Why Most Bid Teams Use AI Wrong (And What High-Performers Do Instead) ✹ Interview with Jamie Horsnell, CPO at mytender.io

Chris: Hi Jamie, very good, how are you?

Jamie: No problem.

Yeah, so hello everyone. I'm Jamie, co-founder and chief product officer at MyTender.io. I was at university, had finished my second year, and was keen to start an AI business because I saw the opportunity there. Me and my co-founder, Sam, were trying to provide AI automations to a bunch of different businesses.

Through that, we came across the pain point of bid writing: bids were very time consuming and people were looking to automate certain parts of it with AI. Once we came across that idea, we doubled down on it and started building the product.

Chris: And in terms of your bid writing experience?

Jamie: Yeah. I mean, it's mainly through MyTender.io. We've done lots of applications to awards and grants. There's a few grants at the university which we won, all of them using our software. And then we also had an Innovate UK grant, which we didn't win, but scored very highly. It's quite a competitive one. So whilst I'm not a bid writer and I don't proclaim to have the bid writing experience, I've done my fair share of grant writing, and I've spoken to, by this point, hundreds of professional bid writers. So I feel like I'm fairly well-engaged to give an opinion on some of the bid writing domain.

Chris: So how did you discover this space?

Jamie: Yeah, great question. It was a facility management CEO who we outreached to. He was doing bids on the weekends, taking time away from his family, taking the whole weekend. He didn't enjoy doing it at all. So he was desperate for any way we could help automate that for him. He said he has five to ten people he knows, some of his friends, who also own businesses doing the same thing and have that same pain.

At the time I had no idea what a bid and tender was. After the call I was researching, finding all about it. Then I found out how big the market is, how common it is. Pretty much every industry does some form of bidding. Once I saw the opportunity and the chance to build a product, as opposed to more service and consulting, I much preferred the product route. So I decided to find a technical co-founder and started building from there.

Chris: And what's your role now?

Jamie: Yeah. Me and Sam at the start were very typical startupy co-founders doing pretty much everything across the business. After six months, twelve months, you start to get more mature. I was doing a lot more of the prompting, the AI side. I was quite interested in that stuff. Doing sales calls, landing sales, speaking to partners. But my interest was into the AI side of things and into building a great product.

I feel very lucky that at my age I've got quite a good direction in terms of what I want to do. I love being a product guy. I love speaking to customers, getting their feedback. And in the modern day, an AI product role or AI product manager, I think it also favours people that are kind of technical. That's something I've become over the period of MyTender. Because I understand the AI elements, so much of product in AI applications is how you're wanting the AI to interact with the customer, the interfaces you're selecting to get the best out of the AI. I find all that very fascinating.

Chris: Where do you think AI in proposals is heading?

Jamie: I think in terms of the AI space, the application layer on top of all these foundational models, I believe we're still very early days. With the foundational models, they're showing diminishing returns. Each latest model is not quite as intelligent, or the jump between each one is less. So I don't believe we're going to see too much improvement from that side.

On the application layer, though, I think that's where there's still a massive amount of innovation to be had, which can truly improve all sorts of domains and vertical-specific applications, especially bid writing. In two to three years, we're going to start seeing the gen AI applications become much more sophisticated. The writing is going to get higher quality.

All of the tooling around the bid creation process is going to improve as well, all the way from the capture phase through to doing the reviewing and post-bid submission, feeding that back into your AI model so it can be used and fed in as memory into your bids.

Basically, the role of the proposal manager is probably going to shift from still doing some of that manual, admin-heavy work of trying to find the right data, and potentially some of the collaboration interaction piece that bid teams have to do between one another, to being much more strategic. Much more focused on the customer, interacting with the customer, and working on the higher-value tasks which AI helps to enable you to do.

Chris: So the bid writer stays in the loop?

Jamie: Yeah. I think that's good news in the sense that the bid writer is always going to be involved in that final process. Human in the loop throughout. Really making sure that final submission is very high quality and refined. One of our philosophies as a product is to give as much control and flexibility to the bid writer as possible so that they can use their knowledge, use their skills to get to that final submission-ready draft as quickly as possible.

When you have inflexible systems or a rigid workflow within the product, it limits and hinders the bid writer's ability to integrate their intelligence and the skills they've spent their whole career building up into what they're producing. So I believe there's always going to be a bid writer throughout that process. But the amount that it gets you closer to that final draft with intelligent systems, from that initial couple of drafts and generations, will be closer to that final draft than it is currently.

Chris: What skepticism do you encounter?

Jamie: I think the most amount of skepticism I've experienced is getting people to understand how AI is going to truly adopt and generate in the style that they write their bids in. The formatting, all of the information and knowledge in that organisation, how the AI system is going to generate it in that similar fashion, to that same tone and formatting requirements, and the general knowledge of the business being transferred into the paper.

That's something where, back to the application layer, I think we'll see improvements more and more. There's already things we do like tone of voice, certain other elements. Templates allow users to generate the proposal with standard template features that help get it closer to that point. But I definitely think as the application layer evolves and you get more cool techniques and innovations developed, the AI will be able to get a much better understanding and really be a junior bid writer to help them on pretty much every bid.

Chris: And that education piece for more experienced bid writers?

Jamie: Yeah, I think it's largely around the education piece of some of the capabilities of these models. For example, we have one of our enterprise customers who originally spent quite a lot of time in the capture phase doing their research on the customer. They were unaware of that capability of AI to automate that significantly.

It's about making them aware of the different use cases that can be used with AI within your platform. And helping them through that process. When you have someone who's used to a certain way of doing things, a process they've done for their whole career, to then slightly tweak it and change it into a new way of doing things is quite a massive change in their day to day. You do definitely need to provide that support, help offer the guidance. And through that, you can both end up massively enhancing their efficiency and quality, and also improving their AI skills as well.

Chris: How should bid writers evaluate and buy AI tools?

Jamie: Yeah, I think firstly, it's the bid writer getting slightly educated on the benefits AI can offer within the bid process, where it's going to offer them at each phase. And then also getting a little bit educated on AI more broadly, because certain organisations come to us with much more granular, detailed, high-quality specs where they're asking the right questions. I think if you can get slightly educated, I'm not saying to try and become an AI expert, but if you have a general sense of what high-quality AI looks like, some of the latest innovations in technologies which can be used in software, then you can definitely select a solution better.

When you are taking that to the CRO, you're going to be able to have quite a strong case for being able to adopt a solution. And it's also going to help the CRO make a much more informed, better decision, which they can then derive off the increase in win rate, the increase in revenue, and the actual economics of what an adoption of AI would look like for their business.

Chris: So the bid writer should drive the buying decision?

Jamie: 100%. The bid writers who are going to be using it day in, day out, they're the ones who've got the domain knowledge of what they want a solution to look like and how it would actually help and benefit them. So they should have an awful lot of power in that buying decision, because of those reasons.

From the CRO perspective, this is a global, every industry, every business is having to come up with their AI strategy and how they're going to adopt it in the most effective way. Something which so far, I think a lot of businesses are not getting correct.

I'm also an ambassador for the Centre for Gen.Ops, which is how you can build and integrate gen AI applications to deliver real value within your business. They have a framework which can help you with that process and understand the right use cases in your business. It's a community interest company for people or CROs who want to understand how they can integrate AI into their business as effectively as possible.

Chris: What's the quickest way a bid writer can start using AI today?

Jamie: I think the most important thing is to get the excitement for the technology in terms of knowing what it can do for you. So that's probably the first thing. The most effective ways you could try tomorrow: firstly, search up Perplexity and do all your research that you would typically be doing on Perplexity. I think that alone would speed up your capture phase, your bid/no-bid decision massively.

The second thing, which I think is one of the things AI is very good at, is the analysis part. It may not be ten out of ten writing, but if you wanted to feed in the evaluation criteria to one of the answers you've already done, and prompt it to say, can I improve this answer to score higher? That's a very effective use case and can get your brain thinking in a slightly different way. I use that for pretty much loads of different strategic decisions I make as a business or on the product. I'll do a similar sort of thing. So I'd definitely recommend trying that. I think that will help them understand how AI can help them.

Chris: What about using AI in the solution development phase?

Jamie: Good question. I think that potentially in the solution development phase, where you're coming up with the proposed solution you're wanting to offer to the customer, that's an area where I think you can definitely get a lot of value out by running a conversation with it, refining it, improving it, adding all of the right context in. It's something we do in our system. You can add in the subject matter expert knowledge.

You have the understanding of the customer and what they're after, and any things which you can add into the chat history or context: if you've had a meeting with the customer, they've explained some of their pain points, what they're trying to achieve, or anything you pick up from the RFP or the tender. Then that can all be used to really generate a solution that's very aligned to what the customer is looking for.

I think that alongside a bid writer and maybe even a subject matter expert who can back and forth different ideas, keep on refining, is an area which I think could deliver lots of value for businesses.

Chris: And you follow the APMP process in the workflow?

Jamie: Yeah, I mean, that's very much how we've structured the workflow of our software, following APMP and that traditional process. We also accommodate a lot for flexibility because not every business's process is the same. I'm very big on: your AI has to be as flexible as possible and be able to do everything in the background. Whether that's retrieve different information, improve something, or even sending an email through to some of your subject matter experts to then put their input. I think all of that in an AI system is what people should be expecting: a super simple, flexible system that will fit around your current process and let the bid writer or the person responding add their input and knowledge and skill into it.

Chris: What about bid/no-bid?

Jamie: That's something I heard in the very early days of MyTender: there's two people who went on a bid. The person who wins the contract and the person who decides not to bid. Emphasising the importance of that initial bid/no-bid decision. Really being selective of the contracts that you do go after, because even with an AI system which is going to make it much quicker and easier, you still could be prioritising likely better opportunities that you are more suited to, which looks after the resource within your organisation.

Chris: What separates the highest-performing bid teams?

Jamie: The teams which in our initial calls, and after they've signed on, I sense that they've got a really good grasp on what they're doing, they've got a good win rate, and they've got a very slick, functioning bid process. It's where they take it very seriously, the whole end-to-end journey of the bid.

Taking very seriously that initial bid/no-bid decision. I think a lot of it stems from that, because if you're bidding on the wrong contract that you're not very suited to, or it's maybe not your strong suit compared to some of your competitors, then it's a waste of the team's time.

Another main part is definitely on that storyboarding phase. Refining the win themes and understanding the pain points and your unique selling points. The teams that I speak to that do that, they definitely come across like they're going about it in the right way.

Now, that does depend on the timescales of the contract. If it's a very quick turnaround, you don't always get that luxury. But spending time at that initial strategic phase of really understanding what you're offering to the customer, that is going to pay significant dividends throughout the process.

And then lastly, ensuring that once you have it submitted, you've got the score back, you've got the feedback from the buyer, then you do have a debriefing meeting where you look to see what you can improve and optimise, and take that into your next bid.

Chris: And you encourage that in the product?

Jamie: Yeah, part of my role is that we have a system where the workflow encourages the user to do that. But it's not mandatory for the final output, because sometimes bid writers won't spend that time at the strategic input stage and the storyboarding. So I think it's about having that within the workflow and in the product, but not as a mandatory must-do step to click on next and be able to go to the next page. It's flexible and there if they want to take that extra step in improving their strategy.

Chris: What about specific industries, like construction and facility management?

Jamie: Yeah, it's an interesting one. They're fairly different, construction and facility management. I wouldn't want to necessarily put them in the same bracket. In construction, you're still wanting to propose quite a good solution. As an industry, it's a lot more relationship-based. There's still quite an emphasis on the technical solution which you're offering. My cousin works in construction, in building hospitals, so I've got a family heritage of construction. I've got a fairly good understanding.

For the bids, I think it's technical and then being able to integrate a lot of the regulation in the construction industry, which is very important in ensuring that you abide by all the construction standards.

And then facility management: from my experience, they have quite a high turnaround. It's not necessarily as technical as an IT sector. So focus on the solution, but it's much more around the complete wider bid you're submitting. For them it's probably wanting a solution which iterates very quickly and learns on what you submit frequently. The better it is at learning from your bids and how you write, the more quickly you're going to see that improvement curve.

Chris: Any final words of wisdom?

Jamie: I'd say enjoy using AI. It's quite an exciting thing once you do make that change to using more AI products and seeing what it can do for your workflow. And I suppose it wouldn't be right if I didn't say, if you do want a flexible solution, then come have a chat and we can find out more about your bid process.

Chris: And where can people find you?

Jamie: Yeah, you can either email me at Jamie at MyTender, or go visit our website. There's a book a demo button right in the top right-hand corner which you can click on. And yeah, I'd love to speak to you.

Chris: Cool, thank you Jamie.

Jamie: Nice. Thank you, Chris.