StarZero All shows Try the podcast clipper →

Feb 29, 2024 · 18m · top-founders

He Did $1m last year selling South American Business Data to firms like CapIQ

Alex Abujamra · 10m spoken Nathan Latka · 6m spoken
0:00 / 0:00

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

In this episode of Conversations with Nathan Latka, Klooks founder and CEO Alex Abujamra discusses how his bootstrapped company reached $1 million in revenue by extracting, cleaning, and selling unstructured private Brazilian corporate financial data to global intelligence firms like Bloomberg and S&P Capital IQ. Abujamra details their diversified DaaS and SaaS business model, proprietary OCR extraction pipeline with 500 accounting validation checks, and capital-efficient reseller distribution strategy.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Nathan holds 38.8% of the talking time here. How this is scored →

Nathan as informed peer 4.5 Guest teaching 3.5 Guest disagreement 1.8 Nathan pushing back 3.0
05100:0010:000:38–2:51 · Nathan as informed peer 5/10 Episode Overview: Klooks Business Model and Metrics Latka opens by recalling historical metrics and revenue breakdown from their prior interview, showing familiarity with the business model. Abujamra collaboratively provides updated ARR targets and revenue proportions.2:52–5:24 · Nathan as informed peer 4/10 Enterprise Clients and Potential M&A Positioning Latka floats an M&A hypothesis suggesting CapIQ or Bloomberg would buy Klooks to block competitors. Abujamra counters this premise, explaining that Brazilian data represents too marginal of a market to hit enterprise M&A thresholds at Klooks' current scale.5:24–9:04 · Nathan as informed peer 6/10 Distribution Channels and Value-Added Resellers Latka drills into channel partner economics, pressing Abujamra on contract terms to prevent price cannibalization between VARs and direct sales. Latka also corrects the specific domain name of reseller TTR Data.9:04–11:42 · Nathan as informed peer 4/10 Data Extraction Pipeline and Team Structure Latka summarizes his understanding of data gathering, which Abujamra expands upon by detailing the intricate crawler, PDF parsing, and human QA pipeline required to extract Brazilian corporate filings.11:42–14:59 · Nathan as informed peer 4/10 OCR Limitations, AI Comparison, and Quality Assurance Latka challenges the defensibility of the business against LLMs like OpenAI. Abujamra thoroughly breaks down why LLMs fail at tabular OCR parsing, detailing table alignment errors and Klooks' proprietary 500-rule QA engine.15:01–17:43 · Nathan as informed peer 4/10 Growth Opportunities: International Expansion and SaaS Latka explores expansion opportunities within a bootstrapped framework before transitioning into the standard Famous Five rapid-fire questions, which Abujamra answers candidly.0:38–2:51 · Guest teaching 1/10 Episode Overview: Klooks Business Model and Metrics Latka opens by recalling historical metrics and revenue breakdown from their prior interview, showing familiarity with the business model. Abujamra collaboratively provides updated ARR targets and revenue proportions.2:52–5:24 · Guest teaching 5/10 Enterprise Clients and Potential M&A Positioning Latka floats an M&A hypothesis suggesting CapIQ or Bloomberg would buy Klooks to block competitors. Abujamra counters this premise, explaining that Brazilian data represents too marginal of a market to hit enterprise M&A thresholds at Klooks' current scale.5:24–9:04 · Guest teaching 2/10 Distribution Channels and Value-Added Resellers Latka drills into channel partner economics, pressing Abujamra on contract terms to prevent price cannibalization between VARs and direct sales. Latka also corrects the specific domain name of reseller TTR Data.9:04–11:42 · Guest teaching 5/10 Data Extraction Pipeline and Team Structure Latka summarizes his understanding of data gathering, which Abujamra expands upon by detailing the intricate crawler, PDF parsing, and human QA pipeline required to extract Brazilian corporate filings.11:42–14:59 · Guest teaching 7/10 OCR Limitations, AI Comparison, and Quality Assurance Latka challenges the defensibility of the business against LLMs like OpenAI. Abujamra thoroughly breaks down why LLMs fail at tabular OCR parsing, detailing table alignment errors and Klooks' proprietary 500-rule QA engine.15:01–17:43 · Guest teaching 1/10 Growth Opportunities: International Expansion and SaaS Latka explores expansion opportunities within a bootstrapped framework before transitioning into the standard Famous Five rapid-fire questions, which Abujamra answers candidly.0:38–2:51 · Guest disagreement 1/10 Episode Overview: Klooks Business Model and Metrics Latka opens by recalling historical metrics and revenue breakdown from their prior interview, showing familiarity with the business model. Abujamra collaboratively provides updated ARR targets and revenue proportions.2:52–5:24 · Guest disagreement 3/10 Enterprise Clients and Potential M&A Positioning Latka floats an M&A hypothesis suggesting CapIQ or Bloomberg would buy Klooks to block competitors. Abujamra counters this premise, explaining that Brazilian data represents too marginal of a market to hit enterprise M&A thresholds at Klooks' current scale.5:24–9:04 · Guest disagreement 2/10 Distribution Channels and Value-Added Resellers Latka drills into channel partner economics, pressing Abujamra on contract terms to prevent price cannibalization between VARs and direct sales. Latka also corrects the specific domain name of reseller TTR Data.9:04–11:42 · Guest disagreement 1/10 Data Extraction Pipeline and Team Structure Latka summarizes his understanding of data gathering, which Abujamra expands upon by detailing the intricate crawler, PDF parsing, and human QA pipeline required to extract Brazilian corporate filings.11:42–14:59 · Guest disagreement 3/10 OCR Limitations, AI Comparison, and Quality Assurance Latka challenges the defensibility of the business against LLMs like OpenAI. Abujamra thoroughly breaks down why LLMs fail at tabular OCR parsing, detailing table alignment errors and Klooks' proprietary 500-rule QA engine.15:01–17:43 · Guest disagreement 1/10 Growth Opportunities: International Expansion and SaaS Latka explores expansion opportunities within a bootstrapped framework before transitioning into the standard Famous Five rapid-fire questions, which Abujamra answers candidly.0:38–2:51 · Nathan pushing back 2/10 Episode Overview: Klooks Business Model and Metrics Latka opens by recalling historical metrics and revenue breakdown from their prior interview, showing familiarity with the business model. Abujamra collaboratively provides updated ARR targets and revenue proportions.2:52–5:24 · Nathan pushing back 3/10 Enterprise Clients and Potential M&A Positioning Latka floats an M&A hypothesis suggesting CapIQ or Bloomberg would buy Klooks to block competitors. Abujamra counters this premise, explaining that Brazilian data represents too marginal of a market to hit enterprise M&A thresholds at Klooks' current scale.5:24–9:04 · Nathan pushing back 6/10 Distribution Channels and Value-Added Resellers Latka drills into channel partner economics, pressing Abujamra on contract terms to prevent price cannibalization between VARs and direct sales. Latka also corrects the specific domain name of reseller TTR Data.9:04–11:42 · Nathan pushing back 2/10 Data Extraction Pipeline and Team Structure Latka summarizes his understanding of data gathering, which Abujamra expands upon by detailing the intricate crawler, PDF parsing, and human QA pipeline required to extract Brazilian corporate filings.11:42–14:59 · Nathan pushing back 4/10 OCR Limitations, AI Comparison, and Quality Assurance Latka challenges the defensibility of the business against LLMs like OpenAI. Abujamra thoroughly breaks down why LLMs fail at tabular OCR parsing, detailing table alignment errors and Klooks' proprietary 500-rule QA engine.15:01–17:43 · Nathan pushing back 1/10 Growth Opportunities: International Expansion and SaaS Latka explores expansion opportunities within a bootstrapped framework before transitioning into the standard Famous Five rapid-fire questions, which Abujamra answers candidly.

speaking balance: gold is Nathan, purple is the guest (3 minute bins)

0:00 · Nathan 80% · guest 20%0:00 · Nathan 80% · guest 20%3:00 · Nathan 17.2% · guest 82.8%3:00 · Nathan 17.2% · guest 82.8%6:00 · Nathan 34% · guest 66%6:00 · Nathan 34% · guest 66%9:00 · Nathan 32.4% · guest 67.6%9:00 · Nathan 32.4% · guest 67.6%12:00 · Nathan 17% · guest 83%12:00 · Nathan 17% · guest 83%15:00 · Nathan 40% · guest 60%15:00 · Nathan 40% · guest 60%18:00 · Nathan 95.5% · guest 4.5%18:00 · Nathan 95.5% · guest 4.5%
Sharpest disagreement ▶ 4:20 Rebuttal of M&A Target Logic

Abujamra directly rejects Latka's premise that Bloomberg or CapIQ would rush to acquire Klooks to lock out rivals, citing the marginal scale of Brazilian financial data in global portfolios.

 Hardest push from Nathan ▶ 7:20 Pushing on Partner Cannibalization

Latka challenges Abujamra on how he legally prevents value-added resellers like Neoway from undercutting direct sales prices.

 Biggest teaching moment ▶ 12:11 Explaining OCR Limitations over AI

Abujamra educates Latka on why modern LLMs cannot easily replace OCR pipelines due to character misreads and table misalignment issues in unformatted PDFs.

 Nathan holds their own ▶ 8:42 Domain Correction and Channel Synthesis

Latka demonstrates sharp research by immediately verifying and correcting the exact domain name for partner TTR Data while mapping out Klooks' distribution mix.
the scores for every segment, with the reasoning behind each
ChapterTopicNathan as informed peerGuest teachingGuest disagreementNathan pushing backWhy
Episode Overview: Klooks Business Model and Metrics 5112 Latka opens by recalling historical metrics and revenue breakdown from their prior interview, showing familiarity with the business model. Abujamra collaboratively provides updated ARR targets and revenue proportions.
Enterprise Clients and Potential M&A Positioning 4533 Latka floats an M&A hypothesis suggesting CapIQ or Bloomberg would buy Klooks to block competitors. Abujamra counters this premise, explaining that Brazilian data represents too marginal of a market to hit enterprise M&A thresholds at Klooks' current scale.
Distribution Channels and Value-Added Resellers 6226 Latka drills into channel partner economics, pressing Abujamra on contract terms to prevent price cannibalization between VARs and direct sales. Latka also corrects the specific domain name of reseller TTR Data.
Data Extraction Pipeline and Team Structure 4512 Latka summarizes his understanding of data gathering, which Abujamra expands upon by detailing the intricate crawler, PDF parsing, and human QA pipeline required to extract Brazilian corporate filings.
OCR Limitations, AI Comparison, and Quality Assurance 4734 Latka challenges the defensibility of the business against LLMs like OpenAI. Abujamra thoroughly breaks down why LLMs fail at tabular OCR parsing, detailing table alignment errors and Klooks' proprietary 500-rule QA engine.
Growth Opportunities: International Expansion and SaaS 4111 Latka explores expansion opportunities within a bootstrapped framework before transitioning into the standard Famous Five rapid-fire questions, which Abujamra answers candidly.

Statements from this episode (11)

Disclosure
Abujamra: Klooks revenue is 50% DaaS, 30% services, and 20% SaaS
“Right now we are on around 50% of data service 30% on, on service itself, and 20% on, on SaaS, but everything has grown from there.”
Alex Abujamra Feb 29, 2024 ▶ 2:15
Prediction Not checkable as stated
Abujamra: Klooks will probably reach $1M in revenue in 2024
“We've had some interesting growth last year, kind of 50% growth, and this year we are probably reaching one million dollars in revenues.”
Alex Abujamra Feb 29, 2024 ▶ 2:30
Assertion Not checkable as stated
Abujamra: Klooks finished 2023 with around $800K in revenue
“It was kind of 800,000. Yeah, that was it.”
Alex Abujamra Feb 29, 2024 ▶ 2:48
Disclosure
Abujamra: Klooks sold private Brazilian financial data to Bloomberg, Capital IQ, and Moody's
“Originally the data we would sell to other intelligence platforms. So we would sell to Bloomberg, to Capital IQ Moody's analytics. So these guys would get our data, which is like financial data of private Brazilian companies and just put on their products unde…”
Alex Abujamra Feb 29, 2024 ▶ 3:00
Disclosure
Abujamra: Klooks is prioritizing direct data sales to banks over intelligence platforms
“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, that's part of our priorities at the moment.”
Alex Abujamra Feb 29, 2024 ▶ 3:24
Prediction Not checkable as stated
Abujamra: Major intelligence platforms may consider acquiring Klooks at $5M–$10M revenue
“Maybe once we reach, like, five to ten million dollars in revenues, that might make sense, because our process is really unique.”
Alex Abujamra Feb 29, 2024 ▶ 4:50
Disclosure
Abujamra: Klooks has 50-60 direct and 500-600 indirect customers
“Oh, directly might be 50 or 60, and indirectly might be 506 hundred.”
Alex Abujamra Feb 29, 2024 ▶ 5:28
Disclosure
Abujamra: Klooks makes $300-$400 direct vs. ~$100 monthly via resellers
“I would sell directly for roughly Three to 400 dollars monthly, and they would sell me to, like, a hundred dollars monthly.”
Alex Abujamra Feb 29, 2024 ▶ 6:55
Assertion Not checkable as stated
Abujamra: Klooks is the only firm investing in deep financial crawlers
“And we're the only ones that are investing heavily on, on crawlers too, too hard to find sources.”
Alex Abujamra Feb 29, 2024 ▶ 9:43
Prediction Not checkable as stated
Abujamra: OpenAI will not solve complex PDF table extraction soon
“I don't see open AI and these folks solving this problem so early because when you go into PDF tables, it, there's a lot of things that OCRs don't solve.”
Alex Abujamra Feb 29, 2024 ▶ 12:12
Opinion
Abujamra: Scaling to $100M requires international document extraction, not DaaS
“In our data as a service, I don't see that making the big difference that would change us from a million to a hundred million. Where I see that happening is, is on our service that turns PDFs into structured data. If we do that internationally, if we can find …”
Alex Abujamra Feb 29, 2024 ▶ 15:29
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 2,600 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more