A Product Market Fit Show | Startup Podcast for Founders
00:33:30 · Tushar Mishra
But I've also seen there's usually a selection bias in getting feedback Because the kind of people who would typically respond to give you feedback are more on the early adopter side who anywhere on the frontier of like different things. They have seen a lot more. They love to give import feedback. They're more tech forward. And so building your worldview based on what you're learning from them can also be dangerous. And then in certain markets, can build a large business with early adopters and eventually scale. So understanding your market where, okay, early adopters, how much depth do you have in that market, and is what they're saying makes sense even if they're actually paying you money? Like, I
00:34:10 · Pablo Srugo
[00:34:10] S1: think we all understand that simplicity matters, but at the beginning, when you have effectively a blank canvas, it's easy to forget that. And I think even with AI now, if anything, this is more of a problem because a feature request only costs less than it ever did. So why not do it has a weaker response. But in the past, it's like, well, it's gonna take three days. It's not worth it. Now it's like, it's gonna three minutes. So why why are you fighting it? And what you might miss there, this ties into your early adopter point, is early adopter tech types, like, they understand technology in a way that, by definition, the later adopters don't. And so they're more willing to have a product with a lot of features, a lot of things, like, you know, kind of they can go and configure it the way that they want, and that's maybe fine for them. But the average people, the masses, if you will, like, they don't want all that stuff to be cumbersome. Simplicity... And I I would argue for most people, honestly, and for most products when you really boil it down, like, that's what's hard, is sometimes stripping it versus adding. So it's a risk that it's just kinda hard to see it upfront. You start saying yes to things, you start adding things. You're like, okay. Well, then if you have this view, then we should do that. And if you're this type of person, we do that and this. When they start competing, all of a sudden, you know, your product is just so cumbersome and so hard to understand when you see it for the first time as a new user, it's confusing. And so that's what I think you're you're also watching out for when you when you listen to feedback a little too much.
00:35:31 · Tushar Mishra
[00:35:31] S0: And I think one quick thing to add there is a great example is playing out for us in real time here. So if you look at like OpenClaw, when that first like it was all early adopters and everybody like it was complex. There were some agencies that were charging you to set up your OpenClaw and people don't do it and more complex workflows. And then you compare that with Mules now, which is how do you strip that down, make it super simple for somebody who can actually use it every day. That is extremely important because when we would go against the legacy provider in the market that we were trying to replace, they had a lot more features than we did. But people hated it. They hated it for the reason that it was complex. It was too cumbersome. The UX was a mess. Mobile app did not work. So we knew that if we do the same mistake, we can add the same features. Like, that's not the hard bit. But unless you think about the problem from first principles... And great example of that is the training example. If I want to build an LMS and build it feature by feature, I'll build the same version of an LMS, maybe slightly better UI. But if I think about the problem for first principle, which is I need to teach a 16 year old how to quickly start being productive in a pizza store, then I think about that separately. I know that that guy grew up with Google, so I need to have great AI sorts so they can get just in time answer. I am standing in front of a point of sale system, I need to be able to ask, hey, how to apply this coupon code and do it instead of me teaching that in a video or in a silly motion graphic sort of thing, which most LMSs were doing.
00:37:07 · Pablo Srugo
[00:37:07] S1: So let's change gears here and jump to kind of tactical number two. The first year, especially, like, you think about 2023 to 2025 and maybe still now, but but especially the beginning, it was kinda pure outbound as I understand it. And you had a system for driving experiments in outbound. Obviously, everybody's told you got a b test. So, again, these are things everybody knows high level, and yet when you look at operators, there is a very big difference in terms of execution and how well some people do it versus others. Tell us a little bit about how you structured outbound in order
00:37:35 · Tushar Mishra
[00:37:35] S0: to have experiments and figure out the messaging that worked, what converted, what didn't, etcetera. So number one, we actually made it made it almost very, very experimental and very handcrafted. What we did was we created a list and we divided that list by multiple different parameters. So one, by number of location, which is the size of business. Second, by industry. And third, by the actual role of the ICP in those companies. And there were different permutation of these different roles. We then had multiple emails that we had written down. And for each of the email, we had two sets of subject lines that we were testing. And then in the email copy, we try to keep the messaging consistent for a specific sort of ICP or a specific kind of email and experiment with what grabs their attention in terms of the use case. Like what do I reach out to you to solve like that actually works. So don't change that I'm selling instead of a franchise operating system, I'm selling LMS, but try to change within that franchise operating system, what are things that grab your attention more. And we started measuring sort of the open rate, response rate, all of those things and started sending out these emails like by hand. We did that. Along with that, we also reached out to these people on LinkedIn and tracked that all on a sheet. And at the end of every day, we would track how many opens do we have on them and you would call these people. Because we would know there's some kind of interest and we would call these people and then also note like what are we learning on those calls. So once you start to do this, you get a very good sense of what from a messaging perspective sort of good work. So for the first year 2024, all of the leads that we generated, majority of the leads that we generated were just through outbound emails that we were sending out and doing a little bit of calling on that customized or that curated list that we had.
00:39:30 · Pablo Srugo, Tushar Mishra
[00:39:30] S1: I don't know if you have it in front of you, if you can remember, but I'm curious on hearing some of the differences. Like, you know, what did your best email look like versus your worst email, you know, or best subject line versus worst subject line? Do you do you have some of those examples you could share? [00:39:45] S0: Yep. So number one, most replies actually came from email three, not email one. So we would send a cadence of email. That was one. Second, our best performing email was actually the one where we added a screenshot of our mobile app with the brand's logo on that screenshot. Because how a lot of people, they're just scrolling through, but if they see their brand on that screenshot, they feel that somebody has put actual effort in sending this out and they actually now read it because there's so much noise now more than ever because now AI STRs are writing all emails. But that gave us that right to attention, which is now they were paying attention. So that was sort of our best performing. The worst performing email was where we made outrageous claims. So oftentimes when we think of when you send out an email and say, I can help you make this much. I get hundreds of these email a day. [00:40:42] S1: Yeah. I can I can help you double your views? I can help you double... Whatever. Some big, big thing. [00:40:46] S0: Sure. You can. I mean, I'm I'm doing this, and this is my 100% of focus. If I can't, what magic sauce do you have that you can?