Our Business Makes $60K/Month Turning Corporate Financial Statements Into Standardized Data

Published: April 23rd, 2023
Alexandre Abu-Jamra
Founder, Klooks
from Porto Alegre, RS, Brasil
started March 2012
market size
avg revenue (monthly)
starting costs
gross margin
time to build
210 days
growth channels
business model
best tools
Google Drive, WordPress, Google ads
time investment
Full time
pros & cons
39 Pros & Cons
3 Tips
Discover what tools Alexandre recommends to grow your business!
Discover what books Alexandre recommends to grow your business!
Want more updates on Klooks? Check out these stories:

Hello! Who are you and what business did you start?

My name is Alexandre Abu-Jamra. I'm a Brazilian tech entrepreneur and I run K-Looks. The business is specialized in organizing unstructured financial data, probably the most unstructured type of data around. Basically, we find and turn corporate financial statements (pdf files that are very messy from a structural point of view) into standardized indexed data with "post-OCR" processing.

We go beyond optical character recognition algorithms (OCRs) to solve this problem and deliver accurate, fully customized data solutions (format, chart of accounts, etc.).

That's quite a pain, especially in credit departments of banks, insurance companies, and large corporations. Our process is a combination of engines that extracts and categorizes data automatically with human quality assurance. The process delivers data of higher quality than that generated by any highly trained credit analyst with record TAT and scalability. Seamless integration, and multi-language offering: Portuguese, English, and Spanish. Give me a month and I can do German as well!

We service some of the main Brazilian financial institutions such as Caixa, Citi, and ABC. Also, we are the official data provider to some of the largest global financial data players. Our revenues in 2022 were $600k with triple-digit annual growth in the last 3 years.


What's your backstory and how did you come up with the idea?

In my previous job, I was an M&A advisor and I had constant pain - in finding financial data of private companies. Revenues, EBITDA, growth, debt-to-equity, Net debt/EBITDA, etc. Such info was available mostly for public companies or selected industries. For the broader company population, the data was hidden inside hard-to-find-thousand-page-gazette-like-pdf-files.

So interns gathered and classified the data most inefficiently: printscreen the gazette page and name the file with the corporate tax registry and company name. A full-time equivalent would only scratch the surface of the data pool, not to mention finding the right sources in the first place. When I quit the M&A advisory that problem stuck in my head: there had to be a more efficient way of doing that. The entire finance industry had the same problem - private equities, VCs, etc.

At the start of a business, the easier way to compete is customizing - so you can be more nimble and adaptable.

I found two partners that saw the same opportunity. Our first challenge was to find data sources and catalog pdf files. The second goal was to identify the financial statements in the files and extract the financial data in a standardized way. Sounds simple but it’s a herculean tech challenge.

There was no consolidated source, so we had to develop around 20 bots just for starters. OCRs ('optical character recognition' algorithms) didn't work, so we had to combine spreading automation with a human-based quality assurance process. Nowadays OCRs work better but still not well enough, especially in lower-resolution images or on coffee-stained paper - yes, we receive all sorts of things.

Guess what? We did it. Large database, indexed data. But we had a total of zero committed prospects. Business development was new to us - not recommended! Piece of advice: always try to get a prospect committed to your product before putting up with all that struggle!

So we had to get out of the building.

We sold the dataset to one client or another on a very low ticket. Burn rate to the sky and no VC money onboard. And there came our first pivoting: our target client. Instead of going after M&A advisors or private equity firms - a fragmented client portfolio with very high UX needs - we targeted financial data aggregators. Not a very scalable channel since there were 3 or 4 prospects out there. But the ticket could fund the company in the short term and allow us to breathe while we thought about growth.

So I cold called Capital IQ. Yes, "the" Capital IQ - the largest private financial data aggregator on the planet. And I somehow reached their local director: "Alex, that's wonderful what you've done, I'm very interested but right now our R&D is Asia-oriented. I will get back to you when we get to the South American cycle". I thought "Alright, he is never coming back, let's try selling hot dogs".

6 months later I received a call "Hey Alex, here is Pedro from Capital IQ, not sure you remember me". One month later we were meeting with one of McgrawHill'sl heirs (Mcgraw Hill owned S&P which owned Capital IQ). At that time I realized I had something relevant in my hands. In my mind that was Day 1 for K-Looks.

After some years we mastered our internal financial data structuring process. So we started to offer financial spreading services - give me a raw financial statement, and you will see it in your system in a chart of accounts customized for you in a couple of hours - to banks and insurance companies. This service offering accounts for almost 40% of our revenues today.

Before creating features that YOU think will make a difference, it might be more productive to sell them before developing and make sure someone is willing to pay for them.

Take us through the process of building the first version of your product.

To be honest we didn't think much about design. That hurt us then and it still does. We blindly thought that we knew what the market wanted. Our attention was directed to creating web-crawling robots and structuring the data indexing process. Our energy was directed to cleaning the data, not to generating value out of it. What we had in mind was that we had to deliver data, no matter what format. Could be through a web platform, a CSV file, a SQL database, smoke signs, or sign language.

The first challenges were regarding crawlers' navigation and stability. Not an easy task but no rocket science either. The real challenge was when we got to "extracting the data out of the pdfs".

No OCR algorithm in the world could solve the problem fully, and mismatches between lines and figures were the norm. Maybe not a big deal in some applications but a deal breaker in credit analysis where a wrong comma can mean a million-dollar credit mistake.

We had to set up a quality assurance team. And there came the challenge inside the challenge: making that customizable, error-proof, and scalable. The first versions of the QA process were done in Google Sheets, where we could adjust the process quickly. It was flexible and extremely customizable and we could iterate every minute. We got to a point where we had more than a hundred automatic quality tests without a dev hour spent. That made all the difference. Before we built our QA system we matured it in Google Sheets.

Since launch, what has worked to attract and retain customers?

To attract prospects to our SaaS platform: good content through newsletters. It's very clear when we hit it right. Good quality leads start to pop up and in a few weeks, we have new contracts. And for us, good content usually means a client business case using our solution or demonstrating in some way the market intelligence we can generate with our data.

You might think "what about social media, Instagram, Linkedin,or Google?". Well, they have huge potential but so far we didn't find the right fit. Tried different ways. Super directed audiences, broad ones, with banners, with text, with different sorts of content. Results were not relevant nor conclusive until now.

On the other hand, our newsletter is often forwarded, and new prospects come to the mailing list. And they bring good leads when the content is good. It's a slower growth path than traditional advertising but it has been more efficient. We have also seen a difference in client retention depending on where it came from.

With this product, we've experienced quite relevant churn rates. Our remediation strategy has been similar to what we have used to win customers: we combine good content with usage monitoring.

Every time a client goes low on usage, a yellow flag shines bright. And when that happens we build customized content for that client. How do we do that? We've classified our clients into 15 different personas and 46 different use cases and built nearly 150 "standard content drops" (growing).

We classify our clients and send content that shows how our solution can be more successful. For example, if the client is an M&A advisor we might send a report of a company that has good M&A potential ("Hey, look at his company, they make online games, have $20m EBITDA, and grew 1356% last year").

That strategy has worked for us. Our churn is not zero but is much lower now.

Other services (Financial Statements Spreading Service and Data as a Service) are word of mouth and outbound sales. Also, in those cases, client retention is not something we put much energy into since churn rates are irrelevant as long as we provide quality service.

How are you doing today and what does the future look like?

We've been profitable and cash positive since our first large client - what we call Day 1. Our main challenge is to accelerate growth.

We've come from 3 years of 100% YoY growth, and sustaining it might be challenging. Since our main contracts come through outbound sales, we need to improve our network in the US and EMEA, the largest markets for our data-spreading service.

Through starting the business, have you learned anything particularly helpful or advantageous?

I think the main lesson I've learned is that demand comes before production. What does that mean?Itt means "sell first, deliver later".

All clients can wait a few weeks or even months to have something if they reed it. If timing is a problem in a negotiation, then timing is not rhe problem, the solution might not be that relevant. It is important to offer what you can deliver.

But before creating features that YOU think will make a difference, it might be more productive to sell them before developing and making sure someone is willing to pay for them. This approach will greatly reduce the classic risk of spending months or years building features that you don't know they are going to sell. It's a tricky exercise but if you think out of the box usually there is a way to make it work.

What platform/tools do you use for your business?

Well, I'm a Google Sheets geek. I've built all our operating systems using it. All our controls are customized in Google Sheets systems and we don't see any need to change that. I am a strong believer that mainstream, off-the-shelf operating systems are much worse than a fully customized system, and with Google Sheets ,you can do much more than you think.

For example, all our Customer Success alerts and drops logic is programmed inside a sheet. Would I have something similar with SAP or Oracle without spending a 7-digit figure? I don't believe so.

What have been the most influential books, podcasts, or other resources?

I like to have a look now and then at Startup Owners Manual, and loved The Founder's Dilemmas, Traction, and Four Steps to Epiphany.

Regardingnonbusinesss literature, I like to read about history, human evolution, and biographies.

Advice for other entrepreneurs who want to get started or are just starting out?

I think I have two main takeaways. One I already mentioned: demand comes before production. Besides that, I see a lot of people building solutions hard to customize, trying to make their clients fit in their solutions, instead of the other way around. I know, VCs love that. "More scalable,e" they say.

Each client has its singularities, even though they might have similar needs. Especially at the start of a business, the easier way to compete is customized - you can be more nimble and adaptable. And I believe that mass customization in software is so accessible that you have a good probability of scaling and customizing at the same time with great cost efficiency. In our case at least, that's what happens. Right now we have a growth capacity of approx. 100% monthly with extremely customized deliveries.

Are you looking to hire for certain positions right now?

As I've mentioned, we have huge potential to expand in North America and EMEA, and we don't have any footprint in those regions. We are looking for Sales Reps or even an operational business partner in those areas with connections in the financial industry, especially with credit analysis teams.

Where can we go to learn more?

If you have any questions or comments, drop a comment below!

Want to start a data analysis service? Learn more ➜