Background

Any business that depends on accurate web analytics to drive revenue, whether that’s affiliate marketing, lead generation, performance media, or ecommerce, faces the same fundamental question: can you trust the data you’re making decisions with? For companies operating in high-bot, privacy-sensitive, or ad-blocker-heavy environments, the answer with Google Analytics is increasingly no.

Find.co runs a portfolio of affiliate websites in iGaming and cryptocurrency, two of the web’s most data-intensive, bot-heavy, and heavily regulated sectors. The company’s business model centers on driving page views, generating click-outs to partner brands, and converting those clicks into registrations, deposits, and revenue. 

For Find.co, every interaction matters, and every missed or misattributed click has a direct commercial cost. 

But for years, the company relied on Google Analytics to track all of this. Over time, the analytics team developed a nagging suspicion that GA wasn’t telling them the whole story. And it turned out, they were right. 
‍

Challenge: The Black Box Problem

Find.co’s Senior Data Analyst Theodore Sorokine knew something was off before he could prove it. Page view numbers didn’t add up. Click-out volumes felt low. But with Google Analytics operating as a black box, there was no way to know what was real and what was being hidden.

 This wasn't just GA’s well-known sampling problem, where large datasets get approximated rather than fully processed. GA was silently filtering out traffic it deemed suspicious entirely. It was removing it without flagging it, without explaining why, and with no way for the team to inspect or challenge those decisions. 

‍

“We didn’t see the data that was being filtered out. When we started running Snowplow in parallel to GA, we immediately saw the delta, and how many hits were hidden from us.”

Theodore Sorokine, Senior Data Analyst, Find.co

‍

This isn’t a problem unique to iGaming. Any organization where revenue is tied to digital interactions, ie:comparison sites, marketplaces, fintech platforms, lead generation businesses, faces the same risk: a third-party analytics tool making opaque decisions about which data you get to see and which data quietly disappears.

In the iGaming and crypto space, data analytics is becoming increasingly challenging. Bots are rampant. Ad-blocker usage is sky-high. And GDPR compliance is essential. GA was falling short for Find.co on all three fronts. It was filtering legitimate traffic alongside bots, missing events blocked by browser extensions, and raising serious questions about whether sending behavioral data to Google’s servers was even compliant under European privacy law. 

Scott Cattell, Find.co’s Director of Analytics, ran the numbers on the compliance risk alone and the business case wrote itself.

‍

“When you put the cost of switching tools against a potential GDPR penalty — 20 million euros or 4% of global turnover, whichever is larger, the business case writes itself. The budget question wasn't even worth debating.”

Scott Cattell, Director of Analytics, Find.co

‍

The GDPR calculus applies to any company with European users, regardless of industry. Potential fines of up to 4% of global annual turnover or €20 million, whichever is greater, make the cost of a compliant analytics platform look trivial by comparison.

This prompted the Find.co team to weigh up three options: stick with GA and hope for the best, hire a full-time JavaScript engineer to build a custom tracking solution from scratch, or adopt Snowplow. The decision became immediately clear. Snowplow’s managed service cost significantly less than a single engineering hire. Plus, it came with a breadth of features, data models, and ongoing development that no in-house build could match. 

Solution: Opening the Box

Find.co replaced Google Analytics with Snowplow’s managed Customer Data Infrastructure, and the difference was immediate.
‍

Total Event-Level Visibility

Where GA silently discarded data, Snowplow shows everything. The Find.co team can now inspect every event at the point of collection, including failed events and why they failed. Nothing gets hidden. Nothing gets quietly removed.
‍

“We now have visibility on every hit and every event. We see before the collection, at the point of the collection. We even see and can inspect our failed events and why they failed. Those are not hidden from us.”

Theodore Sorokine, Senior Data Analyst, Find.co

Privacy by Design

Snowplow’s architecture lets Find.co collect rich behavioral data while staying fully GDPR compliant. IP addresses and user IDs are hashed at the point of collection, and none of the data leaves Find.co’s own infrastructure. All the data flows into Google BigQuery, using EU multi-region instances to keep data within European borders. 

For a business operating across European gambling and crypto jurisdictions, that level of control isn’t a nice-to-have, it’s essential. The same applies to any business subject to GDPR, CCPA, or similar privacy frameworks. From fintech and healthtech to media and ecommerce.

‍

“With Snowplow, all IPs are fully hashed. All cookie IDs are fully hashed. And crucially, nothing we collect goes to Google, not for AI training, not for ad targeting, not for anything.”

Theodore Sorokine, Senior Data Analyst, Find.co

Beating the Ad Blockers

Find.co’s audience runs ad blockers at unusually high rates. This is a challenge for businesses whose users tend to be technical, privacy-conscious, or younger. It affects companies across developer tools, fintech, VPNs, and security products, and increasingly, even mainstream consumer brands.

Snowplow’s support for multiple tracking methods gave the team a way around the problem. By running JavaScript tracking alongside pixel image tracking in parallel, the team could finally quantify the gap. And it was bigger than anyone expected.

‍

“For every 100 events GA recorded, Snowplow's JavaScript tracker captured 130. Add pixel tracking on top, which bypasses ad blockers entirely, and that number climbs to 160.”

Theodore Sorokine, Senior Data Analyst, Find.co

Zero DevOps, Zero Hassle

Theodore had previously run Snowplow open source at a previous company, which meant managing servers, monitoring uptime, and keeping DevOps resources on hand. When he set up Snowplow’s managed service for Find.co, all of that disappeared. ‘It was 100% time saved,’ he says. ‘I didn’t touch a terminal window for this setup.’ 
‍

Results: Data You Can Actually Trust

The migration transformed how Find.co operates. The team went from second-guessing every report to making decisions with real confidence.

Clean Bot Detection and Staff Filtering

With full event-level data, Find.co now flags and filters suspicious traffic on its own terms. A recent A/B test surfaced a perfect example whereby the team spotted strange behaviour in their test data, traced it back to staff members running internal tests, and deployed a Snowplow-powered JavaScript tag that let employees self-identify. With GA, they’d never have known.

Smarter Partner Optimization

The completeness of Snowplow’s data has sharpened decision-making across the board. The team now uses it to inform partner link ordering, page prioritization, and A/B testing, all backed by the assurance that they’re working with a complete dataset, not a filtered approximation.

“When we look at reports now, the confidence is much higher. Which brands to prioritize, which pages to optimize, those decisions are much sharper when you know you're looking at the complete picture.”

Theodore Sorokine, Senior Data Analyst, Find.co

Measurable Business Impact

With Snowplow, Find.co now has access to rich contextual data. It can capture information on things like user agents, device technology details, and more accurate geolocation. This led to a critical discovery: the team found it was serving partner listings to users in countries where those partners don't operate, resulting in dead-end click-throughs and wasted traffic. This became a top-priority fix, and it only came to light because Snowplow captured what GA never could. 

Beyond that, the accuracy and completeness of Snowplow's data has driven measurable improvements across Find.co's core performance metrics. 

Year-on-year conversion rates have improved significantly since the switch. Across all sites, conversion rates are up 1.33x, with the top performing site up 2.3x and the second best up 1.19x. Multiple factors contribute to performance improvements of this kind, but having complete trustworthy data as the foundation for every decision has been central to the gains. 

Looking Ahead

Find.co’s future roadmap builds on Snowplow as the foundation. The team is perfecting the interplay between JavaScript and pixel tracking for even more comprehensive capture, while expanding A/B testing with the confidence that experiments sit on clean, complete data.

The biggest initiative is geo-aware partner personalization. Armed with the geo data that exposed the dead-end partner problem, Find.co is building real-time logic to dynamically reorder and filter partner listings by location, so every click leads somewhere useful. It’s the kind of use case that was architecturally impossible under GA.

The team is also exploring Snowplow’s dbt data model packages for user aggregations and deduplication, which they expect will save significant engineering effort.
‍

“We’re investigating Snowplow’s dbt packages for user aggregations and deduplication. They’ll save us huge amounts of time because we can just take the package Snowplow offers and modify it for our own use cases.”

Scott Cattell, Director of Analytics, Find.co

‍

For Find.co, there’s also a newer challenge on the horizon: AI agent traffic. The company has already created a dedicated attribution channel for AI-referred visits, using referral data and user agent strings to separate agent-generated sessions from human ones. But as more agents fail to identify themselves, the classification problem is getting harder. 

“AI traffic is its own attribution channel,” Theodore notes. “You have SEO traffic, social traffic, and then you have AI. The question is how you categorize the agents that aren’t marking themselves.” This is an area where Find.co is looking to Snowplow’s bot and agent intelligence capabilities to go further. 

And as Find.co builds out its marketing team and moves beyond a purely SEO-driven acquisition model, Snowplow will provide the multi-touch attribution infrastructure to match, giving the company the analytical foundation to scale its growth as aggressively as its ambitions.