Life Math Podcast

#8 Startup Status Update: 500 Days of 3veta

Iskren Vankov and Iliya Valchanov Episode 8

[Special Episode]

More than 1 year ago we started working together on our startup project 3veta.com.

This is a status update of sorts where we look back on the good and bad decisions that we have made in this period.

05-2020: Came up with the idea for 3veta.

06-2020: Started assembling the team.

07-2020: We set up a legal entity.

09-2020: Software development began.

01-2021: MVP (minimum viable product) was out together with a closed beta.

05-2021: 3veta entered open beta.

06-2021: Officially launched on Product Hunt.

09-2021: We had a major breakthrough in terms of users.

10-2021: We recorded this episode - our first status update.

***

Get on the email list and never miss an episode. 

Discuss this episode on Reddit.

Connect with us directly on LinkedIn: Iskren Vankov and Iliya Valchanov.

3veta
3veta.com is our startup. One platform to unify your client bookings, video calls, payments & more!

Disclaimer: This post contains affiliate links. If you make a purchase, I may receive a commission at no extra cost to you.

Iskren:

Okay. Well, I think it's time to say it. Should I say it? Should I say it?

Iliya:

Say it.

Iskren:

Hello. Hello.

Iliya:

Oh my God.

Intro Music:

Life math. Life Math. A podcast indescribably tangled, unnecessarily complex. So bad, that it's good. Life math.

Iskren:

So, today we have a special. And, today is the start-up special. What I mean by this is that after having released a number of episodes, we've had a lot of feedback from people. A lot of it came back to us with other questions or just general feeling of how they perceived our podcastwhat they liked, what they didnt like. And, we noticed that many of you were particularly interested in

the startups episode that we did:

The Ontology of Startup Anecdotes. And so we decided to do a special, which is going to be about the past one year [or one year and a bit] which is the exact time span in which we have started and worked on 3veta -our startup. So this is a sort of overview of looking back the first one year of a startup, because many people give you general tips on, or if you want to have a start up: do this, do that, don't do a third thing. But, they're kind of general. So, I just want to try this format where we talk about our firsthand experience as it's happening one year into this business. Where it is, where does it stand, where's it going? How was much of our expectations? What have we learned? What have we didnt? What have we done? Well, what have we done poorly? This kind of stuff.

Iliya:

Im very interested to learn more. Should I subscribe? Is this the point at which I subscribe, leave my email, and follow you everywhere?

Iskren:

Yes. Also give us money.

Iliya:

And tell a friend about the podcast. Well, that was a great episode.

Iskren:

Essentially, this whole thing is one large ad for 3veta. You have to bear through with it.

Iliya:

Actually, to be honest, I asked Iskren why do you want to do this episode? He's like, oh man, everybody's asking me what's going on with 3veta. Let's just simply record it so I can send it to people.

Iskren:

Basically. I just don't want to talk to people. I want to talk into a microphone alone in the room. Perfect communication, purely one way. No feedback received.

Iliya:

Since we haven't plugged 3veta just yet in this podcast, maybe some listeners don't know what our product is.

Iskren:

Alright. Do you want to give the generic sentence?

Iliya:

Yeah. So the platform that we have built and are currently marketing and scaling is called 3veta.com, three with the number. And it's a platform which unifies your client bookings, video calls, workshops, payments, and much more. With 3veta, you can build your own booking page, your own website, completely white label, and have your video calls in the meeting room with your own name and logo. It's absolutely incredible for some people. This whole episode is going to be about our experience at 3veta.

Iskren:

Basically where we stand now is the following. One year later, we have developed a product. It is online by now, enough time has passed that it's not just the first rough version and prototype. No, it's had so many iterations. So many people have looked at it and used it that it's quite polished by now. Also it has quite a few different functionalities and features that it's starting to get like a fully fledged platform with kind of many legs you can use for different things. And then after several different promotional campaigns, finally, we kind of found this one breakthrough, where with this promotional campaign, which was international, essentially we managed to get several hundred paying users to use our system as providers. And this was essentially our big breakthrough, where after more than a year of struggling to find market fit and users, so that now we have actually hundreds of people who do use the platform which couldnt make us happier, obviously. And that's kind of where we are now. So we have found a way to find several hundred people to actively use us, but not yet how to scale it even more. Thats just for the future, right. So that's where we are now compared to one year ago. And have in mind throughout this one year; the first year, for the vast majority of time, we didn't really have users. So that was scary throughout most of the time. And we kept trying to think, okay, what are we doing wrong? Why is this still happening? We kept trying different options. And eventually something succeeds. That was a pretty mentally stressful time as well. You know, are we in the right direction? We dont know. There's no road signs to let us know. So it's brilliant for us that we have actual feedback now. Um, but yeah, one year ago we thought that this feedback would appear. I remember the very first roadmap. We thought that in the first three months, we'll start receiving actual user feedback. In fact, it took a year and three months to get [not some], but like a lot of feedback, right? We're getting some feedback before, but it was a bit patchy, whereas now we have abundant feedback from regular users -more than one year in. Let's see. The first kind of topic I want to look into is: where we stand now compared to where we thought we would be one year in the future when we started the company. So Iliya try to mentally transport yourself back in time one year, and just remember where your mental state was, imagine how you perceived the one-year future of 3veta when we founded the company?

Iliya:

We had this amazing timeline that we created, and we're absolutely off the timeline and we need a new timeline with new highlights, new milestones because it's practically a different beast. We had this timeline that we expected our product to have some MVP -ready for people to use- maybe in January or February of 2021. So we expected that we will start selling the product January-February, and then maybe by March-April, we will have some feedback. We will start implementing it. And then by June-July, we will have some refined product. What actually happened was that we absolutely over-complicated and over killed the whole product. And we were struggling with finishing the product until maybe June. And then in June, we went live with the product. We made the product hunt lounge and so on and so forth. From this point on, we started selling it. Right now, it is the 14th of October when we're recording this. Basically it took us what 3-4 months to get a couple of hundred users from the actual MVP. The problem there was that our definition of an MVP [minimum viable product] was very, very overblown.

Iskren:

Very broad. So, here's definitely we've discussed internally, but it's definitely something worth mentioning in this episode that we completely over-engineered the product. We made it so much more than it could have been for a minimal product. And obviously we regarded this as a mistake, not like a mistake to beat up ourselves retroactively, but something to learn from. Then I do think we have all learned our lesson to some extent in the sense that we think more carefully before over killing solutions, we all do try to find a shortcut to the solution -rather than try to build the most complex structure ever. However, I still think that all of us also still fall in this trap sometimes, right? So basically we're better, but it's not something you can really completely uproot. It just keeps happening and you have to keep fighting against this over-engineering syndrome everyday on each step of everything you do. Like, I was surprised how much on basically every decision, it doesn't matter how big or small the decision, there's always a way to overcomplicate it. And when those things start compounding, it really starts hurting the overall progress. I guess the first really big lesson learned from this is just, don't do complex things. There's no time. Time just flies. And if you do unnecessarily complex stuff, you don't even know if anybody will ever use it. Like we spent so long on certain features that in the end, it didn't even see the light of day. So point number one, do the simple immediate solution. Sooner is better. So another thing I wanted to give an honor and mention, too, is when we were starting the product we defined those four pillars of the functionality that we want to chase.

And they were:

video, payments, website builder, and scheduling. We decided, okay, those are the four core ideas that we work on. And then, we need to also have the connections between all of them. So, you schedule for meeting, you can pay for it, and you can book it through a booking page, with a website builder. So, they're all interrelated. But in essence, those are the four pillars of the product.

Iliya:

Okay. I have an idea. Each one of us picks two and we have to say, what did we do wrong?

Iskren:

Okay, sure. So basically this fits into what I was like going into. So based on this, let me just finish the set up. So, because there's four -and each one has many mini- individual modules inside of itself, it can have many different sides, we had to prioritize. We start building from scratch. So we have two to choose. And out of those four, we're like: Oh guys, we're absolutely certain that we should definitely start with video and payments. So in the very first iteration, what we called the alpha version of the product which got released to a limited number of people for testing, we just had payments in video.

What this misses:

it had a platform where people can sign up. They can send links, which when their clients receive, they get to pay a fixed amount to the provider has declared and then join a video meeting. So it was literally a paid video meeting, right? We're very happy with the result we thought, okay, we've crossed two out of the four pillars. Now, where does the issue come? And I see this now in retrospect, that we should have probably picked differently, is that regarding the payments, they were the very first thing we're excited about. Yes, payments. In fact, now more than a year into the product and having hundreds of people actively using 3veta in various ways to solve their very real-world problems, the reality is that the payments are the most underutilized aspect of the platform. People want to talk, people want to schedule their talks, people want to have a booking page. People want to do all of those things. People want to send files left and right, but people don't actually -turns out- want to pay each other for those calls. Most of the requests in this direction are how to sidestep this complication and have

like:

Oh, he pays me somewhere externally or it was just free meetings here. So basically now that we have the whole platform and enough user data to know what's actually being used, it turns out that we started building it from the least used component. I mean, obviously we couldn't have known at the time and that's just what we assumed would happen its usage. But I do feel a bit silly now that we started from the least used aspect. However, on the upside, the other module we started with is video, which is still the most heavily utilized side of the platform. So I guess we picked the best and the worst.

Iliya:

I remember we were making you like police. Please have three meetings. And you're like, No, this is a paid video-meeting platform. It was very, very, it's a very good use case. But now with all the information that we have about the industry, usually video + payments, they are these calls that are paid by the minute. So this is the use case and we were definitely not positioning ourselves in this realm of pay-by-the-minute.

Iskren:

Yeah, I remembered.

Iliya:

About the payments, I just wanted to add: what people want is to pay in bundles. So they want to pay several meetings upfront, or they want to send a payment link. And this is something we learned the hard way.

Iskren:

Yeah. I remember being against spending time on making it a free video platform

because I was afraid of:

Why would they not choose literally anything else, if it's just free, et cetera, et cetera. And then it turns out that I was very wrong and why was I wrong? Well, because people enjoyed the white labelness, meaning the fact that they can brand their 3veta platform with their own branding -unlike with Zoom or others. But what they liked primarily is the fact that they can white label their video calls and most of all schedule them. So scheduling turned out to be the absolute key problem, which actually we left for last. So seeing this, since we split the four pillars of the product from between ourselves, you can say a couple words about the scheduling. Scheduling we left this component for last as Iskren mentioned. And the reason to do this is that scheduling is very hard from a tech perspective. And we didn't want to deal with time zones. We didn't want to deal with the whole bunch of problems that arise from these automatic email reminders all of these things. Now, there are a bunch of stuff that you have to think about. We were like, okay, let's build everything else. And once we have the whole structure, we can add scheduling and it's going to fit perfectly in everything else. The problem was the biggest problem for me as a sales and marketing person was that we had to sell the platform. And, even in March or April, when we already had video payments and website, we didn't have scheduling. And I was this person trying to sell the product to random people. But the value proposition wasn't there. We were always like scheduling is coming. You will be able to schedule, you will be able to have a booking page. People will be able to book you seamlessly, but they couldn't see it. How do you demo such a product? You write to the people and then you start scheduling this complicated meeting with back-and-forth emails or messages like: let's have a call at 4:00 PM, CT or let's have a meeting at that point. There are no reminders. There are no calendar events. There are no confirmations. So instead of using our product to schedule the meeting and have everything perfect, we didn't have scheduling. So setting this product was absolutely impossible. So with scheduling, two big mistakes. First, we did not realize that this is the most sought after feature. We didn't realize that the video plus the scheduling is the biggest value proposition. And second of all, this was the biggest problem for our sales efforts. We couldn't start selling the product before we had scheduling. It just didn't work. So, yeah. And, of course I could always schedule a meeting with some other software like this, but yeah, come on. I'd rather send 10 emails, than use our competitors.

Iskren:

Looking back and speaking about this, our first choice between the four core modules, which ones to build first and also for the audience, have in mind that this choice means that were committed for months of development, right? So we choose video and payments. And for months in a row, we're busy implementing this. So we can't just add some scheduling. With scheduling, it meant months of development.

Iliya:

Iskren, now, please tell the listeners how many developers were working on the project.

Iskren:

Yeah, so we had myself and then five extra developers, I believe. I mean, you know, through time it kind of changed. There were sometimes more, sometimes less. I think at the peak, it was actually plus six and then plus a QA person on the side as well. So the definite for a startup of significant size and still these things just take time, like development is just disproportionately expensive compared to virtually everything else for a software company. So all of the other expenses for not just human resources, but also things like marketing, salesOkay, sure, they're expensive. But development is absolutely through the roof expensive if you go overboard with designing the platform. And what, all of us I'm sure wish they had avoided in some way, but I don't know how we could have actually really avoided, is the fact that in the end, when we draw the bottom line, I'm sure that we utilize only maybe about half of the development effort that we put in and paid for and got delayed for. So basically our efficiency when choosing[or to develop or how to develop it] I would say it's been about a half. What I mean is that there's many pieces of software that we wrote. And then for some reason removed, overhauled, had to change something like this, which it's part of the normal process. Nobody gets anything right the first time, especially something as complex, but it definitely happens.

Iliya:

Something I want to just add. A very, very good part of our dev strategy was that we had is Cranium House is the CTO. And we were working with this company Camplight a special shout out to Camplight and,

Iskren:

highly recommended.

Iliya:

Yeah. So they are co-operative I believe that's how they call themselves and they are Okay, so I'm going to absolutely cripple the way they present themselves. But let me try. They're a bunch of dev people, and one or two business people. The business people find work for developers and the developers are world-class developers who hop from project to project. And when we decided to work with them, what we wanted was to have the strategy in-house to have our CTO. So the Chief Technical Officer in-house, but use the manpower of highly skilled developers who we don't need to hire because we can always upscale and downscale. So whenever we need more developers, Camplight can help us with that. And if we, for some reason decided to stop developing the product, we can also reduce the number of developers quite fast. And it's literally overnight. We can get like five guys working on our project. Five more guys, lets say. So thank you Camplight for being a part of 3vetas journey.

Iskren:

Yeah, so Kim Plight was a very good team because we could scale up and down very easily as the demand for software was increasingly decreasing. We also did have Mimi, one person in-house. So that was very helpful because she was with us for many months and she built so many core pieces.

Iliya:

Now she's a data science consultant. If you need data science help, we can refer you.

Iskren:

Exactly drop us an email. We have the best data science consultant. So I wanted to ask you having discussed this issue now of the initial decision to go so hard after payments and make them an indispensable part of the system, just to realize that they're the least desired feature, at least as of now. How do you think we should have approached this? Do you think we could have done anything differently at the time having the information we had or just, No, that's just the process. It's trial and error. Because obviously it's very easy to say: Oh, we should have done this other thing, but how could we have known? So what maybe kind of little experiments could we have had done at the time? How could we have received this crucial market feedback that we now have received, but before having the product?

Iliya:

Well, what we did in the end in terms of payments is that we integrated this very [not very], but relatively complex system Stripe connect, which is the best solution in the world when it comes to building a platform where people can create accounts and then other people can pay them. So this was the best in the world. And, what we could have done was prototype this and basically have a simple payment link, maybe use PayPal or something just to see: Do people actually use payments? And not integrate this whole complex system into the core of 3veta. And I say in the core, because whenever we talk about changing payments, adding other payment providers, it is very, very much related with building part of the infrastructure or changing this Stripe connecting. And yeah, we could have prototyped this one. We were building to scale, you know, that was our motto from day one. And if we had 10,000 people using the platform in the first three months, well, Stripe connect is the best way to do it. So, in retrospect could have tried it out with a PayPal link or a simple payment link.

Iskren:

I remember I, I had this one kind of crazy preposterous idea that got shut down quickly. But, in retrospect, I think, it would have been good actually.

Basically I proposed:

let's not build it at all. Not even like a simple solution. Let's not have a payment solution. Just make it into a virtual kind of gimmick, kind of have a payment page where people can pay with just card details, a simple form, but then there's nothing connected to it. It's just a simple front-end visual form. And then when people tried to pay, so when somebody chose to pay for a consultation, let's say $50. Because there's no actual payment, we take it out of our own pocket and pay the provider. So in front of their eyes, everything works fine kind of magically, we take the hit financially, but receive the confirmation people use the payment, right? So we could have committed some budget, maybe like a thousand or something to cover those expenses until we see that people do use the payments. Let's create them. And so it sounds crazy, too, this kind of stuff, but right now we have spent more on development costs to make it happen than we would have spent on covering those fees out of our own pocket if we had never created payments.

Iliya:

We had to prototype this, but again, you know, this is this avalanche of wrong steps. You take the wrong step to do the video and payments first, and then you leave scheduling for last. This cripples our sales, at the same time, you can't prototype payments anymore because you've decided to build payment, then its literally the first thing to build. And then every other process builds on top of this payment system and, yeah, so overall now we're stuck with being amazing for payments.

Iskren:

Yeah. I mean, okay. So we kind of sound like a downers for now, like, Oh, we did this wrong, we did that wrong. However, let's make it clear. Like we're one year and a bit into the business and with the recent events and developments, we are actually very happy because now we have hundreds of active users. We have tons of very useful, real user feedback. Uh, so right now Id say that it's actually going good, right? Slower than we expected initially, but we're just being unreasonable. So yeah, let's not be such downers. And for example, regarding those payments, yeah, we did spend the development effort on it and it did cost us time when it was crucial to be quick to get out to market, okay, we've taken this hit. However, now that we are out, now to have real users, its just an upside that we have it.

Iliya:

Absolutely very important to say this: if things didn't go well for us, that could have been a very big mistake. However, now that things are going well, it is just a cherry on top.

Iskren:

So I want to..

Iliya:

By the way, before you go, I just want to check how many people have onboarded to Stripe. Sorry, I just, yeah. About 10% of all the people on 3veta platform are set up to receive payments. We could have left it for later, but guys, whoever is building a start-up, it is a great retention technique. Once they start receiving payments through this platform, there is no going away, you know, this, one of the things which logs them to the system, if it works properly. If everything is nice and they receive the payments and everything, they never want to change. So yeah, it's a great retention

Iskren:

So discussing those features and what weve done, I was going to ask you, what do you think is the single most preposterous thing we've done? Kind of looking back, what is the single thing that we over killed ridiculously and needlessly, like the one thing you just have an ax to grind with?

Iliya:

Yes. Cookie consent, too. This is the absolute worst thing ever. We built this extremely complex system, which was governing the whole Cookies & Tool in different languages, on different sub domains, all these things. I don't even know how we started doing this and now I'm just looking at it,

and every time I look at it, I'm like:

Uh, it looks so It looks very nice. Works very nice. But come on, come on. Like, we've all seen cookie consent tools. It's absolute overkill, because I remember when we were building it, you know, it was just before Christmas. It was me, you and Yoko. Yoko is one of the most senior guys we have worked with from Camplight and a great developer. So it was me, you and Yoko. It was 23rd of December and it was maybe 11:00 PM, 12:00 PM. We were all in this conference call, building the cookie consent, too. And we had this huge campaign coming up the next day. It was this. I don't even know how to explain it in English, but here in Bulgaria, we have this Banitsa thing, which is a type of pastry which, where we put fortunes kind of like fortune cookies.

Iskren:

Let's say it's like our festive Christmas version of fortune cookies. So kind of made a little interactive game with it where people can draw their fortune cookie for next year, for the new year, just to emphasize the point that, Oh, you can do many things online, including draw fortune cookies with your family if you're stuck somewhere else away from them because of a lockdown and quarantine and stuff like this.

Iliya:

It was a cool idea. It should have been a one-two day project. It was in the end. By the way, it got us 10,000 hits on the website, just in like two days. It was incredible. We were on New Year's Eve. It was like thousands of people were all on our website, like playing the game the mini game. It was very cool. And, so it worked kind of, we got indexed by Google, which was very important. Indexed by Google, we got our domain authority a bit. We started ranking for different things because of this inflow of people. But, we are preparing for this. And how do we prepare? We built the cookie consent, too, because we can't have people's data without them consenting to it. And it's me, you and Yoko building this and I'm kind of coding something. Okay, so Yoko had to do two things. He had to complete the cookie consent, and he had to, I think, work on the structured data. So structured data when people share links from 3veta, they needed to display properly, display the right images, see the right words and not something generic and so on. You gave me this task guys, because it was the easiest task for a non-programmer. And, I was just writing some code and in the end I did it. So it was great. It worked. And I remember asking you like, Okay, amazing. I just have one question, what language do they code in? Because I can code in Python and I could code this, but I had no clue what it is. And then you explained to me that this is TypeScript the new language of the future. So, yeah this was me, Yoko in the meantime, building just cookie consent to writingChucka, Chucka, Chucka, Chucka clicky, clicky, Were speaking to each other, all our wives and girlfriends were absolutely pissed. I remember Masha was pissed. I remember Yoko's wife was pissed. Probably your girlfriend, future wife, was pissed, as well.

Iskren:

I mean, it was the first Christmas I spent with her and her family at their house. And, you know, I was gone the whole evening of the 23rd, around lunchtime on the 24th I had to go as well because we realized there was some issue.

Iliya:

Yeah. Overkill. Cookie consent, too, the most hated feature in my mind. But, how about yours?

Iskren:

Hmm. Okay. So that's athat's a, yeah, you have a very, very strong point that we over killed this. Very interesting, very strong contestant. Mine? I think the one that I just can't let go mentally, you just keep getting annoyed at past events the one for me is the so-called facelift of the video tool.

Iliya:

Going back to the four features, Iso, your turn for the video. You can rant about it right now. So, what did we do wrong with the video?

Iskren:

Right. So, you know, imagine we've been building the platform for several months now. We have something working. We have absolutely nailed payments, which makes no difference to our customers because they don't want payments, but we've nailed them. And so we also have video. But, essentially some background, the video solution as with virtually everything you build, you don't start from scratch. There are those available tools and projects that you start building from. And so we got this video solution and we started building on top of it. And the thing is that very, very, very quickly we reached something that worked fine. It just works fine.

Iliya:

I think you built you spun it up in a day I think.

Iskren:

Actually I remember this day. I was alone in London. Julia had gone somewhere and it was a Saturday. And anyway, I remember like it was one of those days when I had no plans and I was like, okay, I'm just going to sit down and try to smash this. And I did. I didn't really know much about this topic in particular at the time. So we had a working video solution in one day and it was fine. It worked fine. It looked decent. And it was okay. And so at this point, you guys are thinking, okay, you have payments, you have video. Nice, go on. Just move on to scheduling, go to market. No, instead what we did is the absolute overkill of overkills, for me personally,

which we decided:

Ah, people are not going to use this generic looking video. We need to facelift it to make it nicer visually, but obviously being nice visually is a very subjective thing. So maybe for some people, even the generic template would have looked better. So then we ended up spending, actually literally, weeks and weeks of development effort on face lifting the video solution. You know, rounding the edges, changing some colors, changing some buttons. You somehow like oversee it fully, but we somehow already fall into it and you know, once you're midway, you don't want to quit. You just want to kind of finish it, but then more issues just keep arising. And we walked on this side path that nobody wanted to be on for weeks in a row. And then eventually, yeah, we face lifted the whole video. It was completely ours, with our vision, and we're like, okay, those were a lot, but we finished it. And then like a month later, for other reasons, we were forced to actually check all of this work and move to another solution, which has the generic screen anyway.

Iliya:

How many lines of code did you delete that day from the code base?

Iskren:

Many.

Iliya:

I think it was something like 50,000?

Iskren:

I don't remember exactly, but like it's essentially weeks worth of work and investment both time and money wise. And I cant let it go that what we absolutely should have done is use the generic solution the way it looks which is a decent look even if its not our colors and stuff like this, and then move on to scheduling and things. So that's my absolute biggest ax to grind with our process at the time. But, I feel free, too.

Iliya:

Well, I feel a bit detached from this one because basically the data effort was with the dev team and the facelift was with the design team. So it's one of those tasks I was getting updated on, but I was never a very big part of it. So I don't hold this great grudge towards it. It pisses me off as well. I really hate it. I don't want to think about it too much, but yeah, I was not so involved. By the way there is a business case for face lifting. There are other software what they're doing is they're providing some kind of video API with their face lift on it and people fall for it. You know, people are like, Ooh, you have the most beautiful video. And I think this was the thing we were going after. You know, we were like, Oh, we want to have the most beautiful video. Now, I don't want to say it [sorry, design team], but it was not the most beautiful video, anyways so.

Iskren:

Like there's a thing that if you ask a hundred people, whether they prefer the generic visuals of this video solution or our face lifted version, I imagine it will be like 50-50, because they're just different. I don't have a preference. They're both fine in some way. Like, one is not necessarily better[objectively better] than the other.

Iliya:

Yeah. Should I go to the last feature? So we already spoke about video. We spoke about payments. We spoke about scheduling. There is one more, the website builder. Should we get into it?

Iskren:

Yeah. It's your turn.

Iliya:

Oh, man!

Iskren:

What did you fuck up there? Uh, by the way, a side note, because of the previous episode you called me out for swearing, because apple podcasts flag it and you have to bleep it and I just choose the F-word. I read somewhere, I'm not sure if it's true, but for movies [R rated movies] actually get up to three free F words before they get flagged as R rated.

Iliya:

Fuck me. I didn't know.

Iskren:

So sounds like we can say it one more time. Uh, let's keep it for a good moment. Is it going to be you or it's going to be me, who is going to say it? It could be another word as well.

Iliya:

Anyways, website builder. So in the first talks about 3veta, what were the three things? It was video, payments, and website. By the way, I just want to mention this. We didn't even talk about scheduling. Later on, we realized that without scheduling, none of the others can work, because what is common between payments and meetings? Well, you need to schedule it in some way. You can do it without it. We did it without it, but it doesn't make much sense.

Iskren:

Somehow

Iliya:

And we're like, okay, it's going to be this super simple website where people can basically put a photo of themselves, they can put some information and later on, they'll be able to schedule meetings, as well, over there. So, our one pager website so we built this whole website builder, but you have to realize we still don't have scheduling right, at this point in time. It is happening in December and we were still like, okay, we're going to leave scheduling for later. So what is this website, exactly? It is a photo, some text, with URLs, it didn't support them. So basically a URL which didnt work as a URL and a contact form, which didn't work for some time as well, we realized sometime ago. But anyways, we build this websites which are absolutely incredible immaculate website, or it is the most simple to use website builder I have ever seen. And I've used many, many different website builders. Everybody loves it. Whoever touches it, loves it. Everybody loves the website builder. Without the scheduling though, why do you need this website builder? So that was very, very premature on our end. Again, the main issue with it, however, is that once we created the booking capabilities in the scheduling, we realized that many people actually want just a nice booking page and they don't need the fully fledged website. So that was another pitfall that we fell in. And, um, yeah, so if we had done the scheduling before website builder may have never come to life, it would be just very simplistic booking page with a very simple payment link, which maybe is not even related to any system and to the video solution which works. It may be we would have been ready six months before. However,... So that was the problem in the website dealer. However, now we have both a booking page and the website builder. Some people are choosing to have just the booking page. Other people are building fully fledged websites with 3veta, and I have to say, they look amazing some of them. I even have my own website. I posted everywhere because I really like it.

Iskren:

Yeah, to be fair with this one, I was, I guess, putting the website initially, because maybe being a more technical person for me building a simple one-pagers, like who would pay for this, who would actually use it with us? Why us, there's other website builders that are better. Um, so I always thought of it as a glorified booking page. And in the end it kind of became this The scheduling should have been earlier. Alright, anyway, so we've discussed kind of feature wise. We saw that we just ordered them wrong and scheduling should have been much earlier, but we left it for later, which made us go to market a bit delayed compared to expectations, but it's okay because we have people now. And they use all parts of the system in one way or another, including some even pay using the payment module even in Malaysian ringgit.

Iliya:

We're processing payments in Malaysian ringgit.

Iskren:

Which is pretty mind-blowing. Imagine the capability that in several months we've got a system which helps Malaysian people pay to each other in Malaysian ringgit when neither one of us has ever been in Malaysia.

Iliya:

I know a guy from Malaysia.

Iskren:

Ah, yes. Thats the connection.

Iliya:

Actually, I just want a bit of credit to myself. This is one of the things I committed to the one of the odd things that I did.

Iskren:

The Malaysian ringgit, thats because you know a guy from Malaysia, right. That's

Iliya:

I was the most experienced one and it fell on me.

Iskren:

But again, the conversation, and I'm just thinking of locality as a cost. So we try to do the things that are closest to us, but then this has completely shifted in the online era. So, yeah, country, country boundaries just don't seem to matter one bit for us including So we thought they did because, okay, going back again to this payment issue, we thought we wanted to target European Union and United States exclusively initially, because those are the countries that, at the time, were fully supported by Stripe for payments. Our payment processor is Stripe. And so we are

Iliya:

We are a verified Stripe partner.

Iskren:

We are a verified strike partner because we built payments. Now, why do we do payments, though? I don't know.

Iliya:

So if something, if there is one reason why this is worth it its because when you go to Stripe.com, you can find 3veta as a verified Stripe partner.

Iskren:

Which actually does help in ways, right? Because Google thing is high level now, et cetera, et cetera. So it does have its upside.

Iliya:

Yeah, so something, I want to add this for content marketing and SEO. Right now, we are on the very exciting route in terms of SEO. We have worked for maybe almost a year now. So we started writing the first articles on the book in October of last year. And it took us a very long time to start showing up on Google to start ranking with these articles. And in June, so it took us six or seven months to actually start properly ranking some articles and enjoying the first like breakthrough. And then we just like started ranking for more and more and more keywords on Google. And I'm very, very excited about this because when you scale, essentially SEO is what takes you to the next level. And we're already working very hard on it. So far, we've created 63 unique articles, which are completely proprietary that we have put on our blog. And many of them are ranking theyre ranking for different words. And now we're testing and doing new stuff in terms of SEO, which may be, I'm going to share in our next status update. This is most interesting for you, digital marketing people, out there. We're using the ahrefs.com platform to monitor our domain authority, our back links, the keywords that we're ranking for. And actually for one year, we progressed from zero to Main Authority, to 45 out of a hundred, in terms of the main authority. And this is quite remarkable. I have seen many websites, I have built other websites as well. And I know that this is a very, very complex path. It takes a lot of time out of guest blogging. So we have our own guest blogs on Zapier, on thrive global strive

Iskren:

Stripe verified partners.

Iliya:

We're a Stripe verified partner. Very important backlink, from the app, we have from apple. We have from Google. You know, these are very strong indicators for the search engines and we have all these positive marks, but I remember it was March or April, we still had five domain authority or 0.9. It was absolutely incredible. And then once it starts picking up and we started showing up in other places, it rose to 20, 30, 40. Now it's 45. And I'm very excited about this because you have more crawl budget. So Google is crawling your pages more, your rank easier for new words, you make some change to your website. And two days later it's reflected on Google. Otherwise, before I remember quite well, articles which are fundamental for our SEO plan, they were not crawled for months. For three months, Google hasn't come back to this article.

And I was clicking:

request new Crow, request new Crow all the time and it was not happening. And, now we're on a very, very good path in terms of SEO. So anyways, I have several things I want to discuss and we're very ahead in time. What is your favorite feature?

Iskren:

Ah, that was actually, literally my next question for you after discussing the one that you have.

Iliya:

Well, I asked it sorry.

Iskren:

I asked to grind it, now Ill ask the best one or the one we enjoyed building the most? Okay. So let me think. Give me a second.

Iliya:

Really. You prepared this question for me and you don't have an answer for it.

Iskren:

When I prepare a question, usually while you're answering, I had time to reflect and figure out an answer for myself. I asked you back, but now you are too quick. I think, honestly, the part that Im the happiest about when I use it and see it, it just came up neat is the booking page. I just really like it. It's not the most complex. It's not the most original, it's not the prettiest either, but it's just so solid. You know, like you go to the booking page and it has such minimal information. It has exactly what you need. It has like one button you can potentially press, you can never get confused. It's clean, it's simple. It always works. I just really like the booking page. So thatll be my warrior of choice. What is yours?

Iliya:

From day one, my favorite feature was going to be customized links and it still is. I think this is the one of these killer features, one of these things that many companies are not doing well. So when you register with 3veta, you don't get to a page with 3veta.com/JohnDoe. No, you get JohnDOE.3veta.com, which sounds not so impressive, but it actually is quite impressive for me, at least, because I really love the sub domains. They give you opportunity to later spin this up on your custom domain. You can even i-frame it. You can already i-frame your website. It has nothing to do with 3veta. It doesn't have the 3veta header in these things. Say you go on Amazon and you put your product over there. Go on any marketplace and you put your product over there or yourself over there. Upwork, Freelancer.com, Udemy.com, many of these websites. Each and every product there is in the subdirectory of the website, which means it's amazon.com/product. And then you get the Amazon header, Amazon footer, or these Amazon brandings all around the place. Amazon UX, and so on. With the way we have built it, it could be much different because it is your sub domain. And this is very, very nice for me. That's the first one. Now all the meetings happen on the sub domain, which was absolutely incredible for me as well. So this is the one I liked the most. So you send the meeting link to someone, it says your name, like my meetings are iliya.3veta.com. Yes, you have the 3veta in the name as well, but it starts with iliya. And then we also did this extra effort with the structured data which is very nice. So when you send such a link on WhatsApp, on Slack, you share it on Facebook or so on, it is customized or customizable this whole link, which many other platforms don't have. And it's a very good example that I want to give you is when you share a Facebook link outside of Facebook, everything you see is facebook.com login or sign up. This is even links shared from Facebook looks like this, unless you share it on Facebook or on messenger, I think. Probably WhatsApp is fine. Yeah, so Facebook have not solved this issue and we have solved it. And this is very cool for me. From day one, I wanted this feature and we have this feature and I think it's one of these things that people really appreciate. People who care about these things, they really, really appreciate it. And yeah, that's my favorite.

Iskren:

Right. So it's kind of on the white labeling side of things, basically. Yeah. If I have a listing or some marketplace, like if I make a course on Udemy and I send you a link to my course, what you're going to see is an enormous Udemy logo with enormous Udemy letters, then my name somewhere at the bottom, whereas for us it's the actual professional provider's name, rather than ours. We try to be good. We try to be really good about white labeling. I do believe in it a lot as the future and we try to make it as open as possible for people to be able to have their own branding because people do care about this. And yeah, maybe it harms us a bit in the sense that, you know, every provider who uses their branding, instead of ours, we get less exposure. But I think it's only fair if these people pay us for our solution, they should be able to use it with their own branding. Alright, so what are the questions do you have for me? The one year later episode.

Iliya:

Yeah. So I just wanted to wrap up the product. We're talking about product so much, there are other aspects of it as well that people may be interested, With product, last thing that we should say: the feedback that we got, do you want to comment on a couple of things?

Iskren:

Yeah, sure. So for background for the listeners, basically when we did this promotion in September and October, so just now, and we got those hundreds of people signing up and becoming subscribers to the platform and providers there, we received so much feedback from people that for the first time were just flooded

with information:

what to do next, what to prioritize, which was very nice. We were really in need of such real-world feedback. And I think there was a very positive moment for the whole company, not just because finally there's real people telling us what they want from us, but also because we managed to very, very, very quickly turn around and actually implement those things. And when people asked for let's say 10 different features right now, and we managed to get several of them in a matter of days all the way to the live platform and the people from the promotion, our partners, they were replying to us like, Wow, you guys are so quick, nobody has ever deployed so quickly how are you doing it. Actually, to some extent that's because we were so ahead on product, like we had engineered so much that made adding those into pieces a bit easier because we kind of had sufficient material to start from, like we're halfway there with so many of those things. It was just like crystallization of ideas, let's say. People just told us what they want. And we listened and they're very happy about it. So many of the purchases that we got were people who asked for a feature, we gave it to them and only afterwards they replied, Oh, wow. You really did this, that was very quick. I'm going to buy a product because you listened to me and you gave me what I asked you for. So yeah, those were really, really amazing.

Iliya:

Yeah. So you still didn't quote any particular features. Several things that we added to our pipeline that we were never considering doing basically we said far, far in the future: WordPress plugin, a widget that people can embed on their websites, and Chrome extension. So, all three of these were going to be far, far ahead in the future. However, people actually requested them and because our infrastructure was such an overkill, it's quite easy for us to develop those and basically are going to be out maybe in two or three weeks.

Iskren:

Stripe verified partner.

Iliya:

And a Zappier partner as well, don't forget.

Iskren:

Public partner to be.

Iliya:

Well, a public answered to our email. Last thing about product team form. Again, on the strategy side, when we were thinking about this idea, we were kind of imagining professionals, freelancers, or small businesses and we wanted to do it for individuals. And then build it up for teams maybe later on, if there is demand. Well, demand hit us in the face and people were like, Uh, where's the team feature? And we quickly realized that building for teams is much better, of course. So we knew this from before, but we wanted to get acquainted with the individuals first. See how they think, blah, blah, blah. No, no, no, no, no. So teams of like 5, 10 people. Theres one guy who is usually buying all the software for the whole team. And this guy knows exactly what they need and if we provide it, they're going to buy, let's say, five licenses. And the decision-making process is much faster because they're looking for this thing. We realized this a bit later in our development, we're already developing. And then once we started putting, by the way, my kind of mentor or something, but I told him we're live on product hunt. He said, Whatever. Now put it in the hands of clients. Yeah. So once we put it in the hands of clients, we really, really felt this demand for the team feature. And, we did the UX UI stage and we're building the team feature where a whole team can work together.

Iskren:

Right. So yeah, people want to work together. Uh, and I guess it opened the whole direction of sales and marketing towards instead of B2C and instead of ground enterprise B2B, rather small scale B2B. And that's kind of our next main focus which got unlocked by this overwhelming feedback that people want the team feature. Okay, so talking about this, I want to make the general point, it's kind of a meta point because it refers to this conversation, as well, is the fact that we still overkill product in the following sense. We'd have discussions about the company and what to do, what we've done, what we should do in the future Product always takes 90% of the discussions, right? So as you said, there's many other aspects to it. There's marketing, there's sales, there's so many things happening. There's customer success stories. It's very important. We try to help all of our providers utilize the platform to its full capacity, et cetera. So we do many, many things, but somehow we always end up speaking essentially only about product, thinking mainly about product, briefing and teaching product. Were all about product, which show we are a product company, but I do think we overkill a bit. And so this episode is just a testimony to this. We kind of decided, Okay, let's look back in the past one year what happened? You know, it's a startup, right? It doesn't matter what product were building. We just want to introspect a bit about the past one year and ended up talking about product for one hour. So it's just showing how we would perceive things very product first. And maybe we shouldn't.

Iliya:

I generally agree with this, but keep in mind that you do attend all product meetings and you do not attend all marketing meetings. So for you its this bias of, oh, nobody talks about marketing.

Iskren:

Yeah, of course. When it comes to product, I'm like probably as involved in your product as one could possibly like I'm fully on the product side from the team. As Emperor Palpatine would say, I am the product. That was a terrible impersonation of his voice. This was not even in any way related. Um, so that's true, but still, it's true. We just, this one hour now, we did focus on product.

Iliya:

I can recommend you a book about it. It's called Rework. They recommended it to me yesterday and somebody on the team has to read it. So they want to read it. I'm wondering whether to read it myself or like a pitch to someone else. But anyways, it's by the guy who built Ruby on Rails and what he was saying was that, what he's allegedly saying in the book, I have no idea is that once the product was out, they received all this feedback and they basically started reworking the product to get to the version of the product that people actually want. And basically that's what we're doing right now, we're exactly in the rework stage where we thought these are the cool stuff, but now we realize they are not the cool stuff and people want something else. And a very, very important point on this one is you receive a lot of feedback and you have to find which feedback to implement and which feedback to not implement. So this is very, very crucial. At the end of the day, many, many people are talking about this customer is always right, but not always. Basically, that's what they say. And the fact that, let's say, two people are very annoyed with something doesn't mean that the other 98 are not super happy with it being this way.

Iskren:

Let me then just change the phrasing of the common saying that it's not true that the customer is always right, but I would say that: Many customers are always right. So if many customers are claiming something they're right there, there's no argument there. But if one customer is seeing something, maybe theyre not fully right.

Iliya:

Assuming, assuming product market fit that's the right way to rephrase it. But, so on our booking page, when people were booking, they could book

it at 1:00 PM, half past 1, 2:

00 PM. So it was not possible to book at 1:15, 1:45

Iskren:

Which was intentional. Let me just interject we intentionally did it to be every half an hour so that the booking form does have too many possible options, and slots, and just doesn't look cluttered. So we intentionally did half an hour.

Iliya:

If you did every half an hour, you have twice as few slots than if you do it every 15 minutes. And basically these two people told us, Hey, I actually do my meeting set at 15 past whatever hour, can you please implement this? And we implemented it the same day, right? Because since it was intentionally done in this way, it was very easy to go back. And we went back, and then the same person who gave the feedback, told us, Oh my God, it's so cluttered with time slots, people cant possibly choose. And we received this from this guy. And then just today we received it from someone else. And they're like, Guys, it's so cluttered, do something about this. So, customer is always right. But yeah, you should always keep in mind that sometimes you may skip a piece of it back.

Iskren:

Yeah, right. It's got to be statistically significant to really act on it.

Iliya:

But even if let's say 50% of the people wanted this feature, I'm absolutely sure that most of them would be displeased with the outcome. So even if it was very, very sought after, maybe we should have thought about, maybe we should think about, some other way to solve this issue, you know? Okay, enough of product, do you want to talk about something else? What else is there?

Iskren:

Exactly, I'm fully product line. Like there's nothing else in life.

Iliya:

Yeah. Okay. Well, you can ask me questions about sales. You can ask me questions about marketing. You can ask me questions about branding, content strategy. I don't know. What do you want to know about?

Iskren:

Let me ask the following then. We read somewhere in this article, this one, saying that the way you acquire your first 10, your first a hundred, your first thousand, and your first 10,000 subscribers are very, very different. So, you don't do the same thing. Uh, something has to qualitatively change when you scale up. So, right now were at a stage where we have several hundred active users, that's brilliant. And we know how we got here. It's through those promotions that we did with external partners. Now, where do we go from here? How do we scale from the hundreds to the thousands? What's the game plan?

Iliya:

Are you going to hold me accountable?

Iskren:

No, but our listeners will hold you accountable.

Iliya:

Okay, okay. So generally speaking, the first 10 people you acquire in some strange ways: friends, family, friends of friends, whatever. People to just get in the system. Even your team your team could be the first people to use the product. And they bump into every each and every flaw of the product, in the beginning. Then you have to find your first, let's say, 10 paying customers. This could be friends and family, but usually it's done with sales. You know, you have to sell it to someone. You have to make deep discounts. You have to build these personal relationships and explain that they're early adopters, that there's more to the product that is going to come up, that they're going to have some issues. Some people are open to being early adopters. Others are not. So this is something which by the way, we had an issue with this because getting early adopters for a platform which processes payments and does video online... This is very tough. You know, you're hitting two very big buttons, which are like: No, I don't want to risk my payments and no, I don't want to risk my video connection. So this was very hard to get early adopters. And then you'd go to the first 100 people. his is done mainly with sales and you have to reach out to people again, build the relationships, make them use the product, make them pay, close the deals very, very fast. Unfortunately for us, we kind of sucked at this. We are not great salespeople, and that's fair. You know, that's fine. You can't be good at everything, especially when we are so technical. So, I think that people who are more technical, ceteris paribus, all else equal, are worse salespeople. And, the point is that technical people cannot close their eyes to some critical flaw or they cannot oversell the product when they have built it for instance, or they have thought about it. They ask you a question and you just answer straight. And that's a very big issue. Now, if you have a sales person who doesn't know the nuts and bolts of your product, they can confidently sell the product and believe that it is superior or better, because they don't know the flaws that maybe the CTO knows. And I think this is very crucial in such a process. We were not convinced before in our product, because it was beta. It had a bunch of flaws, a bunch of known bugs that we were about to fix, and somebody asks you, and you can't lie to him. You know, ignorance is bliss. So salespeople are great at this. They don't know the product so well, quite often Now, fortunately for us through other means like communities and marketing that we have, we went past this 100 people mark, and we're well past it, of course again, because we suck at sales and we are great [I hope we're great at marketing] we managed to do this at scale and not like one by one. Now, after you reach this critical mass of people, it is very important for these people to actually use the system. So if I, the fact that we have close to a thousand accounts right now, by the way, we have close to a thousand accounts out of which several hundred are paying users out of which some people use the product a lot. They're power users of the product. Now, if you give me a hundred power users, I can do a lot of stuff with them. If I don't have a hundred power users its like a whole different story. So what is the story? So, first of all, everybody's talking about ads, like you scale with us, you do this with us, but who do you target with these ads?How do you find the people? Now, I was convinced that coaches are the best product fit for us. We're perfect for them, but for some reason they just don't trust us. And, I can imagine that they have coaching tools that all of them are aware of. And they're like, I just don't want to use this product. It's fine. We're perfect for you guys, but they're not working on them. But they work on consultants and other people. So coaches and consultants, they have very close workflows, but they're different beasts in terms of marketing, targeting, and so on. And, once I have these first power users, I can see which people identify themselves with our workflow, like the way we organize their workflow because essentially that's what we're doing. We're organizing their workflow in some way. So we can see who are these people? How do they use it? Do they use the payments, for instance. So if 10% of the people use the payments, I'm going to remove this from the ads. This is a feature which is not so relevant. And these users, they give you the information. What are the sought after features? First, messaging. So that is very important messaging, and kind of targeting. Second, first party data. Many people talk about this and you have third-party data and first party data. Third party data usually what people were doing with the ads was Google and Facebook. They have all the data of everyone and somebody else is doing something on, let's say, the website Udemy.com. They do something. And since they are using Google analytics, Google knows what they're doing there. And then you target them based on their activity on Udemy, and sounds pretty dystopian, but that was how ads were working. And that was how everyone was advertising for years. However things are changing. I haven't truly had a theory about this. I think April is pushing a lot towards it towards this change because Google and Facebook are the main winners from third-party cookies, third-party data, because they have the data of everyone and they're advertising better, making more money. Apple, they're not advertising. They have the devices. So they're like, you know what*f you, Google and Facebook. We're just going to make everything super private. And, we don't give the emails to the people. We don't give their activity. Everything is private on an Apple device and people are like, Wow. Apple is so cool. And now with the new iOS updates, 14.5 or something, I can't remember, all apps are asking you, Uh, do you want us to track you or do you want us to not track you? And of course everybody's like, I don't want you to threaten me. 95% of the people have opted to not be tracked. And if you have an iPhone, chances are your data is not shared with anyone. In April its profiting a lot. Like if you want some private usage of your device in your internet, go use an apple device, right? At the same time, Google and Facebook are just like very, very pissed at this. And, Google and Facebook are very, very pissed at this, I guess. I can imagine they can not advertise as well as before, you don't get the same data on your conversions, you don't get the same data on the ad youre running. For instance, we ran an ad, somebody clicks on it, they come to our website, they sign up. If he doesn't know about it, because he signed up from, say, an iPhone and with an iPhone, this data is removed. Facebook doesn't know. App doesn't know that it's better, and it's not being shown to similar people which is a very, very big problem. Now, if you have your own data, your first party data, meaning people come to 3veta, they do stuff in 3veta, this is our data. We have their activity. We know what they're doing on our website. We're not sharing it with any third parties. This is very important, right? So we're maintaining their privacy, but we know [kind of] what they're doing. We know how many meetings each provider has, we know how many customers each provider has, and these are normal things, right? They shouldn't be pissed at us or angry in any way that we have this information. This is needed for the system to work. But this is enough information for me as a marketer to say, John has more meetings than Adam. So find me more like John and less like Adam. And this is absolutely crucial, right? This is the game-changer in marketing. If you have your first party data, you simply somebody clicks the ad on Facebook, they sign up, and you return to Facebook their email. And, you're like, this email signed up on my system and Facebook has the email who clicked the ad. They don't know what happened on your website, but when you return it to Facebook or this email signed up, Facebook can connect the dots and, Aha! Okay. This ad worked. But, later, you can also add value to each customer. As I said, John is much more valuable than Adam. Sorry, Adam. Thank you, John. Thank you, John, for being so valuable! And so we say, find me more like John. And this is absolutely crucial for the marketing. Now, if I don't have a hundred power users, at least, like if you give me 10 K that's much easier, but if I don't have even a hundred, it's very, very hard for me to do any meaningful ads. Now, and apart from this, if I know something about John and Adam, this is very helpful. I can have a call with John. I can have a call with Adam. And I'm like, what are you doing? Oh, I'm a consultant. And then the other guy, I'm a nutritionist. Aha! This is why you use the system more than the other guy, you know? And you can start making these connections, how people use the product, maybe different segments are using it in a different way, by the way, when in doubt, always segment. And when you don't have data, you can't segment and find these patterns. Back to your question. What happens when you have many customers? First, you speak to them, know their use cases, know how they found you, know why they use you, know the features they need, know why they need them because that's what we do. Somebody says, Oh, why don't you have feature X? And the first thing

I ask is:

why do you need feature X? Like, how does it fit in your workflow? And then they explained to me, and I'm like, Oh wow, this makes so much sense. Something that I never thought we're going to build is actually absolutely fundamental for this industry. And then we can build it. And then once we build it, we can market it. And, this is very strong. You have the feature request from an actual person. He chose you because he wants this feature. You build the feature for him, and then you simply market this feature to a thousand other Adams around the world. And this is very, very powerful once you have users. So, game plan is the following. Find what people want, find what people value the most. We already have a very good idea of this: what they value more than others, for instance, recordings. This is extremely devalued. It was never in of ours. We have never promoted the recordings. And, we have them, you know. From this point on I'm like, Okay, each ad should include this somewhere, either in the image or in the text, people want to be able to record. Yeah, know the features, start adding into the messaging, use the words, by the way, this is the best use the same phrases they used. You know, this is a very good marketing strategy. Somebody says, let's say this woman, she asked us, Can I have delay? And I'm like, What is delay?And it is basically, if you try to book someone, and on 3veta you can't book them in the next four hours. After four hours and one minute is the first available slot you can book.

And the words used was:

I want a 12-hour delay instead of a 4-hour delay. I will never call it like this. And I'm not sure if that's a proper term, but I am absolutely certain that there are a thousand people out there who call it delay as well. And this is quite powerful as well for the messaging.

Iskren:

I guess that's the same concept in negotiations. I think it's called mirroring where you literally repeat with the same exact phrasing things, or whatever the person just said, to let them know that you're listening to them and that's what we're doing now. We are listening really intently to our customers. And actually, because we want to wrap this up now I'll zoom out a bit and say that that's the most important thing happening to us right now. We have people to listen to. We have people who tell us their thoughts about the product. They're not just speaking in towards, they have the so-called skin in the game, right. They have actually paid to the service, so they're serious. They're not just giving their two cents without any attachment to the issue. And they're also actually using the platform, so they know they're talking about. For any listeners again, who have, or thinking of having a business -a start up- do not overlook this, for real. Like, it takes actually immense amounts of time. When suddenly you have hundreds of people and they contact you about various

things:

some issues, some proposals, some pesky little bugs, or something that happened, or just visual issues. It's a very slippery slope. You can very quickly start just quickly replying one-liners to some of those people. But don't really. This is your way upwards, I believe, to really engage with them, not just answer their questions. Open the conversation. Ask them more questions. They're happy to talk to you. Most of those people are very happy to talk to you and they'll tell you

Iliya:

Those who ask, they love to talk.

Iskren:

Yeah, essentially. Right? And then if you ask them, you will find out so much more about their workflow and how they think, how they face stuff and easily you can use this in the marketing. So basically being busy with customer support, let's say, and customer success stories is a great thing to have on your hands because there's the key to the next level, we believe. So to wrap it up, essentially, one year in the making 3veta, where we stand now is that we have the product. It's fairly mature by now. It's also very stable. Something we don't talk too much about, but after so many iterations, now it's stable. Now we believe in the product that it just works. It doesn't break. It doesn't show random issues, which is very important for a new product to also build trust with our providers. We have several hundred active users and we are looking at how to implement and understand all of their development desires. So then we can use this as our strong points to market further. That's where we are. Yes. Thank you so much for coming up with this idea for a podcast. For a very long time, we wanted to talk about the lessons learned, essentially, on a very, very high level, because this conversation was on a very, very high level. I would say we don't get into any specifics that we do on a daily basis.

And, so it's just this:

What would you do differently if you had the chance? Well, a couple of things, but overall well, that's a part of the journey.