to make connections and bring data into it. There’s going to be stuff that would show up on page 85 of the search results that you need to know, that needs to be at the top, and the system is smart enough to know that this would affect the sale so it needs to come to the top.
Katie Robbert – 43:30
Any final thoughts, John?
John Wall – 43:32
You can go through all this and just hope that one of the vendors doing the RFP isn’t the cousin of the guy who’s doing it, right? That’s the big backbreaker.
Katie Robbert – 43:40
That you have no control over.
John Wall – 43:43
Stay on the straight and narrow, don’t engage in any questionable graft, and the wins will eventually come your way.
Katie Robbert – 43:52
Check out our new course at TrustInsights.ai/deepresearchcourse. If you want to continue the conversation, you can join our free Slack community at TrustInsights.ai/analyticsformarketers.
We cover a lot of this on the podcast. Chris and I were talking about deep research earlier this week at TrustInsights.ai/tipodcast. Anytime you want to check out the live stream, past episodes, or this episode, go to TrustInsights.ai/youtube. I think that’s it.
John Wall – 44:22
That sounds good. Thanks for stringing this all together. We’ll have to report back after a couple more months after we’ve played with this stuff.
Katie Robbert – 44:28
Absolutely. Until next time.
John Wall – 44:33
Thanks for watching today. Be sure to subscribe to our show wherever you’re watching it. For more resources and to learn more, check out the Trust Insights podcast at TrustInsights.ai/tipodcast and our weekly email newsletter at TrustInsights.ai/newsletter.
Got questions about what you saw in today’s episode?
Katie Robbert – 44:52
Join our free Analytics for Marketers Slack group at TrustInsights.ai/analyticsformarketers.
John Wall – 44:58
See you next time.
|
Need help with your marketing AI and analytics? |
You might also enjoy: |
|
Get unique data, analysis, and perspectives on analytics, insights, machine learning, marketing, and AI in the weekly Trust Insights newsletter, INBOX INSIGHTS. Subscribe now for free; new issues every Wednesday! |
Want to learn more about data, analytics, and insights? Subscribe to In-Ear Insights, the Trust Insights podcast, with new episodes every Wednesday. |
Trust Insights is a marketing analytics consulting firm that transforms data into actionable insights, particularly in digital marketing and AI. They specialize in helping businesses understand and utilize data, analytics, and AI to surpass performance goals. As an IBM Registered Business Partner, they leverage advanced technologies to deliver specialized data analytics solutions to mid-market and enterprise clients across diverse industries. Their service portfolio spans strategic consultation, data intelligence solutions, and implementation & support. Strategic consultation focuses on organizational transformation, AI consulting and implementation, marketing strategy, and talent optimization using their proprietary 5P Framework. Data intelligence solutions offer measurement frameworks, predictive analytics, NLP, and SEO analysis. Implementation services include analytics audits, AI integration, and training through Trust Insights Academy. Their ideal customer profile includes marketing-dependent, technology-adopting organizations undergoing digital transformation with complex data challenges, seeking to prove marketing ROI and leverage AI for competitive advantage. Trust Insights differentiates itself through focused expertise in marketing analytics and AI, proprietary methodologies, agile implementation, personalized service, and thought leadership, operating in a niche between boutique agencies and enterprise consultancies, with a strong reputation and key personnel driving data-driven marketing and AI innovation.
Newsletter
Podcast
