文字稿 🔗
riverside magic episode 03 boardroom to booth: Humans are under representing themselves in an MA deals. And I've got some evidence to show that and what we can do about it, but I'd better introduce introduce who I am. So hi, I'm Hutton. I'm the managing partner beyond MA with the merger synchronization builders. I think that last word is really important to us because we are builders, we are not auditors, we are not assessors or advisors where people get in there and help you out. And I've got a really interesting situation right now where the it follows on from all of my discussions with agents where in this situation we're at the map stage of merge, which is our end-to-end merger synchronization process. So map includes diligence and we're trying to work out, you know, what what is our buyer or investor looking into and and buying and you know how can we help translate between both parties. So Translate is a really good word in this scenario because the customer has or the target company presented a fantastic data room, lots of documentation, it looks really good, they were very fast to do it. And like many companies in the last twelve months, they've actually cr used something like Claude to create all of the documents. Now I don't necessarily have an issue with that. It does cause us a bit of a pain because it means an a lot more for us to read. But at least in that huge data room, there'll be all of the the data that we think. So the the problem that we had is once we started to review and like talk to people in interviews, then we realized there was quite a mismatch between what has been presented in the narrative of the data room and the data. And they're two different things when AI starts to get involved with MA. The narrative is obviously an LLM taking all of your data Presenting a story and you've given it a context and it's working on your behalf. So in that situation, they've literally got this story stating, well, this is how we do things, this is why, and covering every base, but obviously everything is looking reasonably good. The problem is that the AI does a few things there, it fills the gaps, is number one. So what we're seeing is the data does not match what they're saying. So maybe they're saying b fantastic growth of X percent or things are really running smoothly. But when we actually look at the data, which we're not using AI to look at, by the way, then it's it's a real problem, right? Because they don't match. And so the the that caused a a a gut feel and a concern. And as I've said with agents, same thing with LLMs, when they write your documents, we lose any of the human nuance. Which is why I'm saying humans are under representing themselves in deals today. We had to do something about it. So what can I suggest to you? Well, obviously talk to the target. You cannot just keep that silent and go back to the buy side or investor and say, This is what I've found. You need to actually just reset the whole t project. In our so the things that we've learned and what we were doing already is that half of the team we use AI that we've built ourselves to to understand the data room and the information, but the other half of the team keeps it human. So do you see the benefit there? That's exactly why we could see the gaps. So my absolute suggestion is it doesn't matter how much AI you put into your deal, make sure fifty percent of the work is is assessed and run by humans. Otherwise you're not going to get anywhere. The second thing was to explain to the target that usually buyers and s sorry, sellers and people looking for investment and and buyouts, they tend to show the art of the possible. They speak the the language of positivity. And if you're going to work especially with UK investors and buyers, they speak the language of risk. So they have to meet and all the messaging has to show the risk factor. And I think the key thing there was that what we were really uncovering was the target actually assumed we would use AI. to to do the assessment. So that's my agent speaking to your agent, an absolute waste of time. And secondly, I realized that they also had to understand the seat. They had to put themselves in the shoes of the investor or buyer. And one of the things what they have to do is present they've done a great job to the IC. And the IC don't see a great job if all they can see is upside. They want to make sure every single risk is counted for and and and raised And there's a mediation. They don't they don't actually mind or remediation. They don't mind if there's a problem, but they they do mind if you don't know what they are and there's no cost attached. So every problem needs to be named with a person and a cost attached to it. And if I was you, if you're gonna put something to market, then make sure the humans are represented. And if you're on the buy side or the investor side, make sure and help the target. Put more humans into the deal. As you're gonna see over the next few weeks, I'm guessing, we are just doing far more in-person sessions. It's the easiest way, even if it means international travel. It gives you a better sense and you know, with four hours you can get more than a whole data room full of AI data. And I didn't use the word slop, is proper AI information, but you just don't get anywhere with it. So go and see people in person. and help them reset their diligence if it's not to your standard. See you on the next one and please subscribe. Cheers.