The 2026 Proposal & Bid Software Report is live! ✹ get your free copy →

Log In or Sign Up

Why AI Won't Win Your Federal Proposals with Lisa Shea Mundt

Lisa walks through how to apply Aristotle's logos, ethos and pathos as a working audience analysis framework for proposals, capture and AI prompting. She explains why a proposal model tuned to one contracting officer is worthless the moment that person leaves. She makes the case for the kickoff meeting as the single most important development milestone and walks through how to set bid/no-bid gates that fit your specific organization rather than generic best practice. She closes with why Orals presentations are returning and how introverts in particular should approach them.

Transcript

Cold open

Christina Carter [00:08]: Today I am sitting down with the expert Lisa Shea Mundt, a proposal strategist, who is going to talk to us about what so many teams she speaks to still get wrong about federal proposals. Also why presentations now matter more than they ever have. If your team is trying to win federal work, this conversation is worth your time. I'm excited for you to listen to it. Let's get into it.


Welcome

Christina Carter [00:35]: Hey Lisa, it's so good for you to be on The Stargazy Brief. Thank you so much for joining.

Lisa Mundt: Thank you for having me.

Christina Carter: I'm really excited about this conversation because I love speaking with federal GovCon experts who know how to win a deal. They are so complicated. They're getting harder and harder to win. Any insight we can get is gold in my book. One of the things I really like about your background is you started in the rhetoric and the writing space before you moved into the strategy part. When you studied persuasion formally, what did that teach you that you wish most proposal and pursuit people really understood? Because most of us don't have that background.


Part 1: From rhetoric to federal proposals

A foundation in writing and rhetoric

Lisa Mundt [01:20]: I would be happy to talk about that. For a little bit of context, I got my undergraduate degree and my master's from James Madison University in writing, rhetoric, and technical communication. I also had a double major in undergrad.

I ended up taking the introduction to writing course at JMU and I said, I really do like this. They made it into a minor, so I added it as my minor, and then they made it into a major, and then the graduate program. I just kept going with it.

My dad was a big wig in federal contracting. I remember being 18, 19, and he said, no matter what, organizations are going to need somebody who can write, who can communicate, who can take high concept ideas and translate them into ways that are easily understood. That was the entry point, writing versus rhetoric. As I moved through the coursework I went, rhetoric is really fun. What an interesting concept.

People hear that word and they go, you're trying to change my mind, you're trying to convince me one way or another. And the answer is, yeah, kind of. The concepts of rhetoric came about from ancient Greece. What these philosophers, what the sophists were doing was saying, there's a way that when we speak, we are trying to accomplish something. We are trying to get somebody to change their mind, to believe in our way of thinking, to appeal to their emotions. We are trying to move somebody.

Aristotle's three appeals

Lisa Mundt: The foundation in rhetoric that I've carried forward into federal contracting is Aristotle's three heuristics of appeals. Logos, pathos, and ethos. Logos is the appeal to our logic, the facts, the real hard and fast evidence we have. Pathos is the appeal to our emotions. We're trying to move somebody. And we're also trying to appeal to their ethos, their ethics or their credibility. We're trying to say, I am a credible source, and here's how I am trying to communicate that to you.

Is there anything more apt in federal procurement than trying to appeal to those three things? The foundation is strong. When I found federal proposals, competitive writing, there are not a lot of competitive styles of writing out there.

Christina Carter [03:51]: Not after university.

Lisa Mundt: Maybe you're going for a grant or some sort of funding. But when it comes to competing for multi-millions or billions of dollars, this is a hyper-specific, competitive, rhetorically-laden industry. That I carried through with me. I feel like that helped differentiate me as a proposal manager and also as an Orals coach, which is one of my favorite parts of federal procurement to date. I can do the proposals, I can write the proposals, I can manage them, I can herd the cats, I can coordinate them. But when you get to that source selection methodology of an Orals presentation, that's where my interests are really piqued. To continue the thread between early university and today, I have in the last couple of years gone back to get my PhD in writing and rhetoric at George Mason University. The study continues.


Part 2: AI in federal procurement

Where rhetoric fits when AI is doing the writing

Christina Carter [04:53]: Where do you see rhetoric and being a really strong, persuasive writer fitting now with everyone using AI to write?

Lisa Mundt [05:03]: AI, as we know, is nothing new. I have been in federal procurement for the last 15 years, and for at least the last 10, people have been asking me about artificial intelligence as it pertains to federal procurement. I have been a legendary naysayer in leveraging AI. That's not to say I'm a Luddite. I use it. I use it to craft messaging when the genre calls for it.

In my PhD coursework last semester, one of the professors had us, as a weekly assignment, use AI tools to put our readings in conversation with one another, post the prompts that we used in research and critical analysis, and then the conversations that came out, and a reflection of what the outputs were. What I'll tell you is that AI, when it comes to academic research, is still real bad. It cannot cite, it cannot quote, it hallucinates constantly. It will say, here's this quote on page 47, and you'll go to page 47, the quote doesn't exist. Right off the bat, it's not something that can replace the human mind there.

But even so, to use it effectively still requires rhetoric in the development of prompts. You need to situate the AI as a particular audience, and that's why rhetoric is so important. It's knowing your audience, knowing their standpoint, knowing their frames of reference. If you're asking AI to do something, you have to do it from the perspective of whatever audience you are trying to persuade, because we are all persuaded by different things. In federal procurement and sales, we talk about capture being about knowing the needs, wants and biases of a customer. If somebody is cost constrained, you need to make sure you have that as part of their audience analysis. If somebody wants to do it better and faster, that needs to be incorporated. That's where rhetoric still plays a role. It's that human connectivity of what makes us tick and what makes us decide that AI doesn't have on its own.

The federal standardization problem

Lisa Mundt [07:30]: Until there is hard and fast standardization in the way that procurement is released by the United States federal government, until data is standardized, until the solicitation process is more standardized and streamlined and regimented, to respond using AI on the industry side is going to be a constantly evolving and losing battle.

The AI tool tuned to one contracting officer

Lisa Mundt: A story I like to tell. A couple of years ago, a gentleman slid into my LinkedIn DMs and said, I have a tool I want to show you. I took a call with him. He said, your job, proposal writer, is obsolete. You would not believe, Christina, the things that people say to me.

Christina Carter [08:21]: What a way to convince you of something.

Lisa Mundt: Bananas. He said, I have something that can replace you. I have built a model that can win proposals. I said, cool, is this for the entirety of the federal government, point and click? He said, no, Navy. I said, okay, so any solicitation that comes out of Navy you're going to be able to answer with the point of a button. He said, well, no, not Navy, like NAVSEA. I said, okay, so all of NAVSEA. He said, well, no, but it's been proven successful going to this one contracting officer. I went, okay, great, for that one dude until that one dude leaves. Why did we get all the way down for the AI to the myopic level of the individual? Because the individuals are making the decisions and putting out the solicitations. That's why this guy got all the way down to a single individual.

Christina Carter [09:34]: They change all the time. It doesn't matter who's there today.

Lisa Mundt: Even before the 2025 DOGE era, may it stay so far back in the past that we never see it in the rear-view mirror again, people leave. The names of the companies and employers will change. The individuals stay the same, but their roles and responsibilities vary depending on where they are. Going down to that level when it comes to AI, it can be helpful for a short time. It's not a long-term solution.


Part 3: How federal decisions actually get made

The shift from adjectival ratings to confidence ratings

Christina Carter [10:12]: That goes to my next question. I know so many people responding to these and trying to get on vehicles have a lot of misconceptions about how the federal buying decisions are actually made versus how they think they're made. You speak to a ton of people. What are the main misconceptions you keep seeing?

Lisa Mundt [10:34]: I sincerely believe, based on how I speak to a lot of people in industry, that they believe decisions are made based on sterile analysis of data. We have this conception of evaluation criteria. Adjectival ratings, outstanding, marginal, this, that and the other. That is how things were historically done. We've moved more and more towards confidence ratings, whether or not the agencies have high confidence or low confidence in your ability to perform.

That switch has been very much by design. It's to offer contracting officers more flexibility in their right to choose, to down select, and to ultimately award to an organization. A common misconception is that a lot of people think this is a game of numbers, a stats game. If I submit 99 proposals, I'm going to win 30 of them. That's why we have stuff like Pwin. Everybody is trying to take a subjective medium, the written word, and objectify it. I have a 30 percent chance of winning. I have a 50 percent chance. It might make you feel better at night, but nobody knows what it is. You can do everything correct and still lose.

Reputation, trust, and incumbent-itis

Lisa Mundt: That, I think, is the common misconception. What am I doing wrong, I'm not winning. There could be very basic reasons why you're not winning, or why you're being deemed non-compliant. Maybe you don't have the customer intimacy. Maybe people simply don't like you. That's going to impact your ability to do work. Reputation is a huge part of this. Trust is a huge part of this. That's why the confidence ratings are so flexible for the contracting officers. They need to trust that you're going to be able to perform.

A lot of people conflate federal sales with marketing. There's overlap. There's a reason marketing exists in an environment where sales is prevalent. Marketing builds and fosters that goodwill and reputation in industry. You're more likely to sell to somebody who knows you, who recognizes your name, who says, I've heard of your company and I think you do good work, based on these rhetorical appeals seen through marketing.

The most common misconception in industry is to think that it's a puzzle to figure out, that there is a right answer and a wrong answer.

Christina Carter [13:41]: I see a lot of people who are really obsessed with the compliance piece and trying to make it into a math problem. That's important, but if you don't have that confidence, if they don't like you, if they don't trust you, that doesn't matter.

Lisa Mundt: It's funny you talk about not liking you. There are so many incumbents that get that incumbent-itis and they go, the customer loves us. And I go, the customer might like you. The customer doesn't love you. Your mama loves you, probably. But the customer, that person over in the procurement shop, no, they probably don't love you the way that you think they do.

When does the pursuit decision actually get made

Christina Carter [14:14]: They're not going to put their job on the line for you. Where do you see that decision really being made?

Lisa Mundt: Legally, the decision has to be made at proposal, at bid. That's why so many people have the school of thought, if it's on SAM, it's already decided. Yes, there's capture and shaping. Not every opportunity gets shaped. You might think there's a likelihood it has, if it's a high dollar value, if it's a highly valuable contract, then yeah, somebody's probably stuck their little fingers in there, tried to maneuver it in a way that positions them best. But even a shaped opportunity is not an automatic win. It's about meeting the requirements. If the requirements are general enough where multiple vendors can compete, there's still a possibility of a win. It depends on how the contract is set up, on whether it's best value, on whether it's LPTA, on so many things.

Legally, decisions get made at proposal. There's always still a chance. You have to choose your risk tolerance and threshold because it's expensive to put together bids. It's expensive to write a proposal from a B&P perspective. It's expensive to outsource to consultants, to use AI. All of that becomes risk and reward.

Best practice is to influence during the capture phase, during the request for information, during the market research. I'm very much a proponent of responding to market research, getting in front of customers in advance to shape opportunity, writing white papers and saying, here are the solutions to problems you have. I'm also a proponent of, if there is an RFI or a sources sought notice, saying, here's an acquisition strategy we recommend based on how this can better position you, the agency, to offer a more competitive bid in industry. It can't be about, here's how you move it to me. It's, here's how we help you. We're going to act on behalf of the customer.

Lisa Mundt: We're all human beings, so somebody can make a decision in the back of their head well before proposal. But legally, the decision needs to be made at proposal.

The nuance of "if it's on SAM.gov it's already decided"

Christina Carter [17:32]: I'm not going to lie. I am that person who is like, if it's on SAM.gov, you shouldn't bid, but be wary. So I'm part of the problem.

Lisa Mundt: It makes sense though. It is a fine through-line to have for people. It can't be taken so literally by industry. What we're trying to show is that best practice is to do capture, best practice is to influence. However, there are tens of thousands of bids that come out. Some of them are going to be shaped. Some of them are, we got this nut and bolt, that one, no, if it's on SAM that's a need that they have, so you should shoot your shot. There are so many differences between products, between services, between different capability sets. The government will say, no, we're not making the decision in advance, if it's on SAM it means we're looking to compete it. But as industry, we know that somebody might have gotten in and there might be a little earworm, a little brain bug back there saying, this is really who I want to win.


Part 4: Pursuit decisions and process

Choose your own adventure: bid/no-bid gates

Christina Carter [18:43]: These are people making decisions. What decisions should teams be making during their pursuit process?

Lisa Mundt [18:50]: It means something different to every organization. There is so much rigor and process for process's sake in federal procurement. I have worked with too many bootstrapped small businesses to be like, there is a one-size-fits-all. There's not. I'm very much choose your own adventure. You need to let me know what your non-negotiables are.

Some people's gate reviews, some people's bid/no-bid. I had an organization that said, if it requires more than three key personnel, I'm out. Because we're such a small business, we don't keep people on the bench. That's all going to have to be staffed up. Some people have said, if it's more than 30 pages, I'm out, it's a no-bid for me. Everybody has their pain tolerance threshold. Some people say if it's less than two weeks turnaround, they're not going to respond, because that's too quick. Some people say if it's less than such-and-such million dollars, I'm not getting out of bed.

Christina Carter [20:06]: I want that issue.

Lisa Mundt: If it's less than 20 million, I don't even want it. There are some organizations and portfolios that have those dollar thresholds. So I don't have a hard and fast, these are the decisions you need to make.

The kickoff meeting is the most important milestone

Lisa Mundt: I will say, with the exception of, know who is going to be your point person, know who the stakeholders are going to be, and what their roles and responsibilities are. That's where I see teams fall apart. That's why the kickoff meeting is, I believe, the most important proposal development milestone there is. It's where you say, we've shredded the solicitation, here's the annotated writer's outline, here are all the writing assignments, here's the schedule, here's when these are due, here's your pens down. Here's how we're going to coordinate, here are the logistics, here are the iterations of reviews we're going to do, whether you follow the color team review model or not. Make sure everybody knows the direction they're going in advance, because you can lose time and ground fast if you don't. There's a reason it's called the proposal manager and the capture manager. They manage the team and the process.


Part 5: Data and capture intelligence

Data doesn't make decisions, people do

Christina Carter [21:27]: What are some other signals I could be looking at?

Lisa Mundt [21:31]: Data is just that. It's data. It's informational. It's meant to inform, but not to decide. Data doesn't make decisions. People make decisions. Any data you extract is all pontification, because we typically look at data as historic information to inform future progress. A history of dollars spent is not 100 percent indicative of dollars that will be spent in the future. All we are doing is saying, this is trend analysis, this is where the trends were. We might ascertain this is where the trend will go, but we're still guessing.

Should you buy market intelligence?

Lisa Mundt: Should you buy data? It depends on your organization. We've worked with a lot of people. There have been companies that said they wanted market intelligence, and then they get it and go, okay, now what? Because it's not intuitive how to change the data into action. Whether you want to buy data depends on your ability to enact it and to move forward. So yes for some organizations, no for others.


Part 6: What to automate, what to keep human

AI is good for admin tasks (but you still have to check it)

Christina Carter [22:45]: If I'm setting up my process, what parts should I be automating? What parts should only use human judgment? Is it always a mix?

Lisa Mundt [23:04]: I think it's, unfortunately, always a mix. People want to use AI for admin tasks. We've been doing this for as long as I've been in industry. Even before AI, there were Word macros, I'm showing my age here, that would identify acronyms to help you build an acronym list. Before that, it was some poor sap like myself scrolling through hundreds of pages, finding each individual acronym and putting it in a table, because these were printed files and there was no Control F. That's changed. I haven't seen an acronym list required in a long time.

Christina Carter [23:57]: Me neither. Thank goodness.

Lisa Mundt: I haven't seen the requirement, I don't want the requirement, please keep it as far away from me as possible. But that administrative task, yeah, you could use AI to do it. You still have to check it. A lot of people want to use it for shredding a solicitation to build an outline. What I've always said is, it might have taken me four to six hours to build these annotated outlines manually. Maybe the AI did it instantaneously, but I still spent two hours reading through it, looking at it, changing it, adjusting the formatting.

Writing is a knowledge-making activity

Lisa Mundt: And also knowing it. Because writing is a knowledge-making activity. When you are writing, you are learning. Knowledge is being codified in this crazy human brain that we have, and it's being hardwired there. So maybe I saved four hours by using AI, or maybe I know it a whole hell of a lot less. That's the toss-up behind using these tools.

There are always ways to say, make this sentence sound better, here you go, that sounds more like English than what I said. But it still takes so much human interaction. This idea of point and click. There are going to be people who say, I have Claude and I've figured out X, Y, and Z. And yeah, you probably have, for your specific company, customer base and capabilities. Everybody can find the flow that works best for them. But as a general rule on the whole, I think it is always going to be a marriage of people and technology.

Christina Carter [25:43]: I have built Claude skills to automate a lot of the admin, but it's never fully handed off. It's not, okay, I don't have to think about it anymore. It's, you put it in a table for me, but now I need to go read that table in depth.

To me this is so obvious, but the fact that I am seeing so many proposal teams never get hired or automated out for AI, I see legal issues down the road. Revenue loss. Embarrassment. It's going to help you, but it's definitely not a replacement, in my opinion.

Lisa Mundt [26:26]: Even for the admin, the way that I've seen some of these AI inbox management systems hallucinate, collapse threads on top of each other, attribute emails to somebody else's account where they never sent it, it is just not. You still need critical reading ability. You still need to understand how comms works. The whole, I'm just going to outsource this, when people say that, I sincerely believe they simply don't understand how federal procurement works and operates.

Tools are only as good as the team that adopts them

Christina Carter [27:03]: I think it's the same for any team. AI is useful, any tool is useful, but it doesn't know necessarily what good looks like. The only people who know what good looks like are professionals who have been doing it for a long time. If you have somebody reviewing it who doesn't know what good looks like, then that's still not good enough.

Lisa Mundt [27:27]: Absolutely. Something I say to all my small business teams now: a tool is only as good as the team that is going to adopt its usage. If you don't adopt it, if you don't have KPIs attached to the adoption of the tool, people aren't going to use it.


Part 7: Differentiation for early-stage GovCon firms

Most companies don't have as many differentiators as they think

Christina Carter [27:43]: I have an accounting software. I am not an accountant. My accountant looks at my accounting software. That's how it is. You advise so many companies. Let's say you're advising a really early growth stage GovCon firm entering a crowded category. Where would you say they should focus their differentiation? Narrative, pricing, relationships, data, something else?

Lisa Mundt [28:32]: Differentiation is one of those buzzwords in federal procurement. A lot of people believe they have a lot of it, and then come to find that maybe you're not as special as you thought. A lot of people will say, we have 15 differentiators. You might have one or two. From there I would dive into that differentiation. When I say, what makes you different, and you say, our people, and I say, what about them? Well, our people are the best. The best how? They're the best at this work. According to who? They get down to the level of granularity they think is sufficient, and then I say, okay, but then why does that matter? Not in general, but to this specific client. That's where customer intimacy comes into play. That's where understanding your value proposition comes into play.

The government is never the answer

Lisa Mundt: You have to know who you are and what you're selling, and to whom. When you say early-stage market entrant, the first thing I say is, who are you selling to? They'll say, the government. And I go, no, try again. You're absolutely not selling to the capital-G Government. You're not even selling to specific agencies in most cases. You're going down to the sub-agency level. You might be going down to the level of the individual. That's who you're selling to. The government is never the answer. So many early entrants say, I'm going to do federal work. And I go, great, help me narrow the focus here. Then they'll say, maybe I need to figure out who is buying what I'm selling. Perfect, we've gotten to step two. And then what do you do to find that out? You look at data, at trends, at what's happened in the past, to try to tell the future. Once you have that, then it's about getting in front of the customer. Then it's also about having the engine to submit bids, having the people, having the back office that can handle even winning a contract.

There is no real start except for, know who you are and what you're looking to accomplish, know what you're selling and why they should pick you. That's the step one, because then all of the other steps shoot right off of it.


Part 8: What teams should focus on in 2026

Orals presentations are the answer to AI

Christina Carter [30:45]: Let's say I run a proposal team, a pursuit or capture team. What should I really be focusing on right now? The way things are being purchased by the federal government, on an agency-by-agency basis, is different and more difficult. If I'm trying to train my team, what should they be focusing on so they can keep winning instead of falling behind?

Lisa Mundt [31:16]: I'm going to bring it right back to the beginning and say Orals presentations and preparation for that. The answer to AI is Orals presentations as a source selection methodology. Orals have been in the FAR for the last 30 or so years. They are nothing new. They had a pause in 2020 when all of them were canceled because they were in person and we hadn't moved to a virtual environment. Within the year, we started seeing more and more Orals because of the virtual environment. We said, now we don't have the administrative burden of having to book a room, it's just on Zoom, this is super easy, people can do it from their homes. People were more prepared for them too because they were virtual. There was less pressure of being right in front of somebody. Everybody got used to this virtual meeting environment.

Why Orals are returning in-person

Lisa Mundt: Now we're going back to in-person, because people are always going to game the system. The reason that Orals presentations were really spearheaded in the 90s is because somebody on the government side said, industry is getting too good at writing proposals, and now we need to see them do Orals. Somebody on this committee, and there's a video of it from 1994, said, well, with all due respect, won't some people be better at oral presentations? And the answer was, yes, but those are the people we want to work with. That was the moral of the story. When you talk about confidence ratings, if you can get in front of somebody and clearly articulate that you know the subject matter, confidence goes through the roof.

Upskilling SMEs to articulate what they do

Lisa Mundt: Training and upskilling your SMEs on how to talk about what they do. We know that's a hardship in the proposal process. People always say, how do I get my SMEs to write? And I go, you kind of don't. You have to write and you have to extract it from them. So many times I've said, you say you do this, what's the process? And they say, well, there isn't a process. I said, yes, there's a process for everything. They say, no, there's not, I just do it.

Really, when you put your pants on in the morning, do you do it one leg at a time or both legs at a time? They say, one leg at a time. I say, that's a process. You have a process by which you accomplish your tasks. Having them skilled in being able to say exactly what they do, why they're good at it, and what the benefit is, that's where people should be focusing right now in 2026.


Part 9: How introverts get better at Orals

Meeting people where they are

Christina Carter [33:47]: I know so many people in proposals where presentations are anxiety-inducing. I don't know if it's because a lot of us are introverts or just not a lot of practice. What would you suggest to people who are like, I know I have to get better at this. What should I focus on? What should I do?

Lisa Mundt [34:11]: I try to meet people where they are. Not everybody is going to be loquacious, coming out saying, I'm going to charm you in a presentation. That's why I brought up logos, ethos, and pathos. It's not always about that. If it's just appealing to facts, and if you're an introvert, then you stick with the facts.

You have to really understand your communication style in a business perspective. There's the DISC assessment. Have you heard of that before? I'm strong D, dominant. I'm bottom line up front in the way I communicate. But not everybody is. There's the interpersonal communication, the compassionate communication, and others. It's understanding what your strengths are and then using them. It doesn't mean turning yourself into an extrovert. It doesn't mean turning yourself into an actor. But from a process and framework perspective, in a presentation, I need to account for their needs, wants and biases. I need to tell them what I'm going to tell them, tell them, and then tell them what I told them, because we're talking about retention metrics now too. It's finding somebody who can help meet you where you are to get better, and it's not going to be the same for everybody.

Why a little nervousness is healthy

Christina Carter [35:32]: It's never as bad as you think it's going to be. You go and you practice, you get better at it, like anything.

Lisa Mundt [35:41]: I present constantly and I still borderline black out every time I'm on a stage. It doesn't mean I'm not doing a good job. I believe if you're not a little nervous, then you're cocky. And if you're cocky, then you're susceptible to failure. If you're intentional, if you understand what you're trying to accomplish, and you're genuine and vulnerable, those are the things. We're all just trying to connect with other humans, and that's what a presentation is. That's what procurement is in a lot of cases, whether you want to believe it or admit it or not.


Where to find Lisa

Christina Carter [36:19]: I completely agree with you. Where can we find you?

Lisa Mundt [36:22]: I am very easily accessible on LinkedIn. Some might say overexposed. Please look me up, Lisa Shea Mundt on LinkedIn, send me a direct message, and I'd be happy to get back to you.

Christina Carter [36:34]: She gives the best advice over LinkedIn. Go and follow her. Even if you're not in federal procurement, go follow her. You're just going to learn wonderful things. Thank you so much for being on The Stargazy Brief. This has been wonderful.

Lisa Mundt: Thank you for having me. Until next time.