Coach Mike: Riskiest Assumption Tests

Hello everyone, welcome to the Founder Insight Podcast. I'm your host Drew Zayas, and today I'm joined by Mike SuperVici.

Hello everyone, good to be back on this podcast. For those of you that don't know, I lead the Alumni Success Group here at the Founder Institute, and our role is to basically help our companies scale. So we oversee a portfolio of over 4,000 companies, and there are very few things that I haven't seen in regards to startup creation and the entire process. So I'm happy to share that knowledge with you.

And today we're definitely gonna be talking about the creation side as we're talking about one of the earliest things that you can do for your startup to help it succeed, which is the RAT, the Riskiest Assumptions Test. Mike, we aren't talking about the animal here, we're talking about the startup concept. So what is a RAT?

Yeah, so I think a lot of people are probably familiar with the term MVP, minimally viable product, and a RAT is another iteration of that concept. It stands for the riskiest assumption test. So the idea is that to find kind of like what the actual riskiest assumption is in your business and do anything that you can to actually test that out. And sometimes it requires building something, but most of the time it doesn't, right? So that's kind of the idea there. As opposed to going out and building an entire product. Sometimes you have to, but most of the time you don't.

Okay, so what makes it superior to MVPs? People have been making those for a long time, right?

Yeah, I don't know if that's superior, it's just that it's basically much more focused and much easier to kind of like be much more actionable, right? Like when you think about what is the— like basically what I would— if you have a list of things that like you know that if this thing doesn't work, your startup will not succeed, and everything's pending on that, that's actually going to help you get to, you know, validation much faster. Let's take an example. So let's say you're building like a marketplace and I don't know, you're a marketplace that connects pet owners with pet supplies. What is like the riskiest assumption of that market, right? I guess if you were to kind of like sit there and think about it from first principles and like really dig deep in into this thing is probably can you actually create a transaction? Can you actually, you know, have your pet owner purchase a pet supply or something like that on this market, right? And that's something that if you think about it is like not something that you need to actually do a product for. What you would do is you would manually facilitate that transaction to see if it even works and if it's something that people even want, right? And so that's kind of one way to kind of think about it.

Okay, whereas with if you're building an MVP, you'd have to build bare-bones whole platform to—

Yeah, you might build a website for that, you know, and all that. And I don't think there's— it's— I just think that, you know, you— it's better to test as much stuff as you can before you build, because the more you know, the better you're gonna build. Otherwise, what's gonna happen is you're gonna throw away a lot of stuff that you build initially.

Gotcha. So this seems like it's all about testing assumptions, obviously based on the name. And assumptions by definition are kind of difficult for us to detect, as you know, everyone's an assumption holder. Few people admit it because they're pieces of these kind of perceived knowledge where we think they're the truth, right? So how can we defeat that bias? How do we identify our riskiest assumptions?

Yeah, that's, that's the toughest part. I mean, as a founder, right, like just being like brutally honest and realistic with yourself because you want to try to like A practical way to do this is to write stuff down. Like, okay, this is the risk. This is the risk number 1, risk number 2, risk number 3, and just kind of create a list. That's, that's what I actually recommend founders do is they, they start to think about like kind of what are the riskiest pieces, riskiest things in their business. And once you have this list, you can stack rank them, right? Sometimes you can do this with advisors, by the way. This is also a good idea. Like if you have really trusted advisors, but At the very minimum, by yourself, you can like sit there and stack rank them. You can have like the top 3, right? And now stack rank again the top 3 and then start with the first one and then try to see if you can do, you know, a riskiest assumption test on the first one and do another test on the second one.

Yeah. Somebody came to me recently with a question about this that I, I honestly didn't know how to answer because it seemed almost like a team question. They asked me, how do you convince other people in your company that they are relying on unverified assumptions? You know, from the get-go, everyone's often really excited to like build something, right?

Yeah, well, so those are actually kind of— that's kind of a different thing in a way. Like, as a leader, you need to inspire people to join the mission. And if you agree that this is a mission that's worth pursuing and, you know, taking the risk to go and do, then as you try to find that, define and help solve that mission or whatever, you will do various tests, right? Like, this, this is the thing with startups. Like, a startup is basically a big experiment, right? Um, and even like the funding stages are like you're funding projects, essentially. It's kind of how it is. And so in this situation, you know, you're going to be running multiple experiments, and most won't work, and some will work, you know, and then eventually one will work really, really well, and If you do this together with your customers, you're gonna find a lot of things out, like, you know, does the customer actually need this? Do they really need this so badly that they're willing to pay for it? You know, will they be willing to need it so bad that they're willing to preorder it, right? And things like that, right? And so, I mean, that's part of the process. I actually think this is really helpful for teams because you're kind of in this together, right? The more ownership you have of the project for your team members have, even if it's a loss, the better, because then there's an opportunity to improve it.

All right, so rats generally start with a hypothesis, right? There's something that you're testing, and hypotheses usually contain two elements. So it's like, A, I believe something about my customer, right? And then B, oh, I will know that I'm right if X happens, right? So what's an example of doing that poorly? What's a bad hypothesis?

Well, I mean, a bad hypothesis is one where a person doesn't really think critically about the entire hypothesis that they've come up with. So, you know, if you just come up with a hypothesis that I would like to have XYZ product and haven't really thought through who else might need it and why, and what the big benefit is of that is, you know, without thinking critically about like why and how and all that, and that's generally a bad way to kind of think about the hypothesis, right? So to think about it more practically, as you're doing research and you're thinking about like this hypothesis, it's okay for you to start with this, I need XYZ product. But then what you wanna do is you wanna try to kind of study like, what the competitive landscape looks like, what, how do people solve this problem now, in order for you to figure out other ways that this product can work and what the actual needs are. Are they met by all these other products that are out there and things like that? And then you can say, okay, look, this is clear, there's clearly a need for this, it's unmet because of XYZ company, like, you know, XYZ this and that. You know, a good way is like when you see like a product that really sucks and that has really bad ratings, right? Like in the App Store or something, but yet it has lots and lots of usage and lots of downloads that people are willing to put up with it because even though it's bad, right? So that's like, hmm, like a really interesting kind of thing to think.

Yeah, it's like a market opportunity.

Yeah. Yeah, nothing you should go out and do that. It's more of a mission. It's again, it's come back to your mission. We talked about this in previous podcasts, but like The reality is, is that you really want to be in a situation where, you know, almost like a scientist, like, looked into this and really researched this well so that you understand kind of how to even create a hypothesis really, right? And part of that is talking to your customers. Part of that is, you know, researching for hours and hours on why something could potentially work. That's what I would say as a first step. And that doesn't necessarily happen always with the problem. It also can be like with an opportunity, right? Like, I would argue most social products are not like Solving a problem, but there may be an opportunity for people to create, to connect in a different way or better. So let's say you're building an app that wants to connect people based on, on positive things versus like, I don't want to say like negative things, but you know, like right now a lot of the social media stuff is based on like, you know, does somebody like my thing? But if you can make it more extrinsic, like by like providing people with help or something like that. Then you say, okay, that's interesting, that's an opportunity, right? Then a good way to kind of think about that is just to start either surveying people potentially, or like running some Facebook ads to see if, you know, people may even want to download this app. That's an interesting way to kind of think about it. Is this something even want— do people want?

Right. It seems like that kind of segues into another question you should like ask yourself, which is, is this the smallest possible thing that I could be doing to test this assumption? To avoid overdoing it. How do you reel yourself in from like just straight up building things?

Well, that's one of the good things about startups a lot of times is that we are constrained by money and that allows us to, you know, well, we don't even have a choice sometimes. That being said, as a founder, it's really important to kind of think about like who wants it and as a result of that, be thinking about like distribution. Because you might not be able to build a business if distribution is difficult and it's too expensive, because then the cost of distribution might be less than the cost— it might be higher than the cost that you can, you know, charge for something, right? So like, what you want to do is— this is why it's like, I kind of use that example, like maybe running a Facebook ad or something, you know, right? Or however you want to do it, because you want to kind of see, kind of like Man, what's, what's involved in getting this out to people? And what would I have to charge for something like this to even work?

Yeah, like what would your cost of customer acquisition be and stuff like that?

Yeah, yeah.

Like, do you recommend setting a traction goal from the, like, the get-go of identifying I need X amount of users to be profitable or something like that?

Yeah, I think it's hard to put that in general. As a general rule though, like, I personally am a big, big believer in like activation and retention before like traction. Because it's really difficult to grow unless, you know, unless you retain your customers, right? Otherwise you're kind of like a leaky bucket. But what I would say there is a first step is just to kind of think about more like, you know, how can I distribute this product? Like, how can I get this out to people? What is the most common method, right? And then when you think about that at scale, it's like, well, you know, it's gonna be like There's not a whole lot of word of mouth here. Okay, then that means I really have to work hard to get it. Like, now if there is word of mouth, then that's a whole different thing. And then you have to kind of think about how that word of mouth could go. But all these things are like really helpful in helping you determine kind of like what could potentially be like the cost of acquiring somebody. And then when you have that, then it helps you really think about how you can make money, right? So a lot of times people are like, ah, well, I'll just make money using advertising. Okay, well then take a look at like what, like, the kind of volume that you would need in order for you to make that even, like, sustainable. You would need a tremendous amount of scale for advertising to work, like, as far as, like, eyeballs and things like that. So that means that you would have to figure out a way to get propped up by millions and millions and millions and millions of venture capital for you to have even a realistic shot at this, right? So, and then to do that, you'd have to have, like, incredible, like, distribution and growth to be able to do that. And it's just like, it's just a good difficult thing, right? That doesn't mean you shouldn't do it, it's just something you should think about, right? So back to the riskiest assumption test for a social product like this is like, do people even care about this, right? Like, do people even care? And you can check, you can, you can figure out if people care via a variety of different means, right? Like I've talked about like a Facebook ad, I talked about you actually going talking to people about this. One hack that I always tell founders is just to not, to just not mention that it's your product. When you go out and talk to people, because like when you mention it, it's your product, then people won't give you good— as good of advice because they're gonna be pulling punches because they don't want to like hurt your feelings, you know. So you can just—

some dumb friend of mine came up with this.

Yeah, so you know, so you— yeah, exactly. You can say— you don't have to lie about it, but you could just like not mention it, right? Because then what happens is then they don't— might give you candid feedback, right?

Yeah, that's very hard to get, super valuable. So you pretty much have to almost fill out a business model canvas beforehand or something like that, or at least think through that process.

Yeah, everybody does it differently, so I'm not gonna just say just fill out the business model canvas, although that's a really good tool for a lot of people. But I do think that it is important for you at the very minimum to kind of just write down what you think are like the riskiest assumptions of this potential idea. And then once you do that, see if you can figure out a test to see it, because otherwise it might not be worth you testing out the second or the third one or whatever, because if the first one is kind of holding everything together and you can't really build a business there, then you're screwed anyway.

Well, yeah, that would be a problem. Yeah. So it seems— so you listed out a bunch of different types of, of RATS that people could, you know, run, from just talking to customers in interview style, maybe like making a landing page, Facebook ads, something like that, pretty basic. And I kind of want to dive into interviews. How do you avoid asking leading questions and kind of getting useless answers?

Yeah, one of the best ways to avoid asking leading questions is to ask open-ended questions.

Okay.

So an example there is, would you use something like this? Right, that's a yes or no closed-ended question.

Right.

How do you use this product? That's a broad question. It's like, so anything that kind of starts with like something like a how or why are you using something or something like this, that really just lets the person like talk, right? And that's kind of what you want is like a conversation, and then you can just like learn a lot from those conversations as you're interviewing potential clients and customers.

Okay, so who should you share your RAT with? Should you focus on just one customer slice, or should you try to, you know, try a few different ones?

Well, it really depends on the riskiest assumption, right? And so if the riskiest assumption— if you're building a really complicated B2B product, or let's say a healthcare product that where the buyer is the actual, like, insurance company, right? Like, what you want to think about, like, is, um, you know Is there enough like there for the insurance company to even consider using this, right? And so you can even make like a sales call to their insurance company to ensure— would they, you know, want to do this for a specific company like, you know, the company that they're insuring, right? So there's like multiple layers of this, right? That, you know, sometimes it's an interview, sometimes it's basically like trying to facilitate that transaction as manually as possible using like a Google Form or something like that. Sometimes it's just like going through like the actual like B2B sales process without even building something, right? Because then what you find out in the B2B sales processes is that there's multiple like influencers in the process, and then there's who you maybe thought were the buyers, but the actual buyer turns out to be like the CEO, which has a completely different need and all that. And maybe your product doesn't solve that really the way you were thinking about it does. Huh.

So you mentioned B2B. I kind of have a question on this. So let's say you're building a riskiest assumption test for a B2B product. And your riskiest assumption is that, you know, a handful of specific companies, very specific companies, only a few of them, but who are obviously willing to pay a lot, like enterprise tech, will be interested in your efficiency tech or something like that. How would you approach them to gauge their interest when you don't have any product yet and you kind of want to avoid sinking those relationships before they start or appearing unprofessional?

Well, you can't really worry about that honestly as a startup, mostly because, I mean, you might not even be alive like later. Like, I mean, it's like the— it's— you're kind of like trying to solve like a Friday problem on like a Monday on with this one. I You can't worry about stuff like that. Like, you need to basically go out and like literally execute. And if it— if somebody says no, then it's okay. Like, it's part of the process, but you do have to try. And so what I would say there is that, first of all, in most good sales processes are not really going out like trying to like close people. What they're doing is most good salespeople go out and try to like interview people and trying to learn stuff. So I would approach it more like trying to learn stuff before you even format any sort of solution anyway. So that's the, that's the first step, to try to figure out if this is like a true problem that somebody would want to pay. To look professional, you create like a professional sales, you know, email, which you can learn about just by going online and figuring out what that is. And you go in and you ask and you figure it out, figure out kind of what the process is and all that. And best case scenario could be like, oh, we really need this, we really need this now. Okay, that's a pretty good opportunity. But the most likely scenario is that you're gonna learn that that's not the right person that you should even be talking to, and you need to talk to somebody else. And it might be really difficult for you to get in front of them. And then you might be like, well, okay, it might take me like a year to get this in. How do I know that this is good or bad? I don't know, like, I I don't know if it's even worth putting a year into this because I don't know if I can charge enough, right? So there's ways to do that. The other thing I would say is you can go to places where a lot of these people hang out and just talk to them casually, right? So there's certainly lots of conferences going to like the after-party. If you're like a broke founder, you know, the goal— you shouldn't be spending thousands of dollars to go to these conferences. But what you could do is you could go and figure out where the after-party is and hang out there. And then, you know, you know, that's it. Yeah, that's, that's actually like, that's actually where the magic happens anyway, honestly, like when it comes to like networking and things like that. And you can really learn, and you can ask some of these questions to kind of learn if this is like something that people want to buy. Yeah, it's true that the RAT thing is much more difficult than B2B for sure, but a lot of it is also a lot of B2B stuff is around sales and how can you actually sell efficiently into a product, into a company. And that's what you really need to figure out really in order— before even the product, like if the product works.

Yeah. So off the top of your head, do you have any examples of really great RATs?

Well, I think that there's certainly like a few. You know, one of our companies, like Stock Card, they started out with like an email, you know, and they started getting subscribers. Like, do people even care, right? And they started out with an email list, and then that email list started to grow, and you can tell that people wanted the thing. And, you know, what Stock Card does is they basically recommend kind of stocks that might be good investments, essentially.

Oh, so they're just kind of recommending it via email at first, just to kind of test.

Yeah, yeah, exactly. So that was, that was like an interesting thing. And then eventually they built a really nice product that kind of manages this and helps you actually really like do really deep research with it. So that's, that's an interesting one that I've seen recently. There's several other companies like doing similar stuff in like B2B where they're just like literally going in like interviewing customers to figure out exactly what their market is and who their market is. You know, we have a company right now that's, that's about to finish the Silicon Valley cohort and you know, they, they started with a completely different customer archetype and then they realized it's like this other customer archetype because they kind of went through this process.

Hmm. So once you run your riskiest assumption tests and you've, you've kind of validated your assumption, if you're right, yeah, seems like you're right, what does that mean? You know, should you start pre-ordering the champagne for your IPO? Or—

yeah, totally, you should do that. No, it means that that's a— it's a pretty good— that means that you've— that you probably should go to the second one. You know, like if you have a list, you might have 2 or 3 big ones that are key that you may need to verify. And again, I said sometimes you do need to do more stuff to validate that. It's not every time, you know, you might not be able to get away with just doing an email thing. Sometimes you might actually have to build something, right? And so I, again, depends on the business. There's, there's some things that you need to build on, but there's a lot of things that you don't. Marketplaces being a really good example where you probably can just start facilitating transactions without building anything. And so that's, that's a whole nother topic. But you go down the list, right? And then if you can get the few ones that are kind of risky, like, okay, this is interesting, then that— then you might have something that's really powerful there.

Okay. And if you, if you run a RAT and you get kind of the opposite results and things aren't panning out too well, how many different RATs should you run to try to validate this? Should you like pivot to a different idea entirely, try with different customers? Basis?

Well, look, you can pivot on your product, right? But it's really hard to pivot on your mission, you know. And so make sure that you always have your mission as kind of like your North Star, and that can keep you grounded into what you're, you're gonna do. So I think it's perfectly fine to kind of like do multiple rats within the same kind of space. And the same type of like, you know, mission for the business. I think it's much more difficult to do that if you're changing like the entire, like you're going from like building a game to building a medical device company, right? It's just very difficult to do that and everything changes, right? I think that if the rat fails is not necessarily a bad thing. I think that it just maybe unlocks another opportunity for you to try a rat in a different place.

And, uh, and I got one more question for you, Mike. It's, uh, when should you run a riskiest assumptions test? When's the best time to do it? Uh, not like of year, but of you thinking through your idea. Should you wait until you have like team members with you? Should I run one today on an idea I came up with this afternoon?

Yeah, again, I, I wouldn't run an, uh, a riskiest assumption test until you actually have a list of assumptions first and foremost that's been prioritized pretty, you know, pretty well by you and possibly even some other people. Then when you have that, then I would try to go out and run it. Now, as I've said before, there are certain things that are just like— they're like just difficult to test, right? Like, none of the advice we give here is like one-size-fits-all. We have a company, it's called Andiotix. They basically make You know, swallowable, like, robots that can be driven through the intestines, kind of like a video game using a remote control. It's pretty difficult to do, like, a riskiest assumption test there without dying, right? You know what I mean? Like, you can, like, you know, I'm using that mostly as an example, but there are certain things that are in there. There are, it's like, first of all, can you even create propulsion for this to work, right? So what they did is that they basically have, like, a burrito-sized camera, right, that they can put in a pool. And they created the propulsion, and then it looks like the propulsion could actually work potentially in, uh, in a human if the right conditions were met, right? And so that was like, okay, well, if— well, you can't have the camera that can be like driven through the body without propulsion. So they tested out propulsion, and then they're like, okay, great, now can we make this like half the size? Then can we make it half the size? And now it's like the size of your finger, and pretty soon it'll be able to swallow it, right? And so that, that's kind of how you want to think about this. But, you know, ideally, yeah, I mean, you want to think about this as early as possible. And, you know, also you want to think about it within like the budget that you have as well, because certain risks, riskiest assumption tests, certain RATS are more expensive than others. But overall, like, the earlier the better, right? Like, this is something we And the Core Founder Institute program, this is something that regardless of the stage of the company is coming into the program in, we generally recommend that they do like within the first few weeks.

Wow. Okay. Well, thank you so much for joining, Mike. This is super enlightening. I can't wait to talk about other things in the future, man.

All right, man. Looking forward to it as well. Thanks, Drew.

Yeah, for sure. And, uh, you know, I only got like one rat joke in there and I have another one. I don't want to say it though. No, I mean, it's just a little— it's a little cheesy.

All right, man, thanks.