0:00 Everything starts with the customer, not with the product. How much do they want to solve this problem and how much are they willing to pay? >> This is Roman. He cloned the Shopify app, grew it to $500,000 a year, and sold it for seven figures only 2 years later. Then he did it again. >> When we sold Coco AI, we looked at all the things that made us successful building this tool and selling it. And we saw that the high intent field was something. So we built a tool around that, >> but this time with a new strategy. decided to sell first and build later. And it worked. He made $7,000 before writing a single line of code. Only then he built and grew Goji Berry from 0 to $50,000 a month in only 6 months. So you're probably thinking this guy is a genius developer, but he is not. He is a
0:45 mechanical engineer turned SAS founder. So in this episode you will learn how to validate ideas before building. How to clone profitable apps and launch it in two weeks. The cold outreach playbook that gets 70% close rate on demos. And finally his exact 90-day plan if you had to start over. So Roman you failed trying to copy paste a product text to buy in a new market in Europe. At that time it was 2023 and you said that you were new to SAS. So what I want to know is what did you do before building that first project? >> Sure. So thank you for having me. Um so I'm a mechanical engineer. Uh so I
1:31 graduated I think 2016 or 17. Uh then I got bored very fast. I uh tried freelancing. I tried um doing affiliation a lot. Uh I got some great success with a partner. We made quite a bit of money during COVID. Then I launched a mobile game with friends that we sold. And after that I thought to myself, okay, um I'm jumping from business to business. Now I need something to to settle. I need to compound my knowledge. So I I saw that a few people were doing SAS. It was um it was yeah they were doing great and also they were in SAS um since 2014 2015 so
2:17 they compounded over time and they had huge SAS making a lot of money with employees tractions and so on and so I said to myself okay now it's time to to settle to something a bit more serious so that's how I got into SAS I contacted a dev I um made a team and we just launch whatever we could launch >> and that first product failed. Can you explain why? I think you have the reason what >> Yeah. Yeah. Because um I think everyone does the same. Um they don't try to understand what the customer want. They create a product and then they try to sell it to the the the customer they think needs this product. Um so we did exact exactly this. um
3:04 instead of talking to people, we looked at a project that seemed successful. We tried to copy it. So, for the record, it was um it was called one text. It was an American company that just got into YC. And at that time, I thought that YC meant success. Now, I know that 98% of company that go into YC fail, but at the time, I didn't know. So I I saw this company two guys really successful looked successful and so I said okay let's do this because we spent a lot of time looking at every project going into YC to understand um what we could do fast. So we took this company that was uh SMS. So text to buy um via SMS
3:53 and so we thought how to bring it to France, how to bring it to Europe. SMS buying by SMS was not really a thing in Europe. Uh so we switched it to WhatsApp. We spent months uh rebuilding everything with text to buy on WhatsApp. And when we launched, we had one or two pilots um for like like we had one or two customers trying our our product. They stopped really fast because they were not ready to ask for people to pay by text and also UEIE customers were not ready to pay by text just by saying yes. So the product failed and uh we spent 6 months for nothing just because we did not talk to customers and if we had talked to customers in one week we would have known that um we should not launch
4:39 that and at that time I think you learned from your mistake. You said, "Let's find companies already selling WhatsApp tools in Europe making real money and just copy them." And you found uh one Shopify apps, many no actually few Shopify apps making two to three million. You clone one of them and uh in two weeks you launched it and then you got traction straight away. The first thing I want to ask is where did you find the Shopify the Shopify apps that were making two to three million a year? >> Yeah. So there one of them the founder was sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sh sharing the numbers on LinkedIn and the other one um so if you take a
5:28 tool if you take a chrome extension called store leads you are able to see who downloads a Shopify app um what kind of store they have and and everything. So you pretty much get the data. So based on the data we knew and based also on the number of people working there, we knew they were making money. Um so what we did, we went into the Shopify app store, we typed WhatsApp, we looked at all the companies, we created an account for each one of them. Uh and then we looked at which one was the most successful and was the easiest to clone with everything we had worked on. Um and so we found one. We implemented I think two uh main features. We reconted everyone we had contacted for the text
6:13 to buy and we got almost immediately five to 10 clients uh because no one or almost no one was selling it in in France. Um and so that like in two days everyone wanted to pay which was great. So we knew we had something not the best product in the world, not unicom, but something that could um that could work. >> You said you built in two weeks the features. How did you build that? I know this is not you behind the code, but how did you build it? What did your co-ounder use? >> I'm not going to talk about the technical tools um because I don't know much about it. What I can tell you is what we had built with the text to buy was way more uh difficult to build than
6:59 what the competitors were launching in Europe. So we just had to uh change a few features, change a few rules and that was uh okay. Um I know we were using Firebase. Um, that's like the only tool I know. Uh, because really my part was marketing and I and I don't focus at all on how it's built >> which is interesting because as I hear you have the same background as me like I used to be an affiliate marketer. So you two and your role was marketing. When you say we, who are you? Like you and I think two co-ounders, right? So for um Coco Aai the previous company we sold we were four um and now we are three so two people left and one uh
7:46 joined us. >> And where what were your role? >> Uh so when we were four uh we had one let's say two two people doing marketing one doing product and one doing um the the code. And today we have one at marketing, one product, one um doing the doing the code. >> Interesting. And when you say the code um you grew this app from zero to 50kmr in 6 months almost exclus exclusively through cold outreach. >> How did you do that? So we uh targeted almost only the French market uh which is really a small market
8:33 and then we expanded a bit in Spain uh Germany. So how we did it is quite simple. Um I had a let's say a list of actions to do every day. So I had 100 calls per to do per day. I had I think it was 1,000 or 2,000 cold emails to send. Then I had the demos and then I was also doing cold outreach on um on LinkedIn. So it was really contact I needed to contact as many people as I could in my niche and either give them a blueprint either ask for a demo and I knew that when we add a demo we would close 60 to 70% of people. So the the the game was just getting demos and then it was almost always a win. And to give
9:19 the watcher, the viewer an ID, what who was your ICP? Like who was your ideal customer? Like who were you targeting? >> Yeah. So it was a Shopify store making more than 10k per month. Um and if um they were using AI tools, it was a plus, but Shopify store 10K plus in Europe was uh like the main main thing. >> And where did you find them? So um we like we we had many many ways. Um LinkedIn was one of them. Um so on LinkedIn we would find one person one agency or one person sharing results. We would look at everyone who is interacting. We would have a list then
10:07 of course sales navigator uh store leads as well. We would use um fox leads I think. Um and so all all these type of tools to extract the uh founders. And one thing that worked really really well for us and that made us scale is instead of emailing the founder, we would email the customer service with um a demo video and ask the customer service to forward it to the CEO and they would almost always do it. So CEO would see the the video and he would book a call with us. Um so that was that was really a surprise because everyone was saying yeah you need to target the the founder the customer service will never respond but when you do emailing to the customer service there is also a great upside
10:53 it's that you get almost 100% replies because the customer service always reply to you saying okay I will manage and so your email don't go into spam um so you can do a lot of volume that way >> that's really interesting so I think you pushed coco coai to 500k R. >> Yeah, >> if I'm not wrong. >> And then you sold it for seven figures two years later. So you started in 2023, now we are in 2025. You sell it. My question is why? Why? Why did you want to sell it? >> Yeah. So we had we we had um many reasons. First one um we were only targeting the French market almost and we had reached out to everyone. uh and scaling to another country was not the
11:39 same. We had calls with German people, with uh Spanish people, the closing rate was not 50 to 60%, it was more 15 to 20. So, it's way harder to get a demo, way harder to close the demo. And in the end, uh if you scale, it's it's really hard. Second thing, uh we had two people on the team who wanted to to sell, who like did like the the project. Um and third reason I would say we had no vision about uh this project. So we knew how to make the first steps but then how to expand how to attack new markets that was really really hard and uh so that's like selling wasn't really an option but I sent a few messages. Uh one person
12:26 responded said that he was interested and two months later we were selling. So I mean if no one had responded I would probably still be working on on go. >> That's funny because usually when someone think about selling a business you go on a platform it takes a long time you sent few messages and you sold it in two months like how did you know who to sell the product to and how did it work actually? So we uh tried the acquire.com and uh flipper and all these platform but they didn't want to sell our product since it didn't have one year of revenue. They almost always ask for one year of revenue to do a multiple otherwise the multiple is going to be really low. So they were like giving us
13:14 1/4 or one/3 of what we sold it. Um so how did I know? Um I did two things. First I reached out to all the successful um Shopify app founders that had more than two apps because they were grow they are growing a portfolio of apps. So I contacted all these people saying okay we want to sell. We had a few due diligence but nothing happened. And then by chance I saw someone on on LinkedIn saying okay I'm buying a bees. Um I'm looking for a size between this and this in R. Um contact me please. I contacted him and uh that's it. I mean, maybe it's luck um but it it happened really really fast. >> Do you think you're lucky or do you think like everything that you did was
14:01 because you worked hard or because for another reason because I remember that you did a post and you started by I got lucky. >> Yeah, I think both actually. Work brought luck >> for sure. >> Yeah, I think so. um you sold it for seven figures if you can share it like how much did you get from it and did you have enough to actually take few years off later? Uh, so I can't tell how much uh because it's it's in the contract. What I took for myself is multiple hundred,000. Um, and I yeah, I could um I could Yeah, I I I I'm able not to work for a few years if I if I want to. Uh,
14:48 but that's not the way I am and I want to work. So that's not that's not an option. And we got 90% of the deal up front and 10% uh over three years. So we are still getting a few payments from from the sale. So you did that exactly. You built again but this time you sell you said sell before you code. So you decided to build again and you eat 7K MR before coding a single feature. How did you do that? >> Yeah. So one thing that made us successful in um in Kokoai was reaching out to high intent leads. So we we we found out that when we reach out to people on LinkedIn commenting post about WhatsApp marketing or email marketing
15:35 and when we would reach out to these people they would be more way more responsive just because they already knew what marketing was. They wanted to do uh volume for their e-commerce brand. So when we saw Coco AI, we looked at all the things that made us successful building this tool and selling it. And we saw that the high intent lead was something. So we built a tool around that to see if um like we wanted to see if people were interested by getting better leads and of course they are. So we reach out to a lot of people like I basically targeted the the whole world, the whole USA. I took many industries. I sent maybe I don't know 20,000 emails saying we can
16:21 get you high intent leads uh if you are interested reach out and so I also said um something like uh if it doesn't work out we will refund you or I don't remember the the the exact sentence. So we got a lot of calls a lot of people saying okay let's go. So, we were selling I think between 200 and $300 per month to send people 100 or 200 leads, high intent leads for their business. So, I finish the first call, people say, "Okay, I want to pay." They pay and now I'm like, "Okay, now I need to find them high intent leads." Um, so first I did it manually, then my coworker did it with me and then we hired people on Fiverr with exact blueprint on how to
17:08 find I intent leads. So basically if someone was had a dental clinic and was looking for nurse, we would type on LinkedIn dental clinic nurse. We would look at all the people interacting and then we would extract the right ones and reach them with email and phone number and send them the lead list. So people were really really happy about that because well the quality of the leads were was great. So that's how we sold before building and at one point we had like too many customers um and it was not manageable. So we we stopped almost everything fell to zero and then we built the platform and did this the automated way. Um so yeah and that's was a great lesson because having money before you code is
17:55 a great great sign. It means that really you you are into something. >> Yeah 100% and I think it's a pattern now that I see a lot of B2B founders starting by just selling the selling the service before actually building the product >> of course >> which is really smart. you started um now you're three co-ounder did you just started the service at two co-founder then brought um dev to build the product no we were already three um so also when we goji we had two first versions before doing this one first version we were doing a UGC um so we kept the name goji across the the the the project so goji berry V1 we
18:42 were doing AIGC's uh we did two weeks I think we went up to $500 of Mr. But um we had no interest in the product and we didn't know how to make it evolve. So we shut it down. Goji Berry V2 we did a note taker a bit like Fireflies AI. We took I think two weeks to build an MVP. Then we tried selling it and it was awful. It was really really hard to sell. So when we launched Goji Berry V3 our developer already made two MVPs. So we said okay don't do anything. We'll tell you when when we we can start. So the first month we were two selling the product and the developer was just uh waiting for us. >> Um how long did it take from the exit
19:28 that you did with Coco AI to building this product and getting your first customer? >> Um so you mean for the version one of Goji Berry? >> Um for like yeah the V3 of Goji Berry. >> Okay. Uh V3. Uh so we sold in February and we launched in September. So 6 months around six months. Yeah. >> What did you learn in these six months? >> Um I learned that the project I sold was a great project. Um so when I see the matrix today of Coco AI, um I think to myself, okay, what I sold was really a great project. And it's not easy to make
20:14 a great project. It's not easy to have a um a product where the churn is really low, where you have um great traction, customers are happy. So most of the founders I know that want to sell, they don't realize how hard it is to redo it again. It takes like a lot of energy. M >> so right now it takes way less energy for me to run goji berry than at the beginning when you you need to try you need to find an offer you need to find the first client you don't know if uh you are doing something good so yeah launching a pro a project is painful and um and and it takes time so goji berry went from zero to 33,000
21:00 a month no yeah 0 to 33,000 in 4 months you say no ax just relentless distribution so now I want us as it's your part of the job to deep dive into the distribution so first you talk about Reddit linkadin cold email um influencers newsletter YouTube trust mr you did a lot of things >> can you explain more and just deep dive into that by starting by Reddit Uh yeah, sure. Um so a current um advice like an advice a lot of people give is focus on one or two distribution channel. Um I'm not a big
21:46 fan of that especially in the beginning. So what I do is um my process is posting a lot on one channel, understanding how it works and then adding more and more channels every day. So over time the time I spend on on each channel decrease but I'm always adding new channels uh for tractions. So one of the main channel at the beginning was was Reddit. Um so we got almost 11 millions impressions on Reddit in in four months. It drove hundreds of like tens of thousands of people to to the website. Um and what's really interesting is it's it's free. So you
22:32 don't need um you don't need market marketing budget. You you just need chat GPT. You need your brain and you need to tell stories. And when you have this then you can get some traction. So we started by Reddit. Of course we were still doing outreach um as well since it's the core of of our product. And then we added X, we added um SEO, we added YouTube, we added trust, which is a website where you can buy ads. So we went like this. Once a channel works, we add another one and and another one. And like two days ago, we added threads. Um so that's the way we go. Um interesting. Just for Reddit, you said
23:18 you did three posts a week in 10 SAS subreddit. >> Yeah. >> Can you give me an example of what kind of post you did? >> Yeah, of course. Um, so the what we do in in Reddit is um always tell stories, always back the story with proof um and always include a link. So for example, we uh tried to get into Y combinator and we got rejected after the interview. Um so we made a Reddit post about what we did to apply, what they told us, why we got rejected, the proof that of the rejection, and a link to our SAS. And this kind of content, people love it because they understand, they
24:05 get value out of it that they can use right away. they um love uh hearing stories and also they are curious so they will go into your link and some of them will subscribe. So that's one post that worked well. Also we launched an um a marketing campaign with five influencers and we shared the whole playbook how to find the influencers what we asked them um what the results were the market marketing budget. So I think that sharing a lot of value like this um in interest people and so that's how we how we do it. Um so every time we try something on the business either it's customer service either it sales we make a Reddit post out of it sharing value
24:52 and that's how we that's how we grow. Also there is one hack on Reddit let's say that when you post something you need to get 10 up votes as fast as possible. So what we did is we have a group of friends where we all post on Reddit. Every time someone shares a new post, he shares it in in the in the WhatsApp group and so everyone will go and upload the post. Uh and this allows the post to go viral >> pretty much. >> You have um all these different stream uh as you say like um linking code email. I'm sure that there is probably few of them that you want to go deeper in and maybe the other one that are not so important. Which one for you were the most important in this um strategy? >> So the first one the most important is
25:39 always outreach for me. Um because it's something you can you know right away if it works or not. Uh, so that's really interesting. When you send 5,000 emails, if you don't get any answers, well, you know, your offer is not good or your messaging is not good. So, call out which is great because you can get a response right away. You can scale it very very fast. Um, and if you compare it to Reddit, it's hard to scale Reddit. When you can make three post, you cannot make 30. It's hard to scale YouTube. Same. uh you can make one video per day, but you cannot make 10 videos per day to to make it go faster. So yeah, outreach is definitely the number one. Um you need one LinkedIn account,
26:26 you need a few email inboxes and that's it. You can you can know very fast if uh your project has potential or not. Um >> what do you use just for the cold out reach? What tool do you use? >> So for LinkedIn, we use our own tool goji AI. Um, we have five LinkedIn accounts sending messages every day. And for the cold email, we use Instantly AI. Um, and so if you combine both, we send about uh 5,500 messages per day. LinkedIn plus um plus email, which is which is great, but at the beginning, you don't need that many. >> And did you find any act in cold
27:12 outreach? something that works, a template that you use often. >> Yeah. So, instead of asking for a demo, which is um I I think a bit intrusive for for for most people, they don't want to spend time with you. They don't know you. They haven't seen anything. What we what we do is we offer a blueprint. So we do an email exactly like if we were to ask for a demo, but at the end we say, "Okay, if you agree, we can send you our blueprint that show this, this, and this." And so I mean, saying yes to a blueprint is way easier than saying yes for a demo. So most people say yes. We send the blueprint. It's full of values. It's like 10 pages where the the reader will
27:58 learn a lot of things. At the end, there is a booking link if they want to book. But most of the people they don't book a call, they just go to the website and try with a free trial. So if you want to reach out to I mean if you want to put your offer in front of more people, the blueprint is way way better. >> Interesting. Um, another thing that I want to ask is you did newsletter sponsorship and I think this is something that is underrated because I I always thought for example if I want to build a tool I will subscribe to Justin W newsletter then I will find a problem that he's talking about then I will build a product to solve this problem and I will contact
28:44 Justin W straight away to get in his newsletter and to be sponsored because I feel like a newsletter that's already a curated list of people and you can target them straight away. Did you get a good result with that? So, we only tried French newsletters um by the way. So, on the influencer side and um on the newsletter side, we almost contacted only French influencers because it's way easier because we we know someone who knows someone. So we could get into these newsletters. The results are okay. So I mean we are profitable over two or three months but um we will we will do it again because it brings uh value but
29:31 that's not I mean that's not incredible. Um the the the hardest thing is um you you know what's tracked. You don't know how many people saw it and will buy it later. So it's really hard to know exactly how profitable you are. What we know is in three months we will get the money back that we spent. Um but if I compare it to other channels like for example influencers are way more profitable than newsletters for us >> and influencers is LinkedIn influencers. >> Yes. >> Um how do you find them? Um so mostly by typing keywords um or looking at um what's who who is
30:18 sponsored by our competitors and we re reach out and what's really amazing about LinkedIn influencers is um they give a lot more value than the price you pay because one or two people buying your product can um pay back what what you spent. So, LinkedIn influencers, they don't have a lot of deals. They need money like everyone else. So, it's easier to get um low prices and newsletters. I think a lot of people are asking all the newsletters to sponsor their uh their their SAS or their content. So, there is more competition. So, what what we found out is on LinkedIn influencers, you can easily cut the prices by two or ask for two times
31:03 more posts. Uh so that's why it's so profitable and you only need one post um like going viral to to get extremely profitable and you cannot get viral in your newsletter. So that's the main uh main difference >> and similar you did affiliates with your background did it help? >> Yes. So yeah it helped. Um the most important thing is when people sign up they get a kit of actions to do uh videos to repost text to paste. Uh so yeah the the main problem when you get affiliate is they don't know how to promote. So you have to really help them and give them steps on how to make um like how to make money with with with your SAS.
31:50 And also another thing is that often people are asking themselves how much should I give like commission do you give like 10 30 40% like do you have so we give 30% lifetime but we will raise it to 40% later on then 50% then even 60% if we can so the more you give to affiliates the more they promote um and I've seen it with some friends doing uh affiliate marketing Okay, when you increase the affiliate commission from 30 to 40, well, you make more than that because more people come in and start promoting and also it gives visibility for people that will buy without affiliate links. So yeah, definitely the
32:36 more you give the the the the better it is and lifetime so people they keep sending you traffic. Um, so I think yeah, it's really important to give as much as you can. >> And how do you find the good um affiliate marketer? >> So right now I don't uh I just have a link on my website and uh from time to time I push an email to my list. Um I'm trying to advertise the affiliate program on on on X. Um but right now I mean there is not much that is done on the affiliate part. We get about 10 to 15% of our um revenue from affiliates but I'm sure we can do way better but
33:24 it's not it's not it's not really my focus right now. I mean it's it's here it will increase. I need to do things um but but yeah there are many many many things to do before. >> So what is your focus now? So the focus is uh always increasing cold outreach, getting more traction on um X, getting more visibility on inbound LinkedIn um and I mean yeah the all the channels we have increase the KPI uh so we can increase the number of trials and then the number of customers um but you are right affiliate marketing is definitely it's in the list. Yeah. And the last one on this one, trust Mr. >> Yeah. >> You bought the ND spot and you got 50
34:10 200 visitor a day very profitable. You said um was that a bet or did you see other people talking about it before got result and then you buy one? Um, so I knew even if the ad would not make me money, uh, the visibility I would get on X talking about it would, uh, pay back. So, and if by chance Marlu would retweet me, then that would be jackpot. So, that happened, uh, I think two times and we got a lot of customers, a lot of leads. So yeah, they are I mean when growing on X um like saying uh okay, I use this product is
34:56 great when the founder has has a lot of followers is always a good idea because if they retweet you, if they respond to you, you will get traction. So yeah, we got more than uh what we paid just by one retweet and then on the ad itself um it's profitable. It's bringing um clients not every day but every two or three days which is really really good. And you know what, Marlu, if you look at uh if you watch this video, thanks because you know, Roman, I just saw your post on X and I invited you and I was like, I have to verify if this guy is actually good and then I went on uh trust MR and I saw your profile and I want to know more about it because you have total
35:42 revenue 4.8 million across all your startup. >> Mhm. >> What is it? Is that the revenue since you started Coco AI until now? No. So, Cocoai is not even in in that thing. Um, it's affiliate plus affiliate courses for myself and my co-founder Pot. Uh, so we made about 2.5 million each. Uh, so that's thing from the past, but we used trust to redirect to goji AI. Um but those businesses we share on customer they barely exist today like it's it's coming to an end. It's just um few payments from affiliate website that remain that still uh still arrive but um
36:30 yeah we used it just to to to do some marketing >> pretty smart but that's also a huge number. Yeah. So >> talking about huge number, I am watching this video. I want to build a B2B SAS. What will be your step-by-step process to get the further possible in the next 90 days? >> Yeah. So first step is um knowing of course what are your skills um what is your unfair unfair advantage. Maybe you can take what you did before or or not. Then once you have an idea you look at everyone on the market how they are doing marketing. Are they um
37:16 raising funds? Are they bootstrapping? You have to know of course what what path you want to to take yourself. Then once you know that you open PowerPoint, you write seven uh slides about your offer. So you can do something like um your logo, then the pain points you want to solve, then the solution, then how your product works, then the price. Once you have this and you you need a really basic one. Uh you send messages to let's say 100 500 people. You get them on a call. You ask them about their problems and then you take the point and you show how you want to solve that problem because of course even if you
38:02 don't know how to code, you still know how you will solve the problem. So you take a lot of feedback from people. Um you understand what the pain point are. Usually they are different from what you what you wrote on your on your PowerPoint. So you adapt your point every time and once you get someone who wants to pay either you can deliver it as a as a service or if you can't you just say okay I will build this I I I'll be back and then you find a solution to build the product. Either you bring in a co-founder or uh you pay for a developer. So you need someone to to to develop the platform. Once you do this, you do it with let's say five to 10 people. You validate the idea and then you decide if you want to go all in on that project or not.
38:48 If you want to go all in, then it's time to do marketing. So start with cold outreach. You take any autom LinkedIn automation tool. You take any emailing tool. You need about let's say $300 to $500 to to run it. Uh both. And so you start sending email, reaching out, try to get 10 more demos. Once you see that cold outreach works and you can put it almost on autopilot, you will start doing inbound. So you post on LinkedIn, you post on X, you post on Reddit. Um on Reddit, you make a post about how you are trying to solve this problem and ask people for review for feedback. And so that's how I would start actually. Um and then try to
39:34 grow every platform with uh what you do. So of course it really depends which target um like which audience you are looking for. For me it's easier. I'm targeting people want to do sales. If you are targeting dental clinics um well of course you will not share um content on on LinkedIn or not the same way I guess. So you always have to take that into um into account. Um and then I would just go like this. Um first month focusing on outreach. Second month building outreach plus uh inbound and third month adding uh some source of traffic. Uh maybe some paid ones, maybe some influencers. It really depends how much money you made if you had money before because it's quite easier for me
40:20 to to to talk about this because I just exited a project. So, I didn't have any problems um paying influencers, paying upfront marketing. When you have zero and you start, it's way harder, but you can still make it. >> Amazing. So, I want to go back at the beginning. What is for you the best way to find a good idea? Um that's a that's a great question. Um so usually what works best is um a problem you solve for yourself or or for others um that you can make into a SAS um or talk to people a certain category of
41:06 people and try to find uh something that would make their life easier. Um but everything starts with the customer not with the product. That's the the main thing. Uh so so yeah making a great product and finding a great idea is uh listening to the right people and uh are they how how much do they want to solve this problem and how much are they willing to pay? Um and you cannot guess I mean we got lucky by copying something. So I mean you can copy something that works um but you have to make sure that like they are solving a real problem. They are solving the they are targeting the right audience and and so on. Um but taking a bet on copying
41:52 another company without knowing what their unfair advantage is. Maybe they are lying on the on the numbers. So it's really hard. So you really need to know which problem you are solving. Are people willing to pay for it? Then you look at who is solving this problem. How can you do better? which distribution channel they use and try to copy what you can copy. >> Perfect. And another question uh from this is you said you do code outreach until you um you can delegate it or you can automate it. >> How do you do that? So once you get a great response rates on LinkedIn and on email, your only job is to add more leads and you can keep the same campaign until it dies and then you relaunch
42:38 another one. So for example, the campaigns I'm I'm running on Goji Berry, they've been running for the past maybe two months, something like this, and I haven't changed the word. The only thing I need to do is find the leads, put them in the campaign. I do it once a week. And my only job with outreach today is responding to to people which doesn't take a lot of time. >> Interesting. Where is Goji Berry going? What is your goal for 2026? >> So actually I I don't have I don't really know. Um we want to grow of course as much as possible, keep the team as lean as possible as well. we don't plan to hire uh soon. So
43:25 what we want to do is create one of the best tools out there in terms of GTM copilot. Um so a lot of LinkedIn tools, outreach tools. So we we we we get compared a lot to LinkedIn outreach tools and we are a bit different than that. So what we want to do is either um so instead of of being a tool that sends messages, a tool that tells you who to contact, when to contact them, and how to contact them. So think about um you are targeting sales people. We will only take the best sales people. We know they will respond to you if they have a pain point to get the most amount of responses. And we will also tell you when to um send another message,
44:10 how to send a message and so on. So yeah, we we are thinking in terms of AI copyot instead of um LinkedIn outreach. So right now we have about 500 customers. Um and I mean it's going really really well. >> So when you did Coco Aai you sold it because you said you had no vision for the tool and now when I ask you this question I feel like you're really focused on the present but not really the future. Do you think that's one of the key to actually get a successful product to just focus on the present and be focused on your customer and what they want? So we we have both. Um we know that everything we do um has to be aligned with the customer getting more demos, more replies when doing outreach.
44:58 Uh and so everything we build is um is is focused on that. So you are right, we are focused on the present but also on the on on the long-term vision. We want to add more marketing channels. We want to add more AI features. So when I was running Kokoi, I couldn't tell what I wanted to do two features from what we were doing. Um on goji berry, it's way different because I've been doing outreach for the past five years. I know exactly what I want, how I want to do it, and I'm the core user of my product. I know I'm a customer of my own product. >> Um so yeah, I guess I guess that that's why it works as well. We we exactly know how to
45:43 develop what we want to develop. >> Another question mistakes. What are the be the biggest mistake that you wish you have avoided? >> Um for um Kokoi for example or for um >> everything since 2023. >> Okay. Uh so yeah, Goji of course building something customers didn't want. Uh um sorry Cocoai building something customers didn't want for God. Um maybe we went too fast on on on some marketing operations. We paid influencers that were fake influences. So we lost a bit of money there. But I mean it's it's cost of doing business. It always happen. Um so right now on good delivery I I won't
46:30 say we made any mistakes. So I'm quite happy with that. But on on on Coco AI, sure we we made tons of uh tons tons of mistakes mostly with the product. >> Okay. And what is the thing that move the needle like the 20% that do 80% of the job? >> Yeah. So it's definitely doing more doing doing more of what works. So when you see that when sending 10 messages you get one demo then send 100, send 1,000 and do as much as you can. Um people I think underestimate the the energy you need to build a successful SAS. It's getting harder and harder. You have people that are highly
47:16 motivated from all around the world. So you are competing competing um against a lot of people. Some of them have capital. Some of them have marketing skills you don't have. So you will need to be active um and and work a lot. I mean >> well Roman thank you for >> thank you so much >> your time. That was amazing and I'm sure this one will help a lot of people mainly B2B founder. >> Great. Thank you so much.
