.jpg)
Strix is a security and tracking technology company operating in Argentina, Chile, and Uruguay. Its offering spans vehicle protection, home alarm systems, personal security, and corporate fleet management, all centralized in a single mobile app.
With a 93% success rate in vehicle recovery, it competes in a segment where product urgency drives the consumer's decision speed.
The context
When we started working with Strix, its paid media operation had a speed problem the team had already come to accept as the cost of operating at scale: weekly reports consumed one to two full workdays, creative performance analyses took two hours each, and account audits required six hours to complete. The time left for business decisions was whatever remained after consolidating the data.
On top of that, the picture got more complex because Strix operates across multiple markets with different data sources: media platforms, tracking tools, commercial data. Every reporting cycle started with figuring out where the information even was before it could be interpreted.
The challenge
- How do we free up analytical capacity so the team can focus on business problems instead of data problems?
- How do we validate whether activating programmatic campaigns on DV360 was having a real impact on Search, in an environment where conventional attribution models don't capture that kind of relationship?
The approach
The first decision was redefining which tasks needed to disappear from the human workflow and which ones needed to be improved. We layered in an agentic system connected to the media platforms, GA4, Search Console, and the commercial data source, respecting the client's business logic. It centralizes and structures the data and produces a first layer of interpretation.
But the agentic operation didn't replace human judgment on the decisions that mattered.
The programmatic case is a good illustration of this. We activated DV360 campaigns (DKS-Awareness and DKS-Consideration) on May 17 and ran a comparative analysis of 17 days PRE vs. 17 days POST. GA4's attribution model showed cross-channel DV360→Search assistance below 2% across multichannel paths, which on its own would suggest programmatic wasn't affecting Search.
However, the agentic analysis, cross-referencing Search Console, GA4, and search term data, pointed to something else: impressions down 13% with clicks up 63% (CTR 1.87%→3.51%). Before attributing this to priming, we ruled out keyword mix, auction dynamics, seasonality, and events. With those alternatives controlled for, priming turned out to be the most consistent explanation.
The priming hypothesis — users exposed to programmatic ads who then search with higher intent — held up in the data, even though the attribution model didn't capture it directly. Generic Search campaign CPA improved 28.6% in the POST period.
That insight wouldn't have surfaced under the previous operating cycle. The volume of cross-referencing needed to reach that conclusion, pulling data from four different sources, would have taken several days of dedicated analytical work.
The results
With the centralized system in place, data started reaching the team already structured and with an initial layer of interpretation.
- What used to take 120 minutes of daily/weekly analysis now takes 30 minutes.
- Weekly reports went from 180 minutes to 30 minutes.
- Creative analysis dropped from 120 to 15 minutes. Account audits, which used to take 360 minutes, are now down to 60.
- In the post-DV360-activation period, total search clicks grew 63% and CTR went from 1.87% to 3.51% (+88%).
- Generic Search campaign CPA improved 28.6%, dropping from 10,456 ARS to 7,469 ARS per conversion.
- Sessions generated by DV360 posted a 99.4% engagement rate, a sign that the incremental traffic was genuinely high quality.
-
Operationally, total time saved per task landed at 78.4%, freeing up 63.42 hours a month for the team. With that 78% of reclaimed time, the team redesigned the strategy, which translated into +63% clicks and -28% CPA. Time saved shifted from being an input to being an outcome.
Operational efficiency by process
.png)
.png)
The business takeaway
The priming finding in this case — Search CTR nearly doubling on the same impression volume, generic CPA improving 28.6% — wasn't sitting in any dashboard. It surfaced by cross-referencing four data sources in an analysis that, under the previous operating model, wouldn't have happened within the reporting cycle. That's the underlying pattern: freeing up analytical capacity to spot relationships that conventional attribution models miss.
In Strix's case, the consumer's decision structure matters. Vehicle security is an urgency category: someone searching "car GPS tracker" after seeing a Strix ad has a short intent window and a high likelihood of clicking. That explains the size of the effect.
In categories with longer decision cycles — high-ticket products, B2B services, extended-comparison categories — the same mechanism can show up at a different speed and in a different way. The priming hypothesis still holds; measurement timing and control metrics need to be adjusted to match the segment's decision pace.
What this case left us with is a cross-reading methodology applicable to any media activation where direct attribution falls short of showing the real impact.
.png)
