Marketing Analytics with Fexingo

The episode delivers a focused, technically grounded topic with several non-obvious sub-points (click ID stripping by misconfigured redirects, iOS Safari-specific UTM drops, redirect chain latency destroying conversions). However, the central thesis - broken links corrupt attribution - is already known to experienced analysts, and the runtime is short with some padding near the end.

“I've seen campaigns where the discrepancy was over 20 percent between what the ad platform reported and what the internal analytics showed”

“one retailer I worked with found that their mobile redirect dropped the UTM parameters only on iOS Safari”

The reframe of attribution error as a data-quality problem rather than a model-choice problem is a legitimate contrarian nudge, but the underlying content is well-trodden SEO/analytics hygiene. There are no first-principles arguments or genuinely surprising conceptual leaps - the episode competently synthesises known issues rather than advancing new thinking.

“A lot of attribution bias gets blamed on model choice - last touch versus multi-touch versus algorithmic. But the underlying data quality is often the real culprit”

“The irony is that teams spend thousands on attribution software and consultants, but they won't spend an hour a month checking their own links”

There is no external guest - this is a two-host format where Lucas acts as the practitioner and Luna as the questioner. Lucas references real campaign work, but listeners receive no context on his background, seniority, or track record, and the format inherently lacks the independent credibility a qualified outside practitioner would bring.

“In my experience, maybe one in twenty marketing teams has a regular URL audit”

“I was looking at a mid-size retail campaign from late 2025 - about 2 million dollars in ad spend across paid social, search, and email”

The episode is anchored by concrete figures throughout: a named $2M campaign, 12% broken-URL conversion rate, 4% hard-404 click rate, a 20% platform-vs-analytics discrepancy, 8% iOS conversion loss, and a cited Google statistic on load-time bounce rates. Named tools (Screaming Frog, DeepCrawl) and specific platform parameters (fbclid, gclid) add further grounding, though the campaign data cannot be independently verified.

“roughly 12 percent of their tracked conversions came through URLs that had broken at some point in the customer journey”

“Google's own data shows that as page load time goes from one second to three seconds, the probability of bounce increases by 32 percent”

Luna functions as a competent foil who surfaces useful follow-up angles - the email channel, redirect-chain latency, and mobile testing - and her early scepticism ('That seems almost too simple to be a big deal') creates a productive setup. However, questions are largely predictable and no claim goes genuinely challenged or probed; the conversation is a structured explainer rather than a rigorous dialogue.

“That's a direct, measurable budget decision based on bad data. And the team probably never knew the root cause was a broken link”

“That's the kind of thing that would never show up in a standard QA check. Most teams just click the link on their desktop once and call it done”