Converting from Translation Agency to 1m ARR AI Translation SaaS with Marko Hozjan of Taia.io

Welcome to the Big Break Software podcast. We’ll be talking with software startup founders, software coaches, and consultants and how they found their own software success. And now let’s get started with the show.

Hi, everyone, this is Geordie Wardman here, host of the Big Break software podcast, where we talk to top leaders in the software field like Seth Godin, Andrew Warner of Mixergy, and many more. This is a show where we talk to proven founders about their zero to 30,000 MR journey and beyond. Today’s episode is brought to you by onestopdevshop.io. We have access to hundreds of developers and we’re waiting to take your idea to fruition. If you want a reliable, full-stack development team with top talent, it costs half as much as in-house developers, and you know you can trust your SaaS or mobile app with us. We’ll give you the first 30 days no risk, and we guarantee to be on time and on budget, or we finish a project at no extra cost. Whether you have a software project already or want to quickly augment your team, or you have a software idea, give us a call and we can get you going in a few days at no risk. Today on the Big Break Software podcast, I have Marko Hozjan of Taia translations, an A.I. platform that helps their clients with faster translations, with human services assisting with the translations. Marko will tell us how he came up with the idea for Taia, how the MVP was funded, and where and how he gained his first few customers and was able to navigate his zero-to-product-market fit journey to earn one million euros ARR and beyond. How are you today, Marko?

Very well. So, as usual, overwhelmed with meetings, just a lot of work. The task list never gets done, but otherwise great.

Well that’s great, hopefully, a podcast is a nice break from that. Even though it’s a meeting, it’s sort of a loose – you know, for me, it’s sort of brain candy, just to sit back and relax and just have a friendly conversation with another entrepreneur. So, welcome to the show. Why don’t you start off by giving us a quick breakdown on what exactly Taia translations do, like what problem does Taia translations solve, say, that Google Translate or some of the other ones can’t do?

Yeah, easily. So Taia, in general, is a translation platform. We’re both focused on B2B, so medium to big size companies, and we actually solve many things. So, if you look at the market itself, how can you translate things? You have one side where you can use, for example, Google Translate when sending something to…

That would be just like, say, a private individual that was sending like a no-risk e-mail to a friend or something, right?

Exactly. And then you have the other spectrum where you have high-quality human translations, which are done by professional translators, and of course, accompanied by translation agencies. So, we are actually a platform that combines everything, because we’ve seen that companies have different needs at different times; sometimes companies today, one department needs high-quality human translation, the other times they need high quantity machine translation, and so on. So firstly, we have automated the whole process of ordering translations, so it’s very simple – drag and drop any kind of document, almost any kind. We support almost 70 file types.

The document is immediately analyzed and then you choose your service, your rate of quality, the time that you need it in, and your order, and everything is very transparent and safe. And because of this automation, we can be much more efficient and cheaper. And then the other part is that the big trend is coming in the industry because of automation and machine translation, and it enables companies to translate on their own, to in-source it. And we have developed a tool that enables companies to in-source that, which means that the process is actually the same – drag and drop the document in the app and the document is immediately pre-translated by the MT, and you just post-edit the text, click finish and you receive the document back in the same formatting.

So, imagine having a 100-page catalog or a PowerPoint presentation, or even a PDF or a code exported from an app, a JSON file, an HTML, and so on. So, the formatting stays intact. So, our main USP is, of course, the speed, simplicity, formatting, and file types that we support, together with security. So, this means that files aren’t sent over emails, but of course, everything is in a secured cloud.

Ok, that sounds pretty straightforward. I’m sure there’s a lot more to it, but, for example, what would be – you mentioned medium to enterprise client – so what would be the cost for this, like, what is your tiered pricing?

Sure. So, it depends on what you choose. When you come into our app, we actually ask you a simple question – would you like to outsource the translation job, or do you want to do it yourself? If you outsource, the business model is pay-per-word, which means that you choose the level of quality and the time you need it in, and it is automatically calculated. So, the pay-per-word, for example, so for translation, one word can cost 10 cents. It depends on the level of quality…

Okay, that would be on the high end, right, because to me that sounds like it would be like on the high end.

Yeah, high end, exactly. That’s for the TEP – translation, editing and proofreading of a text – while the software tools are subscription-based, so a subscription model, going from 40 – we operate mostly in euros, but in dollars as well – so 40 euros per month, up to 140 euros per month. It depends on the package that you choose. So, it’s based on the number of users and how many words you need.

Okay, so for example, the one that’s 40 per month you get, say, like 100,000 words or something like that, and it’s based that way. Correct?

Yeah, exactly. So more than 10,000 words per month you get. And of course, everything else is unlimited and you can use it. And of course, it speeds up the whole process. So, what we have learned is that post-editing, instead of just translating or copy-pasting from Google Translate, Amazon, or anywhere else, is four times faster with our tool. You drag and drop a document, you post-edit the text, you click finish and you have your formatted document.

I’d like to get into some accuracy because I’m sure – you know the one thing that I know about at least translating things into English, so if I took say like a French block of copy and translated into English, there are different interpretations that can come from using, say like an AI or machine learning algorithm, and it really sorts of depends on, almost like the author’s intent. Can you speak to me in terms of the accuracy of what, say, like a Google Translate would be compared to some of the other ones in your service? Like what kind of accuracy rate are you generally giving your customers?

Sure. When it comes to MT, the answer is not simple. So, in the world, there are now about 15 very good neural networks, I would say, that are available for public use or even private use. And then companies or LSPs create custom MT engines for different companies.

Sorry, can I just interject – when you use the term MT, is that machine translated? Is that what that term is? Okay, I just wanted to clarify.

Exactly, exactly. No, I’m sorry, that comes with the work.

No worries.

So, Machine Translation. So, some companies or LSPs use public neural networks or MTs for their own use, and some build their own. To build your own, you need your own corpora, so the data set, which has to be quite large, and usually the better the data set, the better the output. So firstly, if we speak about publicly available, free or not free (most of them are not free) MTs, it really depends on language combinations and branches. So, when you speak about quality, the quality can be, if I would put it in percentage, from 60% correctness up to 99% correctness. So, there are cases where you input a document – I don’t know, an official legal document – you get it back, even from a service like Google Translate, and it is remarkably correct because the input data for that kind of text was already input.

So, the more the text is technical or rigid, the better the MT output is. The more the text is creative, like marketing text, literature and so on, the worse the output is. So, there is a big variation, but this variation, of course, is narrowing down. So, the quality is getting better, not month by month, but actually week by week, and this is being enhanced by custom MT engines. So, what we’re doing, for example, we’re combining different neural networks, so we’re connected to them. So, we’re trying to source the best part of each one, plus, we’re building our own – our own custom engines. And on top of that, there is an older technology that we’re combining called translation memory. It’s a very simple tech where you have a segment of text, and once you translate it, you know that it’s 100% correct because it’s been translated by a professional translator.

So, for example, we build a special corpus for each customer, which means that what we translate for each customer, it goes into the machine translation, plus it goes into the TM, and whatever has already been translated, we don’t need to translate again. And the same with Catapult, so the tool where people translate on their own; whatever they have translated, they will see next time. If the same segment or a similar segment has been translated, they will see it, so they don’t need to do double the work and the quality, of course, is higher.

So, what would you say in general? It sounds like the custom one is 100% accurate because that’s human-translated. And then the SaaS packages that you have, what sort of accuracy are we talking about, 90% to 99%? Is that sort of the range that you would say, and it’s based on, whether it’s creative or technical?

So the MT? The accuracy of the machine translation? I can give you the average, I could say 90%, but this is far-fetched because some can be much worse and some can be much better.

Based on whether they’re creative or technical or whatever?

Exactly.

Yeah, okay.

And then, first, language combinations. So, for example, the biggest world languages are very well supported. English to French, the outputs are amazing. While if you go, I don’t know, English to one of the African languages that are only spoken there, for example, the output is much worse because the data sets are less.

Okay, fair enough okay. That makes sense. Now, why did you come up with this idea? How did you get involved in this?

So before that, my partner and I had a language school which was a local language school here in Slovenia, Ljubljana. And at the time, we started getting first requests for translations as a language school, and we actually started a sort of traditional translation agency. But very quickly, we realized that we can only compete with price, so there are no USPs. And since we consider ourselves tech-savvy, we quickly figured that the whole market is very much outdated. So still today, this was four years ago, and still today we see that more than 80% of the market is covered by traditional translation agencies that operate in an old fashioned way, which…let me give you an example, you need a translation, you find an agency on Google, you send them an email, they tell you “okay, send us the file”, you send them the file, they put a file in software to analyze it, they use another software to create a quote, they send you a quote… You catch my drift. So, a lot of time is being lost and a lot of manual work, and then if we would dissect the whole process down the road, you would see that a lot of manual work is being done. But that was a previous concept of an agent, so an agent did all that work and organized your stuff, but now all of that can be automated, and in general, in all the industry, the profession of an agent is dying. It doesn’t matter if you’re a real estate agent, translation agency, whatnot. Automation is impacting the agent as a profession because all of this work can be automated, and this is how we started that within the translation industry.

Okay, so that was four years ago, and you guys had a language school. Are you sort of polyglots? I mean, how many languages do you speak?

I wouldn’t consider us as polyglots either way. So, I’m an economist. My partner is. He studied Chinese, that is so, but now when I see our team, most of the team studied languages, that’s true.

So there is an interest in language, right? I mean, some people just have an aptitude for language, they have an interest in it. Would you characterize yourself as someone that does have an interest in languages?

Of course, otherwise, I wouldn’t be in this business. So, it’s business in combination with languages. Firstly, I didn’t know. I’m a serial entrepreneur in general, so I was interested in entrepreneurship. But when you get into the language industry, it really starts to open up. Just to give you an example, the language industry is $50 billion in size and it’s bigger than the music industry. But since it’s not as sexy as music, people don’t even know, but it’s really vast, and it’s changing so much because the tech is influencing the industry, so it’s so interesting to work in the industry these days.

No, I just finished – I’m sure you’re familiar with it – where Benny Thomas, the Irish guy that did Italki, which is a great SaaS that I use. So, I know it’s big business and I’m sure he’s doing very well with his SaaS and his writing and stuff. But he’s definitely into the language. You know, he likes going and diving in, he’s learning Hungarian, you know. I don’t know, his languages are probably up to like ten or something, you know? So that’s what I imagine, you know, the sort of language nerds that get into this stuff because it is very interesting, you know, and you do get a lot by learning the language. So that’s essentially the type of people that are attracted to Taia?

Yes. So, for example, when we get applications for different jobs, if it’s office administration or dev, people usually are interested in languages, so there is a correlation there.

Great, so let’s get into the MVP. How did the MVP evolve? You guys came up with this idea, you saw that there was an inefficiency because people are sort of, you know, maybe copying and pasting and using word docs and, you know, trying to manually figure out word counts. What was the sort of starting point like for the MVP? Like what was the core problem? And walk me through the MVP, how did you guys fund it? Did you do some pre-sales? You know, anything like that?

Sure. So, as I mentioned, we started as a traditional agency, which means that we did stuff like everyone else. So, immediately we found that that’s something that we don’t want to do. So, we started writing the whole process – how it goes from the point where we receive a file and then when we return the translation. And the first USP that we found that we could solve is actually connected to efficiency and, of course, speed, which means that we were thinking, how can we make the process of the file flow from the customer to the PM to the translator to the proofreader and then back, faster and automated? So that was actually the point of the first MVP, so that the customer just drags and drops the file, receives an immediate quote, confirms it, the file goes immediately to the PM or even the dedicated translator, the translator confirms, they start translating, when they click finish, automatically without waiting, it goes to PM for a final check and then to the customer. So, speed was the first thing that we really wanted to solve, and we did. And actually, since it was a front end, in the beginning, it was very interesting for customers to see. At the time we didn’t have anything else, so we said, look, see the first part, the analysis, and the quote is so much more efficient and so much faster for us and for you. Can you imagine how much we can make the rest more efficient? And they said, yeah, this is great, so it was easy to onboard the first customers.

Okay, so it sounds like it’s a really easy translation. You were an agency, you transitioned to a SaaS – can you talk to me about sort of the trends, the transition between agencies to SaaS? Like what sort of revenues were you doing at the time with the agency? How many people were there? Was it just you and your co-founder or what was the situation when you were an agency and how long did it take for you to do the MVP? The agency, I guess, was funding from cash flows to build the SaaS. Is that how you funded it?

Almost. So firstly, the company was founded in 2017, and my partner and I were ignorant enough to do it in a way that we hired a CEO and said now you’re going to do everything, this is the idea, and that’s it. So, we actually were working on other projects at the time. So, this didn’t work out, the company made 10,000 euros of revenue monthly, so it was just peanuts. So, at the beginning of 2018, I stepped into the company because I really believed in the idea, and that’s where the idea actually was starting to be born. So, at the time, when we made the concept, we received our pre-seed funding of 200,000 euros. That was in the summer of 2018 and in autumn of 2018, the MVP was born. So, that’s when we actually became Taia and started to sell Taia. Even though I say we’re four years old, we’re actually about three years old. And from then on, the actual growth and development started. And then, so at the time, there was a team of three people and then it became a team of five people. And in October 2020, so just one year ago, we received our second pre-seed funding of 1.2 million, and that’s where we really started to grow. So today, we are a team of 34 full time, spreading across Europe, so UK, Spain, Slovenia, Cyprus, because we onboard internationally, and recently our new head of sales is American, and we are planning to generate one million euros of revenue this year, 3.5 next year and previously in 2019, we generated 300,000, which means that, for now, each year we’re planning to triple our revenue.

Okay, which is great. Okay, so where did you get that initial investment?

So it was here, a local firm, it was Swiss…

So when you say here, you’re saying Slovenia?

…investor. Yeah, yeah. So, Slovenia is very much in the center of Europe, so everything in Europe is close to us, so we need two hours to Venice, four hours to Munich or Vienna, so everything. So, the heart of Europe is very close and our first investor was actually sort of Slovenian-Swiss, so a Swiss company invested. At the beginning, it was more trust in the founders than the idea itself, but that later changed. I think that’s very, very common for early startups.

What do you mean “later changed”? How did it change?

That the investors started to believe the idea. So, when we started, they said, okay, there are millions of agencies, millions of translators, MT is developing, how are you going to compete with that? So, we needed to prove that our hypothesis or niche is valid.

Why did you need money? I mean, maybe that’s simple, you had other jobs, did you feel like…why did you need the money?

Development. Developers. So, development is expensive, so we put in approximately 100k. We didn’t have any more money, so we needed an external investment.

Okay, and how did you do the valuation? You don’t have to give me exact numbers, but what sort of equity breakdown did you have to give your initial funders?

Sure. So, it’s easy, so the evaluation at the time was a bit less than one million euros.

They took 20%.

I think every investor would agree with me, evaluations of startups is something that is just not…

No, it’s pie in the sky sometimes, yeah.

…usually…exactly, so it was similar with us. So, we had financial projections there, but of course, they were projections, because you don’t know if the idea will fit on the market and so on. So, as I said, mostly, I think it’s the trust in the team, in founders, plus the idea, plus the projections and you come up with a number.

And how difficult was it for you to find that funding? I mean, was it a client or where did you find them?

It was actually very difficult. So, I think to find the first investment, it’s much more difficult than to find your third or fifth because at the beginning you really need to persuade someone to trust you as a person mostly, because, for example, the second investment we got from Fil Rouge Capital and for example, we spoke to them even before the first investment, and they weren’t interested. But then when we proved some traction and revenue, they started to get interested and later on they came in with a much higher valuation. But it is how it is, so the earlier you come in, the more risk you take on but the more you can get out of it.

Okay. So, talk to me about some of the difficulties that you had found it. Like, how long were you looking for investment? I looked quickly into your background. You have an interesting background, particularly the crisis management stuff that you did. I’d like to hear more about that, but were you doing that at the time? Is that what you stepped away from to try to get into the translation stuff?

Exactly. So, my personal, I don’t know, strong points are on two sides. One is leading growing companies, I can say now, and the other is leading or helping insolvent companies that are very much in-depth or in a deep crisis. So those are my two personal subjects of expertise. So, before Taia, I was actually running a company like that. So just to give you a perspective, I needed to personally lay off 60 people one on one. It was quite a big company, so we had to fire 150 employees and I…

You were sort of a hired gun to come in and lay these people off. Is that right?

Exactly. So just running an insolvent company is not easy because at some points you are personally even responsible for the actions. It’s not a limited responsibility anymore when it comes to insolvency, so it’s quite tricky. So, at the time I was there then because I saw a big potential in Taia, I came here and when it comes to investment, yeah, it wasn’t easy, but I think we had some luck with it. So, for each of the investments, we didn’t need much time. So, people usually say you need at least six months from the point where you start to the point where you get the term sheet and then another three months to get money in your account. With us, I think it was less, so with the first investment, I think we closed it in three months, with the second in six months, but we could have closed it much sooner. But there were some just administrative problems that we needed to solve, otherwise, it could have been done in three months as well.

Where were you with revenues when you went? Like, how far along was the MVP and how many customers did you have when you went to – because if it took, you know, you said three months – like at what point did you decide it’s like, okay, we need to go get funding. How far along was the MVP at that point when you decided that you needed to go get funding, and then where was the cusp? You know, just…

The MVP wasn’t made. So we had sort of a demo, a landing page that showed the main functionalities. So, with the demo, how it works, we actually pitched the idea and with the money, we could go deeper. So, at the time…

Okay, so you didn’t even have the MVP ready, you were an agency at this time. Okay. And then you just sort of had like a demo which was what, like a mockup in a PDF or something like that?

Yeah. It was an online demo. It actually worked. But it was ugly, and just so you could understand what the function is.

Okay. So, the Swiss company invested 200,000. How long did that last you until you were looking for more funding? Were you using your own cash flow at one point to sort of fund the business?

Yeah. So, the majority of the money went to development. We spent it in a bit more than six months, but at the time the monthly revenue started to grow, so we were able to live a bit longer. So, then immediately (and that was June 2020), we started looking for a second investment and in October we got the funds into our account.

And then how long was the transition between when you first got funded to the second round, because you said you just closed, right? Was it like two years or something? So, the second fundraising that you did, the second seed round, when did that happen and how long was the transition between the two rounds?

So the first one was in summer 2018, and then the second one was in October, I mean, the money got into our account in October 2020.

Okay, two years, and in that time, it sounds like a year and a half you were sort of…maybe a year, you were running on cash flows, is that correct? Okay. And at that time, I mean, was it always your intent to raise money? Because it sounds like maybe at one time you were thinking about bootstrapping. Can you walk me through that transition from bootstrapping to raising the investment?

Exactly. So, when we built the MVP, we figured that we’re going to need much more funds to build something that is really useful. So, it was very tricky at the time and very difficult because we were low on cash flow, we were barely breaking even each month. But on the other hand, the MVP was still just an MVP. So, we felt like we weren’t ready for a serious investment until one point when we said, okay, now we have to do it and we felt that we were ready enough. So, there was this period of one year when we were a bit stalling, I guess. But on the other hand, some of the development was going on at the time. And this is how it works in startups. So, the more traction that you have, the more you’ve developed, the higher the evaluation you can get. So, it’s sort of speculation of what is the best time for you to tend to the investment part.

OK, that makes sense. Essentially, what you were trying to do was raise the valuation so you could go out and raise more funds by giving less equity. And were you happy with the second round? Like how long is that going to last you?

So we are currently spending it so it will last until May next year. That’s why we’re fundraising now and we’re raising another 3 million.

And how are the subsequent rounds going in terms of like the ease of finding people? Are the current investors, you know, helping you with follow-on? Are they helping with the fundraising? How’s it going?

Yes. So, the upper part of the funnel is much easier. There are a lot of VCs with a lot of money searching for good startups, so it’s easy to find them, easy to speak with them. This part is easier. But then the lower part of the funnel is more difficult than the previous rounds because the due diligence is fierce. So, they go into every detail – what you do, how you do it – because before there isn’t so much to investigate. Now we are a company, which means you need to investigate the tech, how true it is, and need to investigate the revenue, how is it structured, and so on. So, in this second part, it is more difficult. We have received some help from current investors, mostly by connecting us, and with the current investor, we have a pledge of 10% of the investment already, because they have a sort of a physical limit of how much they can invest. But they pledged to cover 10%, so out of three million, we’re actually looking for 2.7.

That makes sense. And so, talk to me about some of the sorts of issues that you’re dealing with now, because it sounds like there’s a lot of corporate governance and, you know, doing the KYCs. How much of your day is spent doing that? Are you the one primarily involved in leading that? It sounds like it would probably almost be a full-time job.

Yeah. So, most of the things regarding corporate governance are managed by our office manager, or we call her the finance officer, actually. So not much of my time is spent with that. So, most of the work regarding this comes with the multiple companies that we have. We are actually a UK company, headquarters in the UK and the UK company owns the Slovenian and Croatian company, which means that a lot of work with that comes with the invoicing between companies. And then we have more than 3,000 translators in our base that we work with, about 400 regularly. And there’s a lot of work with that, with sending invoices from all over the world and different regulations. So, I can say our accounting and our finance officer has a tough job.

So is this part of the…it’s not the agency work, it’s part of the services component of Taia, is that correct? Okay, I didn’t catch on to that. So, how does that structure work if you have 400 – you guys have 35 employees, say. I was presuming that the translation arm was in there. But so, tell me about how you found these translators and how that works in the service.

So in general, our platform has two parts. One part is a service part where we offer high-quality professional translation and the other part includes many SaaS modules; we have, for now, one product and we’re going to have a series of products and they’re all going to be software and subscription-based. But this first product is actually complementary to other software products. So, this first part requires a lot of work. So, we have our project management team, vendor management team, we have three translators that are employed, but the rest are out there. So, the service part of Taia is not a marketplace because we want to ensure the highest quality. We organize the actual translations. So, when we get an order, this part operates as an LSP or as a translation agency but is highly optimized. So, we have found 3,000 translators, we have vetted them, tested them, onboarded them, and they are now in our base for the variety of language combinations and service categories. So that’s how we can ensure the highest quality service part.

Okay, can you walk me through when you sort of branched that out? Did that evolve from the agency? You know, getting all of these – because, to me, it sounds like a marketplace. How is it not a marketplace?

The Marketplace would be when we would connect the vendor with the buyer…

Okay, it’s all sort of behind the platform. You guys are doing all that. The client doesn’t really know.

Exactly, they just order a service.

That makes sense. And what’s the breakdown now in terms of revenues?

Currently, 80% of the revenue still comes from the service part, 20% from SaaS. But next year, I’m sure this is going to change because the service part can only grow organically, and the SaaS part can grow exponentially. So, of the next three SaaS modules that we’re going to input into the platform, one will be machine translation as a professional service, because we’re getting more and more demand for only machine translation, not a post-editing tool, but machine translation, where you drag and drop the document, and get it immediately translated with the formatting intact. Then, the second will be a TM management solution where companies can have that solution for themselves. So, translation memory management solution, and then the third will be – actually the whole platform we call translation management software. It’s sort of an ERP that we built for ourselves. We’re going to start as a subscription, offering it to other companies. Those are more complex companies that have their own localization departments. They have teams and then they want to organize the whole translation flow by themselves. So that will actually be our fourth SaaS product.

Okay, so 80% is coming from the services. And what kind of margins are you able to get on that? Because that seems like it would be very labor-intensive. Is it pretty limited to growth? Are you going to try and phase that side of your business out? Or it sounds like you wouldn’t really ever be able to…it would be quite handy to have that network of translators to always be checking the work, but you’re trying to build up and grow the SaaS component of the business. Is that what you’re mostly focusing on? Like, even though 80% is from the services component, you’re trying to put your resources into the software, and… how’s that going?

Exactly. So, for example, the margin in the service part is about 60%, while in the software part much higher; it’s about 90%. It depends a bit on how well the software is done, but yes. And so, until the previous year, we were mostly focused on setting the pillars, which means that we were focused a lot on the service part. But today we’re almost 90% focused on the SaaS part because we know that that can grow. But again, it’s very important to understand the concept of the whole platform because the service part is the crucial part of the platform. So, a customer, when they come into the platform, they want to solve any kind of translation needs and high-quality human translations are still going to be very much a fact for the next 10 years. And when you want a 10-million-euro contract translated, you won’t be able to trust the machine yet, and it’s the same with a catalog or any other things. But what we’ve been able to do is automate the processes. One is the whole logistics of the file, and the second is the translation process itself. So today, none of the documents are being translated from scratch. Every document is done in such a way that professional translators are doing only machine translation post-editing, and they are much faster. And let me give you a statistics example. So, for example, you input a document and the document is split into segments (a segment is a sentence or a paragraph). So, for example, a document has 1000 segments that need to be translated. And, for example, the translator confirms that the segment is perfect or corrects it and then confirms. And what we figured is that the MT is getting so good that more than 70% of segments are being only confirmed by a professional translator and not corrected anymore. And you see the efficiency here. The machine translation, in general, it doesn’t matter where it comes from, but the combinations that we use, our translators are much more efficient, much faster, with the same quality.

That makes sense. In terms of combining the two, I gather that you’re using your own software for the translators and that’s what’s making it so much faster and easier. So even so, that service component, you still have a big software component of that because you’re using your own tech for them, and that’s why the margins can be relatively…I mean, because 60% is still good for a services business, right? And you’re using your own software. Has there been any sort of…You’re a big crisis management guy; what are some of the lows that the big problems that you’ve had to solve – you know, getting this going and making the trend, the transition from the agency to the software – what are some of the big things that you’ve had to go in and use those skills that you have in crisis management?

Yeah, every company has crises. So, the first was definitely when we ran out of money with the first investment. So that year of just bootstrapping was difficult. So, a lot of reorganization needed to be done so…

Give me an example of what was difficult. Like, what were some of the things going on, some of the core challenges? Do you have…?

So for example, we were not liquid. So, we were late with payments for not only the vendors, even our own staff, or for rent. So, for example, I needed to renegotiate how we will pay rent. I needed to lay off. I needed to apologize to vendors and so on. So, when we had real liquidity problems, I had tons of work with only debt, with dealing with day-to-day problems, and motivating myself that we were going to make it and then motivating others even more because your employees aren’t always as motivated as you.

Yeah, that’s right. It doesn’t look good if you can’t pay your bills. Can you talk to me about that period? Like, where were you living? Did you have to move back in with your parents or anything like that?

No, so our personal lives, in that case, weren’t impacted. Okay, we had a low wage, but either way, so I personally had some accumulation from previous businesses before starting Taia. I actually made two small exits, so I had some personal accumulation. So, my personal life, in that case, wasn’t impacted. But the business itself was. But I didn’t have additional funds to put into the business. So we had to make it.

So what was getting you through that time? I mean, you know, you’re at a point, the business is sort of not being able to pay your bills. Some of the employees are questioning whether this project was going to work. What made you so sure that it was going to work?

I don’t know. I guess it’s the mindset. It’s faith in the business model and in the business. Firstly, you have to have general motivation or ambition to do something good so that is the basic plateau that you need to have as a person. Each entrepreneur needs to have that. And then I just believed in the business model, that the idea was good enough and that we just needed to go through it.

Did you have any coaching or anything to help you through that? Because mindset can be so important for entrepreneurs, and it sounds like you have a very positive mindset. Were you doing anything to work on that?

At the time, I didn’t. But I don’t know, I consider myself as a person that has a very, very high-stress threshold. I don’t know where I got that, but sometimes when it comes to this when other people say, I can’t do it anymore, I just can. So, I don’t know, it’s just a personal trait…

But that’s what you were saying about crisis management as one of your top skills, right, which is great for a CEO to be able to navigate those. Because every company has them, of course, you know, it’s not always just roses. So, you’ve got to the point where you are now, how are things going now? How is COVID for you in terms of growth? It sounds like it was probably pretty good.

So there was some impact on the industry in general – only a 6% drop in the whole industry. The industry was not impacted directly, but indirectly. For example, the main buyers of translations canceled a lot of international projects because of COVID. And then subsequently, of course, there were fewer projects there. So we had some setbacks, but not as much. And organization-wise, we are a very, I would call it, liberal modern organization. For example, we really have autonomy. It doesn’t matter when you work, where you work, so…

Are you a remote-first company?

Not so much remote first, but when you start working for Taia, it doesn’t matter where you work from. So, if you have an option to work from the office somewhere, of course, you are very welcome to come. But far from obligatory. So, in this case, we really harvest the new ways of working motivation 3.0, agile, scrum, and so on. And with the new generations, Generations Y, Z, millennials, it proves to be very good.

Yeah, that makes sense. And in terms of your future for this software, any big projects that you have in mind? What are some of your big goals in the upcoming years?

Sure. So, we want to become THE platform for business translations in general, so the platform you go to as a company when you have any kind of translation need, small or big. So, we are currently, as I said, finishing our MT module that we’re going to offer as a SaaS – so machine translation as a professional service – and then a variety of different services. I could say our business model is becoming similar to HubSpot, for example, where you come into the platform, some of it is free, and then you just pick different modules and different packages of subscriptions that you want to use. Because the service itself, when it comes to translations, it’s so, so vastly different. There are such different needs, so many different services, so many different languages in the end. So, you really need to give that option, that freedom to just choose what they want.

And currently, who is the leading SaaS in this space?

Wow, it’s difficult to say who is leading. For example, I could mention a company, Unbabel.

I’ve heard of them, yeah.

They are very big, they got 90 million funding, but they’re mostly focused on customer support. Other things as well. So, there are plenty; when I mentioned how the market looks, 80% is still held by traditional agencies, while 20% by these progressive tech companies that are changing the system. So, there is still a lot of potentials to grow. And these 20% are actually our competitors, direct and indirect, it depends on how you call them. But either way, some of them are doing very well.

You see, you mentioned you’re a serial entrepreneur, do you have an end goal? I mean, are you trying to – if you want to become number one, is this a potential unicorn play, like billion-dollar, you know, for you to get to grow? Or if somebody came in and offered you, say, 20 million right now, would you get out and go and do something new?

A very difficult question. So, either way, we have a unicorn potential, but it’s all up to us and of course, some market scenarios, some luck in the end if we can make it or not. So, it’s a dilemma that a lot of entrepreneurs have. So, there were cases where entrepreneurs said no and then they sold their company for 100 million. And there were cases when they said yes, and the company went bankrupt the next year. So, it’s really a million-dollar question.

Yeah, I mean, obviously you’re early on in the growth curve, but you know, it’s just interesting to hear what entrepreneurs…you know, what their vision is for their company. Especially ones that…

Yeah. Today I would probably say no. But sometimes you need to sit. It depends. Some entrepreneurs are late and tired, and there’s a variety of reasons why they choose something. Maybe they see potential in the business, but they’re tired. So, I personally am not tired, so I’m ready to fight and have many fights. So, in this regard, my answer would probably be no. But still, the end result for me is the exit. These are the end goals, yes.

Yeah, well it sounds like the company has a lot in terms of assets, the network of translators, very valuable, and the IP, and it sounds like you’re on the right path. I want to thank you so much for your time, Marko. The last question is, what would you tell your former self in 2017? If you could tell yourself one thing at that time from all of the things that you’ve learned in your journey, what would you tell your former self?

I would say, firstly, keep the faith. Persistence is key. And secondly, I would say build habits. I would give myself some suggestions of which books to read.

Which books would you have yourself read then, you know, presuming that those books were out, then?

For example, The Power of Habit, and then for example, Never Split the Difference by Voss and some books that really gave me a great insight into…

Sorry, what was the first one, Atomic Habits?

Yeah. Atomic Habits or The Power of Habit. They’re quite similar. But I read now about 50 books a year. Actually, read and listen, mostly audio, and they really give me a lot. So, it’s sort of a brainstorming process while reading.

So the reading – would you say that’s a new habit that you started since you started this startup and that’s contributed to some of your success?

Exactly. There were many habits – reading was definitely one of them and then regarding health and just exercising and so on. But reading definitely was one of the important ones. And another book I would say is Mastery, by Greene.

Greene, okay. Mastery, I’ve not read that one. I’ve read the other two. I had the resolution to read one book a week for the year. And it’s the same as you right? To us, it’s not that big of a deal. But now I’m getting closer to 70 and I’m on my fourth year of doing that and I don’t even count it as a resolution anymore. It’s just a habit, as you say. So it’s very powerful.

It’s a great habit to have.

It really is. And where can people find you if they want to follow up or find out about your great service? Obviously Taia, we’ll have it in the show notes. But if people want to reach out and talk to you about specifics of any topics that we’ve discussed, how can people reach you?

So either way, our website is Taia.io. Otherwise, my email, [email protected]. I’m available on LinkedIn as well. Like, I guess, every entrepreneur, I’m always open. I read all the LinkedIn messages that I get. So, as an entrepreneur, I always have my eyes open for anything interesting that I see, so, anyone that writes to me, I won’t keep my door shut.

Okay, thank you so much for your time, Marko. Very interesting chat.

Thank you. Likewise.

Thanks for listening to the Big Break Software podcast with your host, Geordie Wardman. Be sure to click, subscribe and check us out on the web. Keep listening and your software “big break” could be right around the corner.