How I Spent Three Years Mapping CDS Operations (And What Actually Works)
I didn't set out to study cartels. I was a customs analyst in Matamoros doing routine freight inspections when I noticed something odd about the paperwork. A shipping container from Guadalajara would arrive with one set of bills of lading, but the driver would present a different manifest at the gate. Not dramatically different — close enough to pass casual scrutiny, wrong enough to raise flags if you actually knew how to read it. That was my introduction to the CDS shadow network. The full title people throw around now — The CDS Cartel Unraveling The Shadow Network Behind Mexico's Most Dangerous Drug Empire — sounds like a documentary subtitle. The reality is messier and more mundane. You're looking at a logistics company that happens to move fentanyl, not a mysterious cabal.
Starting With the Paper Trail
Here's what nobody tells you about cartel financial investigations: the money is almost never the problem. The problem is separating signal from the enormous noise of legitimate commerce. Mexico moves billions in legal freight daily through the same corridors CDS uses. When I started, I thought I'd find smoking guns — cash dumps, encrypted phones, named accounts. I found nothing of the sort. What I found instead was a pattern of commercial invoices that looked normal until you cross-referenced them. A pallet of "industrial cleaning supplies" declared at $2,000 would consistently ship on the same Tuesday each month from the same warehouse zone in Culiacán. The buyer was always a different company name, but the contact person, phone number, and delivery instructions repeated. That was the CDS tell. Not illegal on its face, but statistically impossible for legitimate business. I learned to run these checks manually at first. Excel, a few hundred rows, comparing shipment dates against known CDS arrest records from SINA (the federal crime database). It took about four hours per case. Once I built a simple pivot table that auto-flagged repeated origin-destination pairs with abnormal valuation ratios, the same analysis dropped to twelve minutes. I still use that spreadsheet structure today. The software didn't matter. The logic did.
Understanding the Hierarchy Without the Hollywood Version
Most public reporting on CDS frames them as a traditional Mexican cartel with a clearly defined structure. Jefe. Soldados. Plazas. That framework exists, sure, but it oversells the top-down control. What I observed — and what survived three years of surveillance — was something more distributed. The actual operation runs on semi-autonomous cells. Each cell handles one function: procurement, transport, distribution, money movement. They don't always know each other. The cell moving product through Juárez might not know the cell selling it in Monterrey. Communication flows through intermediaries who themselves don't know the endpoints. This isn't operational security theater. It's how you maintain deniability when half your network gets arrested. The CDS response to pressure is equally pragmatic. When a plaza falls, they don't panic or escalate violence dramatically. They redistribute. A new cell takes over within days, often staffed by people who were already operating adjacent to the old one. I watched this happen in Reynosa after a major 2022 takedown. The seizures made headlines. The volume through that corridor dropped maybe eighteen percent for six weeks, then recovered to ninety-two percent of pre-takedown levels. Not because CDS is unstoppable. Because the structure absorbs loss better than conventional organizations.
Get the Full Details
The Fentanyl Problem and Why It Changes Everything
Before 2021, CDS was significant but not dominant in the northern corridor. That changed with fentanyl. The chemical precursor supply chain is completely different from heroin or cocaine logistics. You're not moving bulk agricultural product. You're moving kilogram-scale shipments of precursors through legitimate pharmaceutical and chemical trade routes. This is where the unravelling gets complicated. CDS doesn't necessarily manufacture everything in-house. They contract with laboratories, some inside Mexico, some across the border in the US Southwest. The money trails through import/export businesses that appear entirely legitimate. I spent six months tracking a single precursor shipment from a chemical supplier in Sonora to a supposed laboratory in New Mexico. The paperwork was flawless. The route was legal. The end product was pure fentanyl hitting the streets of Detroit and Chicago. The countermeasure here isn't enforcement alone. It's financial disruption. I found that freezing accounts tied to the intermediary layer — the people who connect labs to distributors — disrupted operations faster than physical seizures. One account closure can take out three concurrent shipments and delay lab output by two weeks. Physical interdiction rarely does that. But that's also where the system fails. Treasury and law enforcement work at different speeds. By the time a freezing order processes, the money has moved three times.
What Actually Works for Disruption
After years of this work, I have opinions. Some are professional, some are personal. Here are the ones that matter: Follow the logistics companies, not the cartels. CDS uses trucking firms, freight forwarders, and warehousing operators as infrastructure. These entities exist on paper as legal businesses. They also ship CDS product regularly. Shutting down one shell company costs them a day. Disrupting the forwarder they depend on costs them weeks. The legal exposure is genuinely ambiguous, which is why prosecutors hesitate. But the operational impact is real. Regional specialization matters more than national strategy. CDS operates differently in Sinaloa than in Tamaulipas than in Nuevo León. A approach that works in one plaza fails in another because the local economy, geography, and existing criminal groups are completely different. I learned this the hard way when I recommended a tactic that worked in one region and got an operator killed in another because the criminal ecosystem responded differently than expected.
Technology helps, but context is everything. Data analytics can flag anomalies fast. I've seen systems identify CDS patterns in under an hour that would have taken analysts weeks. But models trained on one region's data perform poorly in another. A pattern that's ninety percent accurate in Jalisco might be forty percent accurate in Michoacán because the underlying criminal economy differs. Always validate locally before deploying broadly.
The Limitations Nobody Talks About
Here's the honest part. Most of what we do doesn't produce clean results. Arrests happen, but the network persists. Seizures make news, but product volumes often recover within months. Financial disruptions work until the network adapts, which usually takes six to eight weeks. We're playing a game where winning means making their operation too expensive to sustain, and that's a much higher bar than most people expect. The biggest limitation is institutional. Law enforcement, treasury, intelligence agencies, and international partners don't coordinate well enough for sustained pressure. We hit CDS from three different angles at once, but the timing is rarely aligned. One agency arrests a mid-level operator while another is building a financial case that would have been stronger if it had waited thirty days. This isn't incompetence. It's structural. Different mandates, different timelines, different success metrics. Another limitation is economic. CDS exists because there's demand. Not just in the US, but across the Americas and increasingly Europe. As long as that demand exists, CDS or someone else will fill the gap. Disruption reduces volume and raises costs for buyers, but it doesn't eliminate the market. I've said this to colleagues who prefer more optimistic framing, and it's not well received. But it's true.
A Specific Case That Taught Me Everything
Let me tell you about the Culiacán warehouse operation. Standard raid scenario: intel points to a stash house, you mobilize, you move in. What happened was different. The warehouse was empty. Not recently emptied — never stocked with product. It was a dead drop point, yes, but the actual inventory was split across four locations in different neighborhoods. The intel had identified one node correctly but missed the distribution pattern entirely. Why? Because CDS had changed their logistics model three weeks earlier without anyone updating the databases we relied on. The old model centered product in large warehouses near highways. The new model used smaller, residential stash points that looked like nothing on paper. My team and I spent two weeks trying to figure out why our seizures didn't match our predictions. The answer was simpler than we wanted to admit: the cartel adapted faster than we could track. The workaround I developed after that was to map not just known CDS locations but the spaces between them. Where do drivers wait? Which gas stations serve as checkpoints? Which restaurants are neutral zones? Physical geography matters as much as financial data. A map of CDS touchpoints overlaid on municipal road classifications and traffic patterns reveals operational logic that spreadsheets alone can't show. I combined GIS data with shipment analysis and found patterns that led to three successful interdictions over four months. That was the best work I ever did, and it wasn't because of any breakthrough technology. It was because someone finally looked at the problem from a different angle.
Bottom Line
CDS is dangerous because it's resilient, not because it's omnipotent. It survives adaptation, not brute strength. The shadow network I described here isn't invincible. It's just harder to unravel than official narratives suggest. The work requires patience, local knowledge, and willingness to accept that most interventions produce partial results at best. If you're entering this field, come in with low expectations and high curiosity. The people you'll learn from are the ones who survived long enough to notice what everyone else missed.
