Name: From Adoption to ROI: What the Data Says About AI in Capture and Proposals
Uploaded: 2026-10-15T19:30:45.449Z
Duration: 1 h 52 s
Description: From Adoption to ROI: What the Data Says About AI in Capture and Proposals

Transcript for "From Adoption to ROI: What the Data Says About AI in Capture and Proposals": Hello, everyone, and thank you for joining us for today's webinar, From Adoption to ROI, What the Data Says About AI in Capture and Proposals. Before we get underway, a few administrative items I'd like to note. For the best experience today, you wanna go ahead and use Google Chrome. The audio is being streamed through your computer. There is no dial in, so just make sure you go ahead and ensure that your volume is turned up. If you do have any questions today, we would be more than happy, to have you submit them to our speakers. We will be taking and addressing questions throughout today's presentation. It's gonna be a very interactive session here today. So at any time, you can go ahead and enter your questions into the q and a box. We're going to try and get you as many as possible. I have a large group today. So if we don't get to your question, we will do our best to try and follow-up with you offline using the contact information you provided at registration. For those of you that presubmitted questions, our speakers have been, provided those questions as well, and we will try and read them in throughout today's presentation, as well. You'll also, be, note that we do have the slides available for today's presentation. If you are interested in downloading them, you can go over to our docs section on the right hand side of your screen, and you will see that the slides are available for you to download. We also have some great resources for you available, for you today, for you to download. Our GovCon AI gap, adoption adoption about... Adoption and maturity, excuse me, which you can go ahead and download again from that document section. We also have our GovCon clarity study, which is our, annual study where we, speak to close to a thousand government contractors and provide some great industry benchmarks and insights. And then lastly, we have the, link to the book that we promised every attendee that attended today's session. You can go ahead and download that. We'll also be sending that out post webinar. And then just keep in mind that that GovCon AI gap guide is going to be, very in tune with some of the things that we're talking about today, about AI adoption and AI maturity and where AI is showing up throughout the project life cycle. So very, very related resource if you wanna go ahead and grab that while we are, going through the presentation today. Some great resources. And then, lastly, I just wanna mention that you will also be receiving a on demand recording of today's webinar, and that will be, sent to you via email within twenty four hours after the webinar ends today. So with that, I would like to go ahead and introduce our speakers, for today. We are thrilled to have with us today, Brenda Kress from Lowfeld Consulting, group vice president, APM and coauthor of From Prompts Proposals in AI Maturity Model, which is the book that is being made available to you today. So you wanna go ahead and grab Brenda's, book. She's a coauthor of that. Brenda has more than twenty five years of experience providing capture, proposal, and program management support for companies serving the federal market. And joining her today is Deltek's very own Nora Beshore, a product director of AI proposals. Welcome to our speakers. And, Nora, I'll turn it over to you. Great. Thank you, Leslie. Hey, Brenda. It's good to see you. Good to see you too. We are so excited about today's discussion because GovCon has been cultivating this data for the past three years on AI and GovCon growth. So the audience gets to hear both an expert in the field, and they'll also get insight into this unique dataset. And, Brenda, when I heard about the dataset, you know, my eyes popped open because I peppered you with questions. So I can't wait to dive into your presentation today. Just a little reminder, as Leslie mentioned, we will be taking questions during the presentation, so feel free to just populate that q and a. Brenda's presentation is segmented into four components. And during each segment, submit your questions. We'll, pick a couple, and then we'll also have time for q and a at the end. So are you ready to roll, Brenda? I am. If you could just pop up the presentation. Oh, there we go. Okay. Well, let's get... Go ahead, Leslie. Do you wanna have any other introductions or you wanna get started? Oh, sorry. Let's go ahead and get started. For the first section, and I can bounce back to that title slide, where the ROI really is, we were talking about how AI is saving time across growth processes. For contractors, what do you think the bigger question is in terms of, like, if you save time, what should it and can it be put toward, you know, stronger bids or better pursuit choices? Where are you seeing the clearest measurable return? And... First of all, it's not more bids. Not more bids. It is making better decisions about your bids. You know, making better bid, no bid decisions or improving the quality of your proposals, not more bids. That, I think, is incredibly insightful, and I'm gonna turn it over to you. So let me know when you want me to advance the slide. Yeah. Could you advance the slide of the first slide, please? Yes. First of all, before we start off, I wanna just give you all a disclaimer. Will Pill Consulting is a thin and proposal services company. We are not an AI tool company. We we are totally tool agnostic, and most of the data we have here were taken from independent polls that we sponsored on LinkedIn, or they are from, meaningful studies like the clarity study that you can download. We also have data from Census Bureau and some other industry recognized experts. So it's more of industry agnostic data. So with that little disclaimer, let's get started. So right here on table one, we asked a question, are you maximizing the benefits of your AI implementation? And we started these surveys in 2024 when AI was still quite new to the capture and proposal industry. And here's what they said. We had four options here. Is AI exceeding your expectations, meeting most needs? Is there room for improvement, or is it not meeting your expectations? And we see between 2024, 2025, and 2026, the answers changed up a little bit. So in 2024, you know, we were saying for the most part, only, you know, 10% says it's exceeding and only 10% said it's meeting the needs. Whereas, you know, 42%, I love those healthy skeptics. Also, there was room for improvement and that's pretty consistent and that's where I would stand across across this spectrum of the past three years, and not meeting expectations. I think we had a kind of a higher tolerance against AI in 2024 because we weren't quite sure about it. But dramatic changes started happening in 2025 and 2026 when we all got much more familiar with AI. We saw that the exceeding expectations rate doubled as well as meeting most of our needs. That also doubled. And then we also said not meeting our expectations fell from 11, you know, from 38% to 11%, then it just crept up a little bit more. But I think this is a great, survey to showcase how our feelings about AI have changed in the past in the past three years. We're getting more comfortable with it. You know, we have a healthy skepticism still, but it is meeting more of our needs. So if you could go to the next survey question, that would be wonderful, Nora. There you go. Okay. And table two, we asked, what is your estimated cost savings achieved using AI for bids and proposals? And we had four categories here, either 20 to 25% savings, 26 to 50% savings, seven... 51 to 75% savings, and more than 75% savings. And these surveys were taken eighteen months apart. And what we found in 2024 is that, you know, people were saying, hey, you know, it saved me, you know, it paid off, 44% in 2024. And in, from '26 to 50% savings was 28%. And we saw that zero to 25%, savings increased dramatically in 2026 and the rest of the numbers decreased. And I think that's because people had, gained more information about AI, they're more familiar with it. You know, when 2024 we probably didn't have many measurements except for that gee whiz, oh, I can get something to write for me, you know, something fast or that was great. And I think by 2026, we have more measurements that we're measuring not just for timeliness, but for quality and risk. You know, verification habits have taught... Caught up with adoption. Maybe we were a little less skeptical... Had more skepticism in 2024. And, we have a lot of verification that we're doing in 2026. You know? And also to ease... Easy wins were banked in 2024. Now we're getting not only using it for writing our first drafts or using it for many, many other purposes. And so we're taking a better look at our metrics. And I think in 2026, we have brought a representation of the people who are using AI as, you know... I I would be curious if you could even put in the comments, you know, is there anyone not using AI who's listening to that? That would be super interesting, or even where you think AI has paid off for you. So that's interesting to hear. Yeah. Let's go on to the next slide. So where is AI paid off? And I think the... It's going to be, you know, the same for 2024 and 2026. Most people are still using it for drafting content. As we can see, the number went down slightly from twenty twenty six from 66 to 20... For... To 58%, but we also saw a rise in options two and four on this page. We see more proposal reviews, and scoring that increased from 15 to 19%. And likewise, compliance, increased from seven to 12% whereas, brainstorming and solutioning stayed a little bit constant. I think if we were to take this poll again in 2027, you'll even see a decrease in drafting content. And as AI slips in to more and more of our use cases and more and more of our processes, connected with capture and proposal development, you'll see that number increasing and just drafting decrease along that continuum. Okay. Let's go on to the next slide here. Okay. K. So we asked the... This is a very interesting question. We asked, why do review t... And it changed dramatically between 2024 and 2026. So we gave them four options. The the review teams falter because of value proposition, compliance, lack of specificity in content, and... Or lack of risk mitigation. And we were even thinking about, AI in 2024. But here we see, this big change in 2024. We saw not, you know, specifying your value proposition. Was it this is reason why red teams falter? And then look again on 2026. The biggest reason was lack of specificity in content. And why do you think that happened? It's probably because of what I call AI speak. Some people call it AI slop. You know, people were generating content as you can see by the last the last survey here using it to generate content. But that content might have sounded professional and vaguely respond to requirements, but yet it lacked the specificity to generate a strength or score for the proposal for the proposal evaluator. And I think that was a huge downside of AI. We kind of adopted it quickly, but maybe didn't think about all the consequences. And I think that that has... We'll see some differences in 2027 because I think most people are very aware of that problem and are trying to correct it. Okay. So let's go on to the next one, Nora. Okay. Okay. And, so I like to put a pin in our talk right now. We've we've talked about several surveys right now. And I see also right now competitive separation widening between the people who are using AI and those in a very disciplined manner and those who are just using it at the prompt level are not using it in a very disciplined way. You know, people who are using it in a disciplined manner, knowledge becomes a strategic asset that keeps improving with both AI performance and human decision making. In essence, it's kind of like a flywheel, if you will, that builds on top of each other. Better governance equals more content discipline, which produces better AI output, which frees up more capacity for things that really matter, like making, you know, better bid, no bid decisions. You know, focusing time on strategy or customer relationships. In fact, we have a quote in our book. AI advantage isn't gained by adopting technology faster than others. It's achieved by operating it more effectively, consistently, and with stronger discipline. And we're gonna talk a little bit more about the book, in a few slides down from here, but, I just wanted to mention that right now that I think that we'll start to see better results from people who are producing a mature implementation of AI. So let's go on to the next slide. I wanted to put a pin in right now and talk about small business. As we've seen in government today, we have fewer small businesses participating in GovCon. That number has gone down. And we have a few facts on this slide that we've gathered. The first one comes from the SBA. We have fewer eight a firm was participating. We have, you know, more realignment of it at at at eight a firms. We've also seen that the small business administration is increasing its size standards. It's in draft right now, but I'm sure it will come to fruition in 2027 to include more small business in these pools. We also see in this slide a growing use of AI by firm size, and this information was taken from the US Census Bureau, this second line here. 37% of firms with 250 plus employees use AI versus under 20% of firms with four or fewer staff members. And here is another fact coming from the clarity study, which you can download in, the box that Leslie, indicated here is that nearly 44% of government contractors use AI for proposal development. And I'm sure that number is only growing since the time that, the clarity study was written even this year. So that's why in the book, for each chapter, we set aside a portion of that to talk directly to small business. And I wanna say to small business, you have an advantage over large businesses. Well, large business have the resources to become an early adopter. Small business, you also have advantages right now that your small business size means there's less bureaucracy, less red tape. You're able to implement these processes. You're able to start them up faster, look at the results, and maybe react faster than large business can. So in the book, we give step by step processes and procedures implementing AI, training your staff at AI so you can use it as a force multiplier. Okay. So that's about it. I think we're ready to, roll on to part two, Nora, whenever you're ready. Okay. That sounds good. Can we take one question? And I'm not sure if it applies to this segment or a future segment, But Carlo is asking, I've seen AI tools broaden their offerings to include things such as opportunity identification. Where do you see the greatest opportunity for AI tools to add additional value to capture and proposal processes? That is a great question. And you know what? I would say it depends on your company and your use case. Look at your own company. Where are you spending the most time either collecting information, processing information, or disseminating it? Where could AI help you in this? So I think each company might be a little bit different. For my own company or just for me personally, here's what I love to use it for. And this is... By no means a commercial, but I do use GovCon just about daily. And I cool... And as a proposal services firm, we collect a lot of information about bids coming up, and I run my searches on a daily basis. And then I use AI to help me parse the information and, you know, come up with trends and things like that. So that's personal. It could be different for every person. That sounds great. We'll get to more questions in the future. For now, let's head into segment two about capturing value at scale. Brenda, what do companies need to put in place to make quality AI results consistent as use expands? You know, I suggest processes and procedures for using AI. And and with that, let me just take a departure and talk about this AI maturity model. So we're all on the same same page here. You know, we have been... We we collected this information, and developed this model after looking at the industry for the past three years. We taught a class in AI, how to prompt, things like that. We've talked to customers. We've talked to vendors. We've talked to the government even. And after all that cogitation, we came up with this, maturity model right here, and I'd like to explain it because it will influence some of the things that we talk about during the course of our presentation. So you can see here is level one is explore, and that is kind of ad hoc curiosity. You know, this tool is just proving that AI is useful and safe to try. You know, you might have downloaded chat GPT and you might have put it within your firewall and you're just kind of experimenting and that's great and you're starting out and that's level one. So level two is apply. You're turning experimentation into a repeatable process, and, you're building a short list of use cases that consistently help you do your work, more efficiently and of a higher quality. Now level three is standardized. This is where AI outputs become consistent across the team depending on who's using it. So no matter whether, you know, staff member a or staff member b uses it, you're using the same, prompt or process or agent, and you're getting similar results. Level four is scale. The goal shifts from speed to quality and ROI. So it makes your, your quality... Your proposals become higher quality, and they just don't come up faster. They come out better. Now level five is transform, and that's where AI orchestration happens. With human sign off at critical parts where a human must approve whatever AI has done, and humans steer the strategy. AI is just helping you along the process. So that is a short summary of our our AI model right here. And I also wanna, mention that, you know, it... To jump from one level to another level, you know, it does take process. It does take intention. And some of you might be at multiple levels at one time, and that's just fine. In fact, that's just that's just normal. And in appendix a of our book, we have a quiz you can take. It's a free quiz, and it just kinda tells you where you are and, you know, what you could do to improve your level and maturity. Okay. Ready to go on to the next slide, Leslie. K. Okay. So there, you know, it... Improving your maturity can sometimes come with challenges. You know? AI can't overcome some some problems that you might have. If you have low quality data, AI is only going to emphasize it. If you have fully tallied tagged or index data, AI is not going to retrieve the data that you need. If you have lack of human control over your, data, you're gonna have to live with the AI hallucinations. And likewise, if you have stale or outdated content or no designated owner, you know, garbage in, garbage out. So when you try to scale AI, when you scale, it's going to only emphasize maybe some of the defects that you have already. So what I would do before I, you know, told anyone to go in and start using AI, kind of take a look at your organization. How was the quality of your data? Because you're gonna get garbage in, garbage out, and AI is only going to magnify any issues that you're having right now. So gonna go on to the next slide, Nora. Okay. Okay. And sometimes AI implementation can be frustrating. You see this little frayed rope here, you know, straight from from the book we have here that, you know, AI can get frayed or issues can occur. You can see two writers coming up with totally different results. You know, you can see AI speak. You can see content drifts from approved messaging. And sometimes when you get started, you see these early ones, it's great. Yeah. Yeah. Yeah. Yeah. But then momentum starts to slow down because you don't have this governance in place. And AI stops being a productivity tool and starts being an operational liability. So... And that's precisely that gap between, you know, that that implementing, a a a tool based on, you know, high quality information, good governance, and those which you're gonna see, you know, as you scale up capacity, you know, you... You're gonna see the holes in it. And then what's gonna happen is your management's gonna come down on you. Get rid of that AI tool or, you know, or modify it or limit it. And that's hopefully what we want to avoid. AI is a wonderful tool. It's a wonderful force multiplier if it's used correctly and within the framework of your organization. Okay. Next. K. Oh, can we take a couple questions that we got in during that segment, Brenda? Yeah. Damien is asking, what advice do you have for engaging with people who are resistant to trying AI tools and their capture and proposal efforts? That is an excellent question, Damien. And so what I would say is I'd start with a small use case within your organization that can really benefit from AI. Identify your metrics, show them the, progress, show them the time saved or the quality improved in just a little kinda use case. Right? And then from there, their confidence can grow. And I think when they see the actual, you know, black and white numbers, those hard facts, they can't dispute them. That's a great fact.... The way you can get it to your your your senior management to embrace AI used correctly. That sounds great, Brenda. And then John is asking about Percentage of savings. He wants to know about what the savings exactly was. He's asking, was it in labor savings, reviewing bids from subcontractors? Do you have more information about that specifically? I think that was in the previous segment. Yeah. And so what I've done is I've just kind of showed the results in some of these. And for every table that I have cited, John, on these, I have the link to where you can find the actual data, all the details of the data, how many responses, everything like that. It's all on our website. And could you move back to that last slide there, Nora, for John?... Heard. This one? No. I think he meant the last numeric slide. Okay. Tell me where to... Stop. Well, no. Not last move for that. Okay. I'm gonna keep going. You tell me... One more one more. Okay. Yeah. The downside of the payoff here. And that was... So this is basically drafting... This is how much it improved, how much time you saved. So... Yeah. So like I say, John, we have links to every single fact we put in this presentation, and you can find them. You can see that link below. I'm I'm pointing to it. You probably can't see it, but maybe Nora can hit it with the cursor. But there's links to everything, much more detail and much more analysis too. I also wanna say that all of the links we have here are kind of the tip of the iceberg. We have so many more surveys on, on our website. They're free to everybody. And you can also see me on LinkedIn pulling people sometimes, doing other studies. So it's... We can go back to, part three by answering John's questions, hopefully. Oh, I think you did. And we're gonna start talking about governance. And so We love the speed of AI, but then governance issues. So I'm gonna turn it over you, Brenda, to dive into that. Thank you. K. So let's go ahead here to table six. Someone knew... Yeah. K. There we go. So this question asks, what's the greatest return on investment or ROI you've gained from using AI on bids and proposals? And so I think the greatest time savings was 65% of people said time on proposal writing. I think that's the normal answer as people have used it most prominently for proposal writing. Then I was very happy to see, that, the next group of people, 30% said in proposal quality. I think time saving is great. But ultimately, what we wanna do is we want to produce a proposal that scores highly, is most competitive, and wins the contract. So it's very hard to see this increase in proposal quality. However, reduce proposal cost and enable you to submit proposals that remains very low. You know, hopefully, most companies are using the free time to focus on win strategy, customer relationship building, improve proposal quality. You know, I also wanna say... Just put a pin in it and say, I think AI is is creating new jobs within the company. You know, I think we are spending more time on knowledge management so we can retrieve that... AI can retrieve the data. We're spending more time on verifying the AI hack is not hallucinating. Or we're spending more time on writing agents or prompts, standardized prompts or agents. And in the future, next year, I think we'll see more people working on agentic AI and writing agents for that and using AI to communicate between our systems. That's my prediction. But I I think it's really important to look at how you're using AI, what quality measures you're using to measure it so you can make sure that you're using that time wisely to produce winning proposals. I agree with you on that. Would you like the next slide? Yes, please. K. So we know... All know and probably agree that AI is a force multiplier. It is giving us time back. Although it has, you know, like I mentioned before, given us new tasks. But for me, it definitely has saved time. And hopefully, we are using that tie... Time to select bids that are most likely to win. You know, we're not producing more and more proposals. I know sometimes, gosh knows that our executives would like to bid more, bid more, but maybe that isn't the right answer. Maybe it is saving that time to bid more wisely. That we have more time to talk to our customer, find out about our competitors, talk to our vendors, stakeholders, sector so we can build a fantastic proposal. And also, maybe we use that time for more deliberate reinvestment in sharpening our win strategies, training our people. Because what we train them in in 2024 isn't what we're gonna be training them in in 2027. That that that dynamic couldn't change. So time savings alone is not going to win us our next proposal, but spending our time on evaluator focused tasks that will increase our score will help us win the proposal or, likewise, you know, coming up with a more competitive price. That makes a lot of sense, and I love hearing you talk about the strategy and how important that is. And so I just wanted to put this slide in as a clarifier here. You always have to have humans in the loop. It's so important. You know, when we look at the jobs on the left, 10 jobs where AI can increase productivity. I love AI. I use it daily. You know, I'll I'll use it to scan SAM.gov or like I mentioned before. I use it to, you know, look at... I run big of wind searches, and I use it to detail the information I want to analyze there. You know, I summarize agency priorities because I don't wanna read a 300 page document. I, you know, I I aggregate information about my competitors. You know, I could even draft strength statements. That's that's all good. That helps. Right? I use it as a solution buddy. When I'm stuck, you know, I'm sitting there typing and talking to it like it's my best friend. On the other hand, there are jobs that I will not involve AI in and that is, you know, our bid, no bid decisions. It's great for collecting data. We're gonna decide which jobs we're gonna bid on. You know, interpret nuances like what the customer is telling us. Machine can't interpret that. Validate competitive intelligence, or even make final calls on pricing. So as you build this, governance model, I think you have to determine what you're gonna do and what AI is gonna do. And certainly, we have suggestions about that in the book. Would you like the next slide? Yes, please. Okay. Alright. K. And using AI can be risky too. So we did this poll, this very fun poll, and it says, you know, what concerns you most about using AI for proposal writing? And by far and away, forty seven percent of the people indicated, nearly half, that it was, like, hallucination. And I agree that is a huge issue, especially if you have data in your company that is not very well vetted or could use... We all have data that could stand room for improvement. Right? Yeah. But often we... I have seen, you know, AI go out there and call data and cite the wrong program or hallucinate facts about that program, and we can't have that. We always have to have a human human in the loop there. And, likewise, as we saw in a previous survey, AI speak, AI slop, whatever you wanna call it, it also is still... I I love it for a first draft, and it can help me fill in the blanks, but it's not going to take the place of of me for writing, persuasive proposal content that's gonna score highly with an evaluator. I also wanna talk about AI security. Even though it did not score highly, only some... 17% of the people saw AI security as a risk. I think it's... It is a big risk. So in the book, we have a very detailed checklist where every proposal professional can sit down and talk with their chief security officer about security. And there's just a huge laundry list of questions right there because I know I'm a proposal professional, capturing proposal professional. I'm not a chief security officer. And I just wanna make sure that I understand it on my part that the security controls are in place. My data is not being shared or learned on by a large language model or some type of proprietary tool. So I just would say even though it didn't rank highly, I do see that as a problem. And the last one is bias. Now I do think that is a problem as our AI agents get smarter and smarter. I know I use one particular one all the time, and I was just asking it, you know, what are some of the best proposal consulting firms? What can I do better? You're assessing it to solution with me, and it came back, oh, Lillefeld Consulting is so great, blah blah blah blah. And I said, I've been talking to you for too long. Right? That is not true. So I thought it was not true. It's it's it's... We do have to watch out that it... That we're not drinking our own bathwater there. So those are four real risks, and we're gonna run this, survey again in December of this year and see how that's changed. But I I still do think that those all are are all very important. So I'm ready to go on to the next slide, Nora. Alrighty. Oh, can I interrupt before we get to this part? We've got from Carol Ann. This is a really thoughtful question. What do you perceive the long term impact will be on human performance and creativity if we promote the thought that the AI bot bot writes better content? That is an excellent question. You know, I think of bots, and they can have very good suggestions. But sometimes they do... They have one thing that will never be duplicated, and that is empathy and understanding that humans have about solutions. Many of the people that we work from GovCon are very mission oriented. And sometimes that requires an empathic response, not only, you know, a very creative one. And I think you have to temper that innovation with empathy and with human creativity. So I don't think... I think that they're great buddies and maybe if we team up, we can come together and create a good response. But, I just don't think that you can substitute, you know, humanity, for machines. Yeah. Definitely. Alright. Part four. We are seeing that the skills needed to use AI well are changing because these LOMs keep getting more and more sophisticated. So what do you think, Brenda, in terms of what growth teams should focus on? What do they need to get better at? And, when leaders measure ROI, should they look beyond the power saved? Yes. Yes. It is. Okay. So... Okay. Well, let's, I say yes to that question. Did you... Would you rather go on to this slide right here and talk about that? Yeah. It's up to you. Yeah. Let's talk let's talk about this and we can circle back. So we asked people what skills will matter most in in 2027. And so here are the answers. Directing AI output, 49, nearly half the folks. Storytelling and differentiation, 39%. In a much lower rates, we have knowledge management at 9% and far compliance and fluency at 3%. So I think what eight of... What executives want is AI outputs that are compelling, they're accurate, they're timely, and they just have high confidence in it and so can the customer. So I think that's a huge skill. And sometimes not only do we have to come up with the the response, it also has to be formatted correctly. So that's kind of a two for there. And storytelling and differentiation, I think that AI can help us with that, but that's more of a human in the loop type and maybe it can spur us on to to better solutions and and creativity. So I also wanna mention here that, you know, capture proposal professionals to increase their ability to, of course, direct meaningful out, outputs by, you know, implementing AI maturity assessments and government... Governance procedures. You know, make sure you have human in the loop controls, learn to build persistent AI project or workspaces that load in con... You know, consistent instructions. So I think that your ability to implement good governance, best practices will be important too aside from these technical skills. I also see that we have changed from 2024 to 2027, if you will. Because in 2024, we were just looking at, can you prompt? Can you type something on that white line? Right? And, you know, in, you know, in 2025, we were building agents. And and I think in 2025, 2026, we're talking about persistent workspaces. That means, you know, building projects with instructions and using that. And I think we're we're gonna move beyond that prompting kind of agent development stage into 2027, which is making all of our tools work together and produce outputs. And so I know that's why we have changed our class in 2027. We'll be coming out with a new class at Littlefeld Consulting that does precisely that because I think as AI evolves, we all have to evolve and and learn new skills. And we do have a list of skills in the book that people can kind of test ourselves against. That's incredible. And I bet people are gonna find that super useful. Are you ready for the next slide? I am. Thank you very much. So we wanna measure what matters. In table nine, what is the primary measure you use to track the effectiveness of using AI for capture and proposal work? And I think most people are view... Looking at time savings right now. So... But, however, I feel very encouraged that, you know, 26% are using it to look at their rates of compliance and rework. And, while also a healthy 16% are looking at win rate or proposal scores and quality and defect indicators. And that suggests to me that probably time savings is great. We all want... And maybe we can use that time more effectively. But, we'll see that growing during the next year as as people start to measure other type of quality indicators, compliance indicators, anything along their continuum of life cycle tasks. I also wanna point out on table 4.4 of the book on page 97, we have concrete suggestions for measurement. And I wanted to name some of those right now if anybody was was taking notes along the way. And in addition to time save, I have a couple of other suggestions which include proposal quality and readiness indicators Compliance and defect reduction, rework review and cycle time efficiency, knowledge reuse and content effectiveness, bias control, evaluator trust indicators, you know, ROI and business impact, and governance and continuous improvement. So you only have to look within your own organization, say what ROI are you trying to achieve and and build metrics around it. Those were just only my my suggestions. I think they sound like really great strategic suggestions. So I flip to the next slide if you're ready for it. Okay. This is this is our summary slide before the... We have some slides about resources for you. But AI's ROI is real, but it's uneven as we've just seen in some of these surveys here. You know, the exceeding expectations number has plateaued about 20% for the past two years, which tells you that, you know, that we're we're doing a great job here, but, you know, we can do dig in and maybe find more, you know, or greater ROI in the next coming year. And we also learned in... We will see a little magnifying scaling of, you know, we'll identify those hidden gaps. You know, if your house is not clean, it'll just magnify its uncleanliness. You know, if your data isn't good or if your processes aren't sound, you will find it as AI starts to retrieve your data and give you output. And next here, reinvest capacity. Just don't spend it. Yes. AI will save your time. There's absolutely no doubt. But I would just caution you. Don't automatically spend that doing more proposals. Spend that time on, tasks that can help you win better no go decisions, and similar similar decisions. And finally here, we have here that, Direct AI just don't use it. The same logic applies to your people. Please train your feed people on AI, help them with the governance, and just increase the effectiveness of your AI implementation. And so remember, direct it at your AI, and thank you for having me, Nora. I really appreciate it. And I think we have some more resources, on the on the subsequent slides and my contact inform... Information. If you have any questions at all, please don't hesitate to contact me, and I guess we're ready for some questions. We have a lot of questions, so let's see what we can get through. This was a tremendous presentation. Alright. What is the number one mistake that you must not make when using AI and proposal preparation? Security. No security leaks. K. Awesome. Work with your CSO. Follow the checklist in the book. And Yadira, I think this might get to your question too. How do you train people to review AI output critically rather than just rubber stamp it? Look for your sources. Require AI to identify any source for a factual point of information so you can go back and verify it. That's a great answer. Oh, here's another really good question, Brenda. How do you sustain adoption after the initial rollout when the novelty fades? Oh, that is... That's, like I said, great question here because you have to have a plan for adoption. You know, if you have a vision and you have a a destination of where you wanna go from, like, a to z, and I think it's one thing the book can help you with. And then you'll... You can kind of chew on each little milestone that you make and know, but, yeah, I'm farther to go and so you have... But you... You know, I think what's that quote with... Without vision, you know, everything fails, so have a vision. Alright. Darren asks, understanding that source selection teams are employing AI to support the evaluation process, are there any changes that we should implement given this reality? I'm I'm sorry I did not take care of that in the presentation. That is a great question because I know right now of at least two commercial tools that gov... Gov con is use... Or the government is using right now, as well as other large language models and things like that. And they are now using it from a to z, everything from market research to even helping them evaluate proposals. It is even looking at your proposal right now and detecting if you use AI. It's also using it as a protest mitigation factor. So they're kind of protest proofing their RFPs before they come out. There has been a huge increase on the government's part in the past few years in their use and maturity of their AI implementation. So just know that many people who are using it from soup to nuts have mature AI implementations and just be cognizant of that. And I have two or more blogs about it on Lofell Consulting website and it's also been published by Federal News Network. Awesome. I'm not sure we'll be able to answer this one, but Sonia is asking, does your book or the presentation up... Offer suggestions for using AI for assisting government agencies when they're preparing requests for proposals? I think my blogs have taken... Talked about that, about how they're using it, how they're using it for market research, how they're using it for preparing, proposals. But just know that they have the tools. They have the know how. You know, they've gone... Government, god bless them, has gone through a a kind of a, you know, rough time these past two years, especially acquisition, eight, departments of agencies. And I think they're coming back, and I think they're implementing these tools as a false false... A force multiplier for themselves. And so they're only going to increase and, they're using just that we're using them to do their work more efficiently. Here's a really interesting question about... Someone submitted, we are seeing a growing... And I don't know if you've seen this, Brenda. We are seeing a growing trend in RFPs and RFQs requiring contractors to certify that a was not used in developing the proposal. Given that we've invested in an AI tool for capture and proposal development, how do we handle that if the government is negating the use of AI? I think there's, yeah, I have I have seen that. I think, first of all, government is kind of blind the fact that everybody's using AI right now. And... But there's there's a difference between using AI to collect information and synthesize in a kind of stress test your your your thoughts and versus, you know, generating one using AI. And I think that's a real differentiator. And so I think maybe we need to our contracting officers about that. We're not using it. You know, we're we're just using it as a efficiency tool more than running. But I also see that government contractors, I feel sorry for them because they've had a lot of AI slop to read in the past two years. But no one has ever won a proposal writing AI slop. They win it because they're strategic and they have a high number of strength and they're competitive. That makes... I I agree with you there for sure. And then John is asking another question. I think this ties back to governments. With CMMC requirements level one, two, and three, with a firm that is an early adopter for DOD projects with high security requirements, how do you advise them that they should integrate AI with increasing safety concerns from the government? Yes. If you're gonna use an AI tool, there are many tools out there that are FedRAMP and CMMC compliant. So I would first start with those tools, and I would work with my chief security officer to verify that you were compliant with all of the normal security requirements for, like, for handling Cooley or something like that, that your that your government agency expects. Claudia has a good question here. In listening to Phil Christie, chief acquisition director VA stated, AI is driving back to LPTA due to similar technical proposals. What are your thoughts on that, Rhonda? See, that's where you need the human in the loop because that's where you need to... You know, you need to differentiate yourself. You And so let me start with the first one here. And, obviously, we have been talking about proposals. And so our first question is, would you like to be contacted by a Deltek sales representative to learn more about improving your proposal process? And so, obviously, here at Deltek, we've got a a lot of great products and solutions. And if you would be interested in learning more about how Deltek can help support your proposal process along with many other, ways we might be able to support your organization and business, go ahead and select yes, contact me, and we will be sure to have somebody reach out to you. Nora, you wanna take the next question while folks answer this, and then we can kinda jump into the next... Yeah. Sure. This is a really good one, Brenda. Alright. Since we know that the government has asked industry for solutions to support their desire to use AI tools to both create RFIs, is it possible to find out how the government's using AI to evaluate our RFI responses? Yeah. And to that end, I'm not gonna name any specific tools here in this conversation, but I would point people to our website and I've written about this. There are tools out there that provide, government agencies with the ability to go from market research all the way to proposal evaluation and take a look at those tools and and how government might be using it now. Government is not, you know, uniform. You know, some use commercial tools, other use large language models, others are building systems from from from scratch. But I think all of them are using them to build use cases to automate the acquisition process. And I would just talk, you know, playing with your contracting officer. You know, how are you using AI? Just doing formal conversation with them to learn, and that really might help you very specifically understand what you have to do on your part to make sure you're giving them the best, support that you can. That is a great answer, Brenda. And then, Leslie, do you want me to turn back over to you or go with another question? Okay. We're gonna jump into the next poll question here. Sounds good. Let me go ahead and open that up for everyone. And the next question is going to be about Lofa Consulting. Would you like to learn more about Lofa Consulting? Obviously, Brenda has done an incredible job today in helping us better understand the proposal, process, and AI, and how it's all woven together. And so if you really be interested in learning more, just go ahead and select yes. I'm interested in learning more, and we will be sure to pass this along to Brenda and Tim, and they will be reaching out. Can we take one more question, Leslie? Absolutely. We've got about, a couple more minutes, actually. So... Yeah. Absolutely. Alright. Brenda, let me know what you think about this question. The person is asking, asking, can we anticipate an increase in demand for all proposal presentations in the acquisition process due to AI assisted proposal development? Absolutely. I've even sat down with government people, and this is lately, who's saying that sometimes they find it hard to differentiate proposals because they can see AI and everything. And so they are using orals as a differentiator. Now I might add that you can use, AI to help you produce your orals too. Now is it gonna write them for you? No. But is it gonna help you stress test them? Is it gonna help you, you know, maybe even do early reviews on them or help you write scripts, you know, maybe if you want to, you know, improve upon them, absolutely. Yes. And then one more question, Leslie. Do we have more... Do we have time or... Okay. We sure we sure do. I'll be quick. Okay. This is a really good one, and I'm gonna mispronounce your name, Jadon. And I'm so sorry if I mispronounced it. How are people integrating more full business development life cycle knowledge? I'm assuming in the proposal or just maybe integrating a... Integrating across business development through proposal? I I think that, people are building AI or other types of tools to integrate their systems right now. I think that's a coming attraction in 2027 that we're just seeing people do right now to integrate all of their business processes, connect their data, and maybe, provide more stringent agents and use cases around using that data so they get better results. I hope that answered her question. Okay. And then maybe one more question. Michelle is asking, and I'm not sure we can answer this, but I'm gonna toss it to you, Brenda, anyway, and then maybe, Michelle could follow-up with you. But have you noticed any AI GSA rejection letters? No. Okay. Not on the grounds of AI, but on the grounds of just a proposal. Okay. That's fine. Alright. Let me look for one more question. I think we got a lot of them. Susan, I didn't I didn't... I scrolled through. I didn't see a question. I'm sorry. Maybe you could resubmit it after the webinar. Do we know... Luana is asking, do we know what AI tools the government is using? Yes. Like I said... Mentioned before, they're using a plethora of tools. There's commercial tools out there just dedicated to them as well as our language models. So, yeah, I've talked to quite a few government people and they're using a variety of tools. Okay. Just like we are. Yes. Sister, we are. She's got a lot of tools. And, Leslie, we're coming up to the top of the hour. Yes. I think we are. Great, number of questions. We're glad we were able to get to quite a few of them. So before we officially conclude, we just wanna remind everybody that you will be receiving an on demand recording of today's webinar that will be sent out via email within twenty four hours. If you are going to or if you have, downloaded the links to access the book or any of the guide, just remember when you go to download, the book to turn off your pop up blocker so you can get into that and get that download. I would just like to take an opportunity to thank our speakers today. Thank you, Nora. Thank you, Brenda, and thank all of you for joining today. If we weren't able to get to your question, we got so many great questions today. We will try and do so, follow-up with you offline. But, otherwise, we hope, that you will visit Deltek.com for future events like this. Thanks everyone for joining us, and have a great afternoon. Thanks, everybody. Thank you. Bye, everybody.