The Red Review - What 2025 holds for AI in Bids and Proposals with Javier Escartin
In this episode Jeremy is joined by Javier Escartin, Founder at bids and proposal AI platform DeepRFP. They talk all things AI in winning work, from understanding the landscape of specific platforms versus ChatGPT, the market of providers, what 2025 holds and the future of AI in our world of business growth.
DeepRFP.com is a kit of AI tools and agents that help bidding professionals prepare better proposals faster. It is mostly used by small and mid-sized businesses across industries and bidding consultants to respond to RFPs. It was founded by Javier Escartin, a Proposal Professional who has been over 13 years in the field and shares his insights in his personal newsletter at jescartin.com
For tools: DeepRFP.com
For insights: jescartin.com
Social: LinkedIn/escartinjavier
Transcript
[0:00:00] Jeremy: Welcome to the red review with me. Jeremy brim, the red review is brought to you by growth ignition, the transformation and capability development business all in the work winning space and the bid toolkit, its product set in bid process and training videos.
[0:00:16] Jeremy: So hello and welcome to the red review podcast with me. Jeremy brim, welcome back. First episode of 2025
[0:00:24] Jeremy: and an exciting one tackling one of the big subjects. I guess AI with my friend Javier. So welcome, sir. How are you?
[0:00:46] Jeremy: So I'll have to brush myself off and make sure I ask the right questions. So would you mind giving do a little bit of an introduction? Give us your, your sort of background, current role, and then we'll get into things.
[0:01:03] Javier: I like to call myself an engineer who loves proposals, because I'm actually an aerospace engineer who started working in the European space industry in technical stuff, like engineering stuff. So after a few years in very technical positions, I moved to project management and then business development. But the thing with the business development in this particular industry is that it's just sort of sexy name for a job that is 80% about building consortiums, managing proposals to win big government contracts. So that's in that role is where I sort of learn RPS, bidding on what's our profession. Then I quit a job, started full time freelancing as a proposal manager at that point of time. Wasn't anymore about just aerospace, but most of my clients were American tech and engineering companies. There's more more, like a small business, businesses that need help with RP some proposals. And I did like that for like, four years full time,
[0:02:11] Javier: until I saw how good AI was becoming, and decided to launch the Papi, sort of with this idea of bringing advantage of AI to our profession.
[0:02:22] Javier: And so now that I founded the Part B, like, one year before chatgpt happened, and so that's, you know, that's kind of, I use that as a joke sometimes, because at that point in time,
[0:02:36] Javier: this was a, like, this idea of using AI for proposals was, like, a very small niche, like, literally, we were, like, three or four, five, maybe companies supporting this worldwide. But then after chatgpt, happened in, what was that? November 2022,
[0:02:54] Javier: you know, the space sort of boom, to a point where I, you know, I just done, sort of count them anymore, like, you know, every week you have someone trying to do something with AI proposals, trying to wrap chatgpt APIs for to do some big product. So, yeah. So that's
[0:03:17] Javier: sort of my background. One important thing to note is that I still keep running the services business, sort of this freelancing for proposals business, but right now that's more like
[0:03:30] Javier: laboratory for testing things. How, you know, testing use cases, seeing what works in the you know, in real life, in real life. But the priorities is debug piece to feed those use cases into, into tools, into
[0:03:47] Javier: into technology and so forth.
[0:03:55] Jeremy: Very good. So, so your products deep. RFP,
[0:03:55] Jeremy: so tell us about it. I guess where, where did the idea come from, bearing in mind like you say it was before that big kind of bubble of chat, GTP, before everybody really started down this line. What was the kind of epiphany moment and then tell us about what the platform does?
[0:04:12] Javier: Sure. So the moment I realized watching that summer,
[0:04:20] Javier: autumn 2021, something like that. I just, I've been, you know, being in an engineer I, you know, always play with with tech and software, and I'm always interested in this stuff. And also in that point in time, I was full time involved in, in doing grid management and be greater, for writing for for customers. So and I just wanted to optimize stuff like in the thing I was providing the services. And so always have this sort of intuition about AI, about software automation for doing some of the repetitive tasks that we that I was I was having to do manually, back those years, and then just discover. There, almost by chance. What was at that time, GPT three, which was this model that started to look very well, but still, you know, very dumb compared with what we have today. But then sort of the predecessor of what ended to be chat GPT. So I started with playing those models with open AI, which at the time, wasn't a very well known company, so sort of had that idea like, Okay, this is this is going this direction, so why don't we basically try to exploit that to my service business? And that was sort of the original idea. But then it started to take form. Decided to make the to take the step to build the sort of start prototyping something with the commercial mindset and building and building tools. And so that was the origin of the company. But right now, dpap is the best way of thinking about deep up is a kit of different AI tools and AI agents that are sort of like mini applications, or Mini Apps focused on solving very specific use cases. And those are use cases I have seen myself when, when you, you know I was full time doing this, but also, I see it peers meet every day. And so we have tools for writing responses to requirement. We have tools for editing a Ruby when we have now built to as such a matter expert, sort of, you know, having AI playing the role of an engineer or a technical expert, instead of used a bit wider or big manager. So we have these mini tools that optimize for doing one specific theme only. And users can run these tools, like, as in a kit, you know, you just have this need you go to that tool running, get your outputs, keep moving. So that's for the tools part. Also recently, we have launched AI agents, which is, you know, these are live since November last year. So 2024 so pretty new. We have a set that work with, with analyzing RPS and generating full reports on on that analysis of that RP reports that look like complex matrices, but can be about anything. Actually, the scope of analysis is free, and then we also have agents to sort of draft full proposals. So this idea of because that's something that users have been demanding since I started deep RP, until basically today, there is this need for this use case where you use, upload your RP, upload your winning points, or your past proposal, or whatever, and you get a full draft that, you know, 50 page draft that is 70% ready. It's not your final proposal, but you don't need to be sort of iterating with them, you know, with the AI, as you would do when using tools or using chat boards and copilots. So that's, that's the way of thinking about the party. Is this key of tools and agents?
[0:08:15] Jeremy: Interesting. How is it different from just using chat GTP, or from the other the platforms in the space that are developing?
[0:08:27] Javier: yeah, so Well, I like, I like to think of depape more in comparison with the rest of the professional tools in the space, and not so much with end consumer applications like telepathy or Microsoft compiler. And anyhow, there are, like, maybe mainly, three ways in which TRP is different. The first one is we focus on solving real use cases. So it's not about which model, AI model powers what, but the use case? Okay, so our priority is to solve a real use case in bidding and proposals, and then for that, you know, sometimes in order to do that very well, we need to use a perpetrator model, or we need to use an operations model that we customize for that particular use case. Or we can use third party model that we have sort of focus on solving that particular use case with documentation, examples, etc. So that's one key difference, is that we are not trying to first, we are not tied to any AI provider or whatever, but that our priority is to solve the real use case in building and we don't really care much about the technical stuff, as long as it solves the problem. But that's one the other way. Deeper piece is different. The second one is that we don't train any AI model on users data and to and to ensure that the. Because, you know, some vendor could promise that, but to ensure that that is actually the case, we don't even keep or save the data from from users, ranks and so, and that's one way of protecting
[0:08:27] Javier: them, sort of to ensuring confidentiality and protecting property info. Because even if, you know, imagine your company, like the company gets acquired in the future, that new management, like cannot train AI on your data because we didn't save your data and so, and that's, you know, that actually was a hard hit to take regarding marketing, because it's, you know, it sells if you if I tell you that a tool is going to learn from usage, and it's going to be better with your use, etc, but for that, I need to keep your data, and then you know that there is a risk of some, some, you know, some sometime down the road that that data use for training AI models, you know, that's like The part of the party learning when you use it, it's a good marketing hook. So we had to compromise that to renews to that in order to ensure the confidential data and and that's actually a big difference between the Part B and what I'm seeing other vendors do in the space. And then finally, this is more like, type of user, type of market we are after. A big way in which tipr P is different is that we are focused on like, if you you can imagine Software as a Service in different quadrants, you have this very expensive solutions that are after the enterprise deal that need to, you know, a sales force to pursue those deals, and then you try to get someone in a multi year lock in contract because, you know, complex software, enterprise level, whatever. So we are in the opposite corner of the quadrant. We are, like, very affordable, self serving, no risk deal part of the software industry. So basically, we offer a pure software as a service experience where anyone can get a free trial. Then, you know, upgrade to a plan with a use a credit card, and if you're not happy with that, you just cancel it. You know, you can do monthly billing so you can cancel it after a month, and that's it. There is no big risk, no big investment. And you could be using the party without having to talk to anyone in sales or, you know, even have a demo. And this is the reason that, why the party is very popular, between or among small to small businesses, freelancers and some mid sized companies also regarding this. You know, of course, you know, we still do the demos, like if someone asks for the demo or you need some customization, we have done that a couple of times, but most of the time you see this idea of this is a easy to use self service, very low risk software. And you know, one evidence of that, one proof of that, is that our pricing is public. So you could go today to the website check the prices for the clients, something that you know, it's not that common in the other vendors that are trying to go after this enterprise type of So, yeah, that you know, that's sort of the three ways in which the puppy is different from what I see in the in the RP and proposal space. And that's,
[0:13:39] Jeremy: that's really interesting. So, so you're different from chat GTP, as in, your platforms, only looking inside the company's data. It's not sharing the data externally, is it? It's, it's just fantastic lens on the company. Stuff
[0:13:55] Javier: that's for sure, the confidentiality part, you know, we improve on, on what end consumer applications do, but also is very different in from chat GPT and other co pilots or chat bots, is because the user experience is also different. So what we have is these two parts of the tool, sort of sorry, them, the kit. So we have tools. Those tools look more like a web form where you use, provide your inputs and run it, and you don't need to care about anything else. You should get these you know the output from that, and you have different tools optimized for different use cases. So depending on what you want to do, you need to choose the tool, run it with your inputs and you get the output. But you don't need to learn sort of how that AI works or interact with it. Sort of work on your how you talk into the AI, how you directing the work. It's more like one of tools like here, my input, give me my response or my summary or my edit or my technical approach, whatever you. Doing. And then the other one is running agents. And the agents are even easier to run because it's just about selecting the right agent, uploading the documents, the inputs, which are full files, in the case of agents, and just run it. And then you go for a coffee or whatever, because, you know, it takes a few minutes to complete, and then you go back and get your output. And the feeling is more like working with a remote coworker than actually running tools. That's, well, we will talk about that in a minute, but basically, that's, you know, the what? I think I'm bullish on the agents idea. So it's not only about data, it's not only about pricing, it's also the user experience is different in the way that people are using chat bots or CO pilots.
[0:15:50] Jeremy: Interesting. Thank you. So yeah, because it's it can be an expensive game onboarding a platform like this at a corporate, big enterprise level. And so I think you've probably got quite a smart plan in terms of, you know, a free trial and monthly billing and all of that kind of stuff, transparent pricing. I think that's really clever, actually. And because there is that, it will be very handy for people to be able to dip their toe in the water and really be able to see what this stuff can do if it's their first time out, beyond looking at just using co pilot or something. So I think that's quite clever.
[0:16:27] Javier: yeah. And also, you know, since, since I've been I've been busy. You know, as I told you, like the whole project started, like one year before chatgpt happened. We, like, we were getting a lot of leads, also from the corporate world, like the enterprise level, like we have talked with, you know, very big companies about this AI thing before chatgpt And after chatgpt happened. But as a small business, you you you realise, like, that's sort of not your game. And so I think it's is the right strategy for us, but also is sort of, you know, then the structure and the business that you need to put in place in order to go after those enterprise deals, that requires, you know, a whole different approach to business. And also it changes your pricing drastically. And what's what I'm seeing happening in this, in particular our space. But also, you know, it's literally in every software as a service in the B to B world is like companies that go for that often raise fund from venture capital, yeah, that that racing requires them to go after those enterprise deals. And then you need to build the sales team. You need to have, you know, your account managers, your sales rep. You need to pursue aggressively to this executive to try to close deals in the, you know, in the ranges of $100,000 and so it's a whole different game that we basically are not trying even to to play and so, but yeah, in the, you know, in the other part of that space, I am, you know, we are very comfortable doing this, more sort of easy software for building teams that you know doesn't break your budget and is, you know, very transparent from the beginning. You know, we are not charging anyone differently because, you know the pricing of public and you don't need, you don't have the hassle and the risk of selecting this vendor, because I'm going to be committing for three years contract, because this is sort of and I need to onboard all my stakeholders, whatever. It's a different game. Like, small businesses need more, like, you know, quicker solutions than that. And I just, personally, I just hate, go to some software company try to use the tool or whatever, and the only way to move forward is booking a sales demo with someone and then move to the next one and the next one until the third iteration. You don't get the pricing, and the quote is like this, you know, for me, it's, you know, I don't like that personally, and also why I focus deep up in this segment.
[0:19:19] Jeremy: understand that's quite smart. And to be honest, I have mixed feelings about private equity and all of that space anyway, it's a really powerful tool for unlocking growth in certain ways. But as a business owner or a leader, it's quite, I think it's quite stressful. It's not, it's not for me. I mean, I, I get people approach me all the time. You know, I left. I was an authorised trainer with the APMP until the middle of last year, and I left because of a a difference of views, let's say, and people approach me all the time. You know, will I start a competitor to
[0:19:19] Jeremy: APMP? And there is, there's a dramatic business opportunity for somebody, it would be easy to absolutely destroy them and do a. Fantastic job. Better quality, good integrity, you know, all of that kind of stuff, because they've got some Cornerstone issues that it would be easy to to overpower them on. But I just you would need big money behind you, private equity stuff, the same as the kind of big software as a service lot, and you'd have to go after lots of big corporate deals, like, say, corporate memberships and all of that kind of thing. And it's just not for me, really. I'd rather just have a nice life and run a great business, enjoy working with my clients, you know, happy days, so I can kind of get the vibes, I can see why people do it, but I have a number of clients that are private equity backed and the metrics that are put on them, the behaviors, the sort of challenges it's, it's not for everyone. It's definitely not for me. So I can, I can sort of get that. So best of luck to people to do that.
[0:20:57] Javier: Yeah, I know. You know, linking to that topic, if you see them, like I see, I would expect this year, and I'm beginning of the next one to like, because the market is like all markets are, but our space also, you know, it's kind of saturated of AI tools right now, like, you know, we're talking about hundreds, and many of those are busy back, for example, I told you the part pi boosted the company, so no risk there. But I see some companies doing something, you know, similar things, or actually going after these enterprise deals, but anyhow, sort of trying to use AI for proposals on our piece stuff. And I don't think all of them are being able to close the many enterprise deals they need to survive the next round. And so it's in I think it's going to be interesting to see who sort of survives that race for the enterprise deals. And I think that you know, if you see the sort of when these companies raise money, and sort of the typical window of one and a half two years, they give you for the first sort of milestones for the next one, I think this would be the year where we see some of them actually winning that game. I'm just some of them just going bankrupt and closing the companies. So I hope maybe you know, I would expect that in a year and a year and a half, we will have less players in that space. But anyhow, that's just my guess.
[0:22:41] Jeremy: No, I could agree. It's usually the sort of thing when you, when you get a sprint like this with a new technology or a new innovation, you do get a sort of, sort of that beta max versus VHS kind of thing, isn't it that you'll, you'll find some will fall by the wayside, and it'll be interesting what happens for the clients that have signed up with the wrong ones that go bust when they've or closed because they've gone and put all their eggs in those baskets. It's, it's a real question that I would say, Yeah. And also,
[0:23:13] Javier: you know, then this, this, and I'm talking about personal experience now, is this gives us kind of an advantage, because, like, we can deep RP can really be focused on trying to lead on agents, lead on research and development, like trying to the last thing, see how to adapt to that, and don't worry much about sales. Or at least, of course, you know, we do care about marketing and sales. But the the thing is, we are not stressed by having to close those many deals this year. Otherwise we are out of business. So, you know, this is like for the wrong run since dpap break even and started to generate more money than it spends, let me use some other of like, like this company can be here for the long term, and so that's also a nice advantage of not being in that race. And also I think that's, of course, my personal opinion, but can bring more value to the space and to the professionals and to the peers that are doing this stuff, because at least you have some companies that can be focused on fully taking advantage of AI to solve real use cases, and not so much the pressure of, you know, going after this big enterprise deals and being in that race. But yeah, you know,
[0:24:41] Jeremy: anyhow, we would, well, it's a nice segue. So we were going to talk next about what 2025 holds for deep RFP, but more widely for AI and bidding more generally as well. What do you think the themes are? What do you think the next steps are, or the next innovations, where, where we're going to take it in? This game.
[0:25:01] Javier: Yeah. So, for for sure, I believe AI agents are going to be important, a very big thing, because of how implementation is going to change with them. So, and I think the I think this is going to be like across the board because, and the main reason is because of how these agents fit into current workflows. So as I was telling you before, it really feels that you're working with a co worker in remote you know, you use ask for some tasks and you get the outputs and you just don't need to do anything else. It's not like writing tools or chat bots or copilots, where you, you know, with copilots and chatbots, you sort of need to micromanage them, but in this case, it's more like delegating tasks. And so I think that's only that is going to be fitting better on most people's most teams way of working. So I think agents will drive implementation of AI, but also, not only because AI will be more used because of agents, but also we have current limitations in AI systems, and I think it's important for everyone to know, but basically, there are, like three main limitations we have with current state of the art models. Okay, one is inputs, sort of it's very easy to overwhelm AIS with too many inputs. So people have this idea of trying to pass the whole content library, and you use, can do that. You can. You need to be strategic about the inputs that go into its request in order to maximize performance. Then you have the limitation of outputs, which is sort of the length that you can generate in one run, which is sort of one thought stream of the AI, if maybe they are not thinking or whatever, but that's more like a philosophical debate. But anyhow, this this idea of, in one run, in one thought process, what they can generate that's right right now, around two to three pages of content per run. And then you have the hallucination problem, which is some of these systems making stuff up, like, you know, using fake data or do something like, they don't know something, but they tell you they do, and they are free with some past project that wasn't there, etc. So these are the limitations you get with AI generative AI large language models, as currently they are. And this, you need to manage these limitations yourself when using tools or using chat bots. Now with agents, you can sort of orchestrate the process behind the scenes so the user doesn't they doesn't need to take care of any of these limitations, because you can select the inputs statistically behind the scenes, depending on what the request was. You can generate full drafts doing this iteration like maybe the AI is just doing two pages at a time, but you behind the an AI agent, at least for example, what's happening with the one that we have that last proposals is that you have this process when one AI is playing the manager role, sort of deciding the table of content, which winning points we want to use for each Section, which inputs we not consider for each section, all that stuff. And then you that manager is calling AI writers to develop every section, and then you iterate on that process and but all this happens behind the scenes. So the user just, you know, uploaded the RP, and now it's downloaded a full draft. So I that's why I think agents are so powerful, because of the way of working with them, which is, you know, more like what we do nowadays, and also because you can build in the them, managing this, workarounds, limitations. So at least what we want to do in this year is we're going to keep expanding. The set of agents we have and the use cases that they discover. And I think this is going to be like across the industry, like what's going to happen with with with tools. But also I wanted to tell you about Jeremy, about a concept I've been exploring, like the few last weeks. It's a concept that, to be honest, I still don't have a good name for it, but is this idea of having a top layer interface where you don't really need even to run tools or agents. You know, it's just sort of you ask what you want to do, and this sort of built a proposal manager, is the one that has access to
[0:25:01] Javier: the whole kit of tools and agents and then ensures that you know you provide the right inputs for the what you want to do. And then runs the tools in the background and runs agents in the background. And sort. You stop having to interface with an application through this software interface, and you just talk to it and it knows, sort of masters what to call and when to call it and with which documents, etc. So I don't know how to call that yet, yet, but we have a prototype, sort of conceptually, in the buildings. And I think this is, Well, right now, this is just too, too much into research and development to say it's even a thing. But then these are, this is a concept I want to explore this year about this idea of, you know, and in other spaces, like, for example, if you talk about personal computer companies like Microsoft and all the big ones are exploring this concept of like, you have a computer, we have a lot of programs and tools, etc, but you could have sort of a layer of AI where you just ask for things, and that layer knows how to open your navigator, or how to draft an email, how to run that tool or this other tool, but you just interface the computer, like through conversation like anyhow. So that's what I what we're gonna focus on, on, on 2025 so sort of developing the agents a lot, and investing there a lot, covering use cases. And also trying to, as always, with deep RP, we are trying to go in the to go to the yet with this technology, and so exploring this concept of interfacing differently with a web application, this idea of you don't need to be to even care about where the tools are, or which you know how to run each tool. You just ask for things and and then we like this AI layer takes care of what to run, depending what you want to do, etc. And of course, you can always access the traditional software application, if you will, and run the tools yourself or run the agents yourself. But sort of way of prediction friction with AI, which I think is key for people to benefit from it. Yes, it's exploring this idea,
[0:32:08] Jeremy: sort of like an AI somalia or something.
[0:32:13] Javier: Yeah, yeah, a compass. I was thinking about a COVID, a compass or a master. But, yeah, it's still too early to even to give it a proper name. No,
[0:32:25] Jeremy: fair enough. And, and what? What do you there more other general themes beyond your business in the world of sort of AI with proposals that you're seeing? What? What are the themes for this year? Do you think
[0:32:34] Javier: so that then agents are right now, big in every space, many companies are exploring this idea of having sort of autonomous AI is doing tasks on behalf of people with restrictions. Of course, because of the limitations of the technology, you cannot have like a like, in proposal, you cannot have an AI to take over a full RP and try to manage the whole process yet, because, you know, not that smart enough, and hasn't, doesn't have the planning capabilities. But anyhow, agencies is, you know, being big. But also, there is this debate on, is AI hitting a wall in terms of development and capabilities or not, and that's a debate that is still, you know, being arguing about we're going to be the good news is that we will have some proof, one way or the other, in the first quarter of this year, because the last Generation of models have been trained so far and are about to be released. We have an answer from open AI, but also, you know, x ai and other companies are building and training this last generation model, so we will see the new capabilities we get with those. And that is going to define, I think it's going to define the next few years in software and AI overall, because if AI sort of hits a wall and stops having these emerging capabilities that have surprised everyone, then we are very much lacked or stuck with the intelligence that we have now. Yeah, you know, models will get a little bit better, but it will be like sort of the same thing. And then everything is going to be focused on how to drive implementation, how to ease the use case experience, user experience, use cases. How to, you know, bring this AI to all the applications and software we use today. But it's not going to be like changing the the approach that you already see in different tools and vendors. That's one option. Then the other option is like the exponential curve keeps up, and then if it keeps going, then nobody really knows, because it's going to. Depend on the capabilities of of what these models are able So imagine you get one model that is able to go through your IP and through your library content, and then it's also able to communicate within your company, and sort of becomes your virtual proposal manager, but in a sense that you don't really need to micromanage the whole process. So if you ever get that capability in AI systems, then the whole software industries is going to sort of adapt to that. And probably the tools that we're going to see at that point in time are not anything like the property today or any of the vendors in the space today, because of how big they live in the capabilities of AI. But again, this is going to depend on, on what happens. And the funny thing about this is, like, I'm not sure enough people understand this, you know, in you know, like overall, in general, but this isn't, haven't been designed to, to to do any particular case. So we're talking about systems that have been about use training particular algorithms on huge amounts of data and give them crazy amounts of compute power, and then you start having these emerging capabilities. But no one designed the models that Power Chat GPT to write proposals that wasn't a design requirement or whatever, it's just like, Okay, let's put like the whole internet into this thing. Give like this many data centers to this thing. And this happened, and now this thing knows how to code, writes very well. Content. Can sort of think, brainstorm with you, etc. And so the thing is, the question is, what happens when you scale the next step? And nobody has the answer yet, even we probably get some in the next few months. And that's going to shape, in my opinion, that's going to shape the the whole industry, depending on what happens at the other side of the new generation models,
[0:37:05] Jeremy: interesting, slash terrifying. It's really fantastic. And as I'm getting a bit older, I guess, yeah, unnerving, but there you go. I think I've been managing the risk of our business in terms of exposure with this stuff. For a while, we've been standing up AI proof businesses like property development stuff, and I focus these days. Most people come to me and want me to teach them to write bids and stuff. And I do need a bit of help this year in developing some extra curricular modules on AI and how to interface with it, generally, not product specific in this space. So we're doing that at the moment. But actually I gravitate more towards capturing Key Account Management and things where it's still about humans and relationships and building trust, et cetera. Although, of course, I'm interested in whether some of the vendors start to point AI to support that in terms of researching prospects, you know, under helping you develop your strategy propositions, you know, understanding the personality DNA of the people will be interfacing with, and who the real decision makers are, and all of that kind of stuff. There's some of that going on in the CRM world, plug ins, the sales force, etc, but I'm interested what the interface will be by the time it gets to proposals,
[0:38:28] Javier: strategy wise, yeah, yeah, that's, that's right, and that's sort of one of the main points right now, is that, you know, AI is gonna is spring up time everywhere, you know. And many use cases are quick wins everywhere. And people, so people, people are that are using AI are getting, you know, free time back. And now that's, that's sort of the newest equilib, like, you're not gonna gain an advantage just because of that. Now what matters is what you do with that type that's going to be what matters. And then you have this sort of debate on if you go for bidding more, or you go for bidding better, and that's honestly, that's something that every company needs to answer, because sometimes it makes sense to be more for more, because you have the capability to execute those sort of, the capacity to execute those contracts, but also you have a huge competitive advantage, and you just need to be more to be more contracts. That's, you know, not the most often case, but sometimes as a solution, most of the time, I think the Wiser strategy is to invest that time in bidding better. And that bidding better, what it means is the sort of things that you have just described, for example, how we're going to capture better insights for from customers and stakeholders to build into proposals
[0:38:28] Javier: strategy wise, yeah, yeah, that's, that's right, and that's sort of one of the main points right now, is that, you know, AI is gonna is spring up time everywhere, you know. And many use cases are quick wins everywhere. And people, so people, people are that are using AI are getting, you know, free time back. And now that's, that's sort of the newest equilib, like, you're not gonna gain an advantage just because of that. Now what matters is what you do with that type that's going to be what matters. And then you have this sort of debate on if you go for bidding more, or you go for bidding better, and that's honestly, that's something that every company needs to answer, because sometimes it makes sense to be more for more, because you have the capability to execute those sort of, the capacity to execute those contracts, but also you have a huge competitive advantage, and you just need to be more to be more contracts. That's, you know, not the most often case, but sometimes as a solution, most of the time, I think the Wiser strategy is to invest that time in bidding better. And that bidding better, what it means is the sort of things that you have just described, for example, how we're going to capture better insights for from customers and stakeholders to build into proposals. Spend time there and. A strategy, building a strategy, and sort of game theory strategy, like where your position in terms of quality and scope and and then pricing all that, all that stuff, and also improving the content itself, like spending time on on even if you are working with AI to improve that content, but spending time driving that process. That process of just not less. Don't use what we already have, but improve that and feed that into the AI. And so it's gonna also a trend we are seeing right now is that, you know, more and more procurement and in live presentations and sort of live interviews. One of the reasons that's happening is because they are having a harder time to differentiate between good and bad vendors just based on the content, because of AI and, you know, a new sort of is not super new field, but you could also invest the time on, on improving your sort of your skills, on winning those deals that require a live presentation, and, and, yeah. And regarding technology, I agree, like, for example, for for the sales part, like, there is a there is a paradox in sales, which is that the best people, the best sales people, are very busy selling. So are the ones that input less content to CR reps, because they are just too busy jumping from call to call, meeting to meeting, and no one takes them accountable because they are bringing them, you know, more revenue than anyone else. So you know, this type of sales rep that they start in the team, and they don't capture insights in the CRM because they don't have the time literally, and then CRMs gets populated with inputs from the average sales reps, which are Probably the ones that have average insights on medical insights on clients, and so AI is called to change that, because, you know, if you could give that very good sales rep a way of, for example, just recording an audio note between meetings while Driving, and then the CRM takes that audio note, digest it, and an AI that is presented on the use case post those insights where they need to be so people in bidding and proposals can leverage those so and I think that's not going to come from vendors in rpm proposals, more like it's going to be like the incumbents in CRM, but I see hope for for that big advancement there, and also across the process like I think AI is going to unlock value in many of these steps. So it's not only about saving time, but you're going to get better insights and, you know, better strategies. But again, what is going to determine the winners of the future, I think, is what you do with that free time that's, that's the advantage, like, what you're going to do with the free time that these automating, you know, tools and AI tools are giving you?
[0:43:20] Jeremy: Yeah, I think that's fascinating. Actually, you've set some hairs running in my mind there. That's really interesting, because I'm on my 16th client now, maybe 17th, actually 17th, where, when I've been working with their board or leadership team, particularly Finance Directors, Chief execs. We've We've realized, looking at the data, that around 80% of their margin is coming from 20% of their clients. And then they have a few clients where they make some more money, and then a very long tail of clients when, if they actually calculate it effectively, which they very rarely do. If you wind in the cost of sale and looking after those clients, they're making a loss. And so if you cut the tail and just add a client or two to that list of the top clients, You'll smash your numbers to bits and have to bid a lot less and be able you'll probably have higher what we tend to find the latest work we're doing is what's in the DNA of the clients and the projects that are in that initial bulge in the in the sort of bell curve, and it's the clients that engage early that are good to work with, where you have a higher propensity for direct awards and negotiations, rather than competitive tenders. So I think it could be really valuable that you release your time from working on tenders so much using AI, and then put more effort in into your solutions to clients, to make sure you deliver more value than anyone else, and then in how you look after those clients, would be my natural reaction. But I'm going to do some more work on that. Actually. Think that's, I think that's really interesting. If I was given more time, what would I do? It's probably those two things, but maybe some others, yeah,
[0:45:07] Javier: and that's, you know, that's a question that people need to answer themselves. And that's actually one of the top tips I've given, not only in the Part B, because, you know, I'm very active in, you know, I basically talk proposals and technology all day long. So you know, my newsletter, discounting.com and also LinkedIn. But the thing is, one of the top tips are beginning, given the most, is this idea of, there's, there's like a two steps task or project, is first analyze your bidding process. And when with analyze, I mean, like actually drawing diagram flow, like, you know, process where you have the inputs, step outputs, and having that very clear, like, what's happening when you know how you screen our piece? What happens when someone says, Okay, we're going for this RP, and the steps you take, how you know stakeholders and integrating contributors, content, whatever you do in the process, like having that picture there. And then, that's the first step. The second step is to understanding current capabilities and limitations of AI systems, which we have talked about today, but basically this idea of internalizing that computers can now read, understand, write, parse information than ever before, plan, execute on task autonomously, like those are capabilities that computers didn't have two years ago. We need to sort of remind ourselves that that's something that we now have. Everyone has in the phone or the laptops. And so what you have those two clear steps, like your process on one hand, and you understand what AI is capable of, or computers are capable of, then you can sort of put those together, and you go bottleneck after bottleneck, asking yourself, like, Can AI help me with this particular bottleneck? Like, I'm doing this for every bit. Is there a way AI can help me with this particular step? It's probably yes, the answer. And then you just go for that, implement that. And so just do by doing this, you're going to be ahead of most people in the space, and also better prepared to look for tools and to avoid being treated by vendors marketing, because you have a clear idea what you need, and you just want a solution for that bottleneck, which is your car bottleneck, etc. And then as a result of that, what happens is like you're going to find yourself freeing dozens of hours every week, and then the next step is more like a strategic and it's a question that everyone needs to answer within the team, with the context of the company, et cetera, there's not one solution for all, which is this idea of what you going to do with that time. And that's very important question to answer. It is,
[0:48:09] Jeremy: it is, Thank you, sir. There's some really good, thought provoking stuff. I mean, 2025, in amongst other things, is the year of AI for me and my business. I have to say I've been behind the curve, slightly fearful of it, and I need to pull my finger out. And just even in my marketing stuff, which I've taken great inspiration from you on, I have to say with your email newsletters, etc, which people must sign up for, will come to you know,
[0:48:09] Jeremy: I think there's a lot of value to be had that can make me even more of a David or powerful David against these big Goliaths that I come up against all the time. So thank you very much. Good coaching, wise counsel. So where do people find you? Where can people follow you? And all of that kind
[0:48:53] Javier: of stuff. Yeah, you know, if they want to follow LinkedIn is a place I should consider engaging conversations. That's also the easiest way to reach me. Use message me on LinkedIn. So that's one. Then if, if they are into insights more like, you know, using to useful information for every day, like to do better at proposals and meeting, then they can sign up for free to the newsletter. I send a couple of emails every week, and they can do that in just cutting.com with the J and that's for free. And also it comes with a lot of perks as you subscribe. And you know, I share some useful insights there, and also I share your positions and give some free resources. So it's, it's like, if you are looking just for resources and information, that's the place. And then if you are looking for tools, then the place is the Part b.com that's the company page for the software as a service. Yeah. Often, what happens like is, you go to the page, get a free trial account. We have free trials there for seven days. You can play with all the tools, no limits, and then,
[0:50:12] Javier: you know, you just decide if that's for you or if you want, want to give it a try for a couple of months. Remember that you can cancel anytime that this is not like the type of enterprise deal that's going to log you in for three years or whatever. And so those are like and that. But I only would recommend to go to the party.com. For people that are proactively looking for, you know, tools and testing what they can do, and testing these agents, concept and that kind of stuff. So for tools, developers.com, for insights, just continue that point.
[0:50:46] Jeremy: Very good. Well. Thank you very much. Really appreciate the time. It's been great to have you on I'll follow with interest. And here's to 2025 it's going to be exciting. Thank you.
[0:50:56] Jeremy: Thank you for the opportunity to be here.
