How Klooks Hit $800K Selling Brazilian Private Company Financials to CapIQ

Guys, klooks.com.br was launched back in 2021, 2022. He broke $500,000 of revenue in 2023, cleaning Brazilian financial data and selling it to big firms like Captivate IQ and Bloomberg, which many of you guys might actually be listening, might actually pay for directly. Some of that data is provided by these groups like Klooks. He has specific OCR technology with his seven engineers, but then it's the human layer on top of that where they clean the data, sort

it, filter it, make sure and check for accuracy before they sell it off to the end market. That really is their secret sauce. 50% of his revenue today, and they'll do about a million dollars this year, 50% of that is data as a service. Another 30% is pure services, and then 20% is true software as a service. Completely bootstrapped, selling through two major resellers in Brazil. That's NeoWay and TTR Data, where he takes a cut of

the sales there. If he goes direct, he's selling for 300 to $400 per month and has over 50 customers to date. Hey, folks. If we haven't met yet, my name is Nathan Latka. I launched and sold my first software company back in 2015 and went on to write a book about it, which you guys made a Wall Street Journal bestseller purchasing over 30,000 copies. Thank you so much for that. After the book, I launched this show

and went on to create founderpath.com. I raised a large fund to do non dilutive deals with B2B software founders. So far, we've invested in over 400 software founders totaling 150,000,000. Here in 2024, we're doing three to four new deals per week. So if you're looking for capital and don't wanna give up equity, go sign up at founderpath.com for free to get your offer. Alright. Let's jump into the interview. Hey, folks. My guest today is Alex Abu

Jamra. He is a former m and a adviser and manufacturing CFO. Today, he's a CEO and founder of Klooks. He graduated in business administration, summa cum laude, and, again, is now building this tool which helps structure unstructured financial data and sell it in a SaaS, DaaS model. That's software as a service, data as a service model. Alex, ready to take us to the top?

Hello, guys. Great pleasure to talk to you again.

Yeah. We appreciate it. We had you back on back in January 2023, about a year ago. At that point, your revenue was about 60% data as a service, 20% software as a service, and then 20% just pure services. What's the revenue mix today? What are you selling?

Right now we are on around 50% of data service, 30% on service itself and 20% on SaaS, but everything has grown from there. We've had some interesting growth last year, kind of 50% growth. And this year we are probably reaching $1,000,000 in revenues.

That's great. So targeting 1,000,000 in revenue this year. What did you end last year at?

It was kind of 800,000. Yeah, that was it.

That's great. Now who are you selling this data as a service data to? Is it banks and financial institutions?

Yes. Actually originally the data we would sell to other intelligence platforms. So we would sell to Bloomberg, to CapIQ, Movies Analytics. So these guys would get our data, which is like financial data of private Brazilian companies and just put on their products, on their SaaS. And that was our original revenue model. And well, after a while we started selling it to banks as well. So

that's our rush right now to put our data directly into banks and not only in intelligence platforms. That's part of our priorities at the moment.

Is CapIQ still paying you today? Are they still a customer?

Sure. Yeah. Yeah. Yeah.

And what are they paying you for? Is it specifically... I know you specialize in financial statements from private companies in Brazil. Are they paying you for all your Brazil company data?

That's right. They are paying to have financial data of private Brazilian companies that we crawl in the web.

And if your data is really valuable to somebody like Bloomberg or CapIQ, eventually don't they wanna come buy you so that you don't sell your data to all their competitors as well?

They might, but I don't think that that's attractive enough at the moment. I mean, it's, Bradesuza is a small target for them. It's not like, well, I'm the only one who has data of private American companies and that's a huge

competitive advantage. It's kind of a marginal market to them. So I'm not sure if that's something that would make sense at the moment, maybe once we reach like 5 to $10,000,000 in revenues that might make sense because our process is really unique. It's been proven operationally and

in practical. So, and financially has been growing. But as soon as we escalate enough, I think that that might be something they might look into because these guys, have thresholds for M and A, they're quite higher than our size right now.

Yeah, that makes sense. How many customers are paying you today?

Oh, directly might be 50 or 60 and indirectly might be 500, 600.

Give me an example of an indirect customer.

So I sell my data to CapIQ and to all these guys, but I have some clients that they don't pay me for my data, but they integrate in their product and they sell my module. So these guys, they have like a 100 clients, they use my module. So that's the indirect customer.

Can you give an example of one of those companies where your module is built in?

Sure. There's some guys in Brazil called NEOWAY, N E O W A Y. They're quite big in the big data environment here. And well, they resell our data in a module inside their products.

So if NIO Way in Brazil sells your module, the Klooks module for a $100 a month to end users, how much about a $100 will you keep versus what NIO Way will keep?

Well, when they distribute me, I keep less than when I sell it directly, obviously. So I have large quantities, but I don't have

the value per customer that I would have selling directly, but I don't have any CAC as well. I don't have any cost of acquisition of customers. So I would sell directly for roughly 3 to $400 monthly, and they would sell me to like a $100 monthly.

So same question though. If NIO Way sells your module for a $100 a month to an end customer, what percent of that revenue will NIO Way keep versus what will they pay you?

Oh, no. That's what I would get. They would sell it for a little bit more, and then we split it.

How do you contractually make sure that your value added resellers like NIO Way don't cannibalize your own direct sales by decreasing your module price point inside of their channels super low?

Yeah, it's an awesome question, really good. And the answer is, well, I'm not sure about it. I don't have a way to make sure that they don't cannibalize. It's more of an

empirical conclusion. Well, I think it's not cannibalizing. I'm not getting any feedbacks that indicate that that's happening. Hence how we conclude it.

I see. How many value added resellers like NIO Way drive you customers every month?

Have a NIO Way and another one called TTR.

Okay. So just two, two big value added resellers.

That's right.

I see. It's ttr.com?

It's ttrrecord.com, I think. It's transactional track records.

They originally sell data of transactions like

EV EBITDA multiples. That was their primary business, but they integrated our solution And they are focusing in emerging markets such as Brazil and other South American countries.

Is it t t r data dot com?

Let me check here.

I'm certain it's ttrdata.com, but you can tell me if I'm wrong there. So those two channels do a lot of a lot of your reselling. Is that right?

Yeah. T t r... Yeah.

Yeah. So they do a lot of reselling. You also go direct. Tell me more about your process. I'm always fascinated because, you know, the whole world... I mean, maybe not the whole world, but anyone potentially can get access, right, to a private company financials in Brazil, I imagine, because they have to file something on some public domain. But you have some unique process you use to get to the data, to transform it, to clean it, to then spit it back out so that these partners can sort of use it. What did I just say that's not accurate? Or is all that pretty much true?

No, that's pretty much true. But the thing that makes us unique is the data is really hard to find. So there are some sources that are easy to find, but most of the sources are hard to find. And we're the only ones that are investing heavily on crawlers to hard to find sources. And when I mean hard to find sources, it's not like, oh, it's a website that had everything structured. And then it's just like a

bot that goes there and gets to the end of this stuff. No. It's a complete mess because the hard to find places is like thousands of different files that you don't know if that's a financial statement or if it's not. So our bots need to find out if it is a financial statement or not. Those files are in PDF format. So the data is unstructured. So we need to have other bots that take the data out

of the PDFs. And then we have a quality assurance process with humans that make sure that everything is

perfectly classified and structured. We got pretty good on this routine of taking financial data out of PDFs and putting that into tables like JSON or Excel or whatever. So we started even selling this process to banks. Banks receive thousands and thousands of financial statements in PDF formats every day. And it's a complete mess to turn that into actionable data. That's something we do very well.

Actually, haven't seen anyone doing that as well as we do anywhere else.

Why does this... Why do you guys have this DNA? You know, how... I guess, let me ask it differently. How many folks are full time today at Klooks?

40.

And how many are engineers? Seven. Engineer. Okay. Are you an engineer or no? If I

am an engineer, no. I'm I'm I'm graduated in business and administration, but I like code a little bit. Just it's been a necessity to try some stuff and make sure that we can develop some things I envision.

So I guess thousands, let's say you get a data dump thousands of documents. It's... They're in weird languages. You're not sure if they're even financial statements or not. Your your team of seven engineers have built some process to go analyze those. I guess my question to you is, isn't this something that one day sort of AI and and sort of these LLM models take over? In other words, have you built some moat? Have your seven engineers come up with something that OpenAI and these other folks have not thought about where you will always be in need in terms of cleaning that data?

I don't see OpenAI and these folks solving this problem so early, because when you go into PDF tables, there's a lot of things that OCRs don't solve. So if you don't have like a system that goes into PDF and structures that perfectly for the AIs to then take, have an opinion over it. It just doesn't work. The technology needed to solve that is not

the ChatGPT or OpenAI technology, it's the OCR technology, which...

What's that stand for, OCR?

OCR is optical character recognition. It's a system that looks at a picture or a PDF and identify characters. So it turns like an image to text. So that's the thing. There is one specific problem, which is like grammar errors. Sometimes very often systems, they think that a G might be an eight, for instance. And these

confusions are not solved yet, be solved off contextual. And LPs that might understand that in the middle of a sentence of a word, there's not a number. That's something that AI might help. But when you go into tables and when the OCR, there's another problem that is very common. They make confusion in the line. So they might put a value that is for assets in

current assets. And this confusion totally deadly for a risk analysis.

Yeah. That's interesting. Do you have an example like on your desktop or something that you could screen share of some PDF you got that OpenAI could never deal with because it's so confusing, but your OCR technology is able to read it and extract sort of characters and things that we can read in a programmatic way?

Yeah, the key thing is not exactly the OCR because the OCR, we use different OCRs bought in the market. Real difference is the process, the quality assurance process we do over the OCR process. So we have a system of alerts like, hey guys, this total assets is not equal to total liabilities. There might be something wrong here. So our quality assurance teams just goes into and fixes the problem. And we have like more

than 500 alerts that make our data get out of this process completely perfect and validated.

Mhmm. That makes a lot of Tell me a little bit more, Alex, as we wrap up. I mean, it's clear the technology here is powerful. You've got big folks like Bloomberg and CapIQ paying for it. How do you get more customers? You know, you launched this. I remember back in 2022, chatted, you're doing like 200,000 of revenue. 2023, you get up to $7,800,000. This year, you wanna break a million. You're bootstrapped, so you're in full control, which I love, but is there any way you can grow this faster with creative marketing and advertising, but also stay bootstrapped?

Yeah, we have three different products, right? So in our data as a service, I don't see that making the big difference that would change us from a million to a 100,000,000. Where I see that happening is on our service that turns PDFs into structured data. If we do that internationally, if we can find the channels to

go international with this service, I think it's very unique and

the big tech folks, they're not focusing on that right now. So that might be an advantage at the moment for at least a few years. And I see our SaaS as a nice catapult of growth. And that's something we've never invested that much, but we've been having really nice results in lead generation for the last four to five months. And that's making us very happy and excited

about what might come in the future for our SaaS solution.

Mhmm. Well, I'm rooting for you. We'll see what happens. In the meantime, let's wrap up here with the famous five, Alex. Number one, your favorite business book.

Oh, favorite business book? Sapiens? It's not a business But book, it's a book I like.

It works, that works. Number two, is there a CEO you're following or studying?

I'm super cliche on these sort of questions. I would say like Elon Musk at the moment.

Number three, what's your favorite online tool for building Klooks?

I'm still with the answer I gave you last year. I love Google Sheets to prototype stuff.

Number four, how many hours of sleep do you get every night?

Oh, six to eight.

Okay, that's fair. And did you have a birthday? Are you 38 now?

I'm 38, that's right.

Okay, still single with one kiddo?

That's correct.

Awesome. Last question, something you wish you knew when you were 20?

Something... I would... Sorry. Say it again.

Something you wish you knew back when you were 20 years old.

Oh, I would like to be less afraid of making mistakes. I think that would have helped. Mhmm.

Guys, klooks.com.br was launched back in 2021, 2022. He broke $500,000 of revenue in 2023, cleaning a Brazilian financial data and selling it to big firms like Captivate IQ and Bloomberg, which many of you guys might actually be, you know, listening might actually pay for directly. Some of that data is provided by these groups like Klooks. He has specific OCR technology with his seven engineers, but then it's the human layer on top of that where they clean

the data, sort it, filter it, make sure and check for accuracy before they sell it off to the end market. That really is their secret sauce. 50% of his revenue today, and they'll do about a million dollars this year, 50% of that is data as a service. Another 30% is pure services, and then 20% is true software as a service. Completely bootstrapped, selling through two major resellers in Brazil. That's NeoWay and TTR Data, where he takes

a cut of the sales there. If he goes direct, he's selling for 300 to $400 per month and has over 50 customers to date. Alex, thanks for taking us to the top.

Thanks, Nathan. Great pleasure to be with you again.