Most writeups about Brazilian corporate email targets treat them like they were stamped out of a single mold. They were not. Magazine Luiza (a Brazilian department store chain, usually shortened to Magalu in everyday conversation) is one of the larger corporate targets in the region, and the blog posts that promise a single magic pattern tend to overstate how predictable its inbox layout really is. I want to walk through what the sampled records show, where the lookup vendors stop agreeing, and what a realistic cold email workflow looks like when you do not want to torch your sender score on day one.
Cold email is a numbers game, but the framing matters. I care about friction, not feature lists. If a guess lands in two tries instead of seven, that is real time back in your day. That is the lens I want to use here. You are not buying a vendor promise. You are picking a workflow that will actually deliver.
What the sampled records actually show
When you strip out the lookup-vendor marketing and look at the raw sampled pool of inboxes, one shape appears more often than the others. It is the lowercase firstname.lastname form, with no middle initials, no digits, no underscores, and no department tags. A worker named Caio Inbox at the retailer would most plausibly sit at caio.inbox@magazineluiza.com.br under that pattern.
The margin between the top shape and the next is real but not dramatic. Roughly 17% of sampled inboxes use the dotted layout. The next shape, where first and last name are jammed together with no separator, sits around 15%. A third form, where the first name is followed by the last initial, sits closer to 11%. None of the top three are runaway winners. The winning pattern is a plurality, not a majority.
Here is the rough ordering of the top five shapes I see in the sampled pool:
- Dotted lowercase form, the strongest single guess, around 17%
- Run-together lowercase form, the strongest second guess, around 15%
- First name plus last initial, around 11%
- First initial plus last name, around 9%
- First name only, around 7%
That stack covers just under 60% of the sampled pool. The rest lives in long-tail shapes that no pattern-based guess can reach. That long tail is the part where a paid lookup service earns its keep, and pretending otherwise is the kind of confidence that loses you a sender score.
Why the vendors cannot agree on a single number
This is the part most writeups skip. The vendors who sell you this data rate the dotted shape very differently from one another, and the gap between the lowest and the highest estimate is wide enough that no single number should be quoted as ground truth.
- One vendor reports the dotted shape at roughly 89% of its records.
- A second vendor puts the same shape at roughly 92%.
- A third climbs as high as 96%.
- A fourth vendor goes a different direction and reports an underscore-shape as the leader at roughly 23%.
Each vendor holds a different slice of the overall pool and scrubs it under different rules. The direction is consistent across all four. The exact share is not. If you treat the highest number as gospel, you will send to a meaningful share of dead inboxes and you will not understand why your reply rate dropped.
The honest read is that the dotted shape leads by a real margin, and the precise percentage is up for debate. Any blog post that quotes a single number with two decimals is reading from a vendor brochure, not from the underlying data.
Verify before you send, every time, no exceptions
A guess can look right and still fail when it hits the receiving mail host. A verifier pings that host and reports whether the address is alive and willing to take delivery. Skipping that step bills you later, and the bill compounds quickly. Hard bounces eat into your sender reputation. A weak reputation pushes the next campaign into junk folders, even the messages to people who never asked to hear from you.
Free verifier tiers are fine for small batches in the dozens. The moment you are running real volume, pay for one. The cheapest paid plan is rarely cheap when your sender score drops a few points and stays there for weeks. A single misfired campaign can take a full quarter to recover from, and most teams never realize they are the cause of their own deliverability decay.
A few habits that keep the bounces down:
- Run the verifier right before send, not when you build the list. Inboxes go stale inside a month.
- Send the high-confidence batch first as a probe. If bounces climb, pause before you touch the long tail.
- Keep the verifier logs. When something breaks, the log tells you whether it was the data or the sender.
What departments break the pattern, and where to look first
Public Brazilian retailers tend to share a corporate naming convention across most departments, but the long tail is uneven. In my sampled look at this retailer, three buckets drifted away from the dotted shape more than the others.
- Executive and director-level staff often use a first-name-only address or sit on a different domain altogether. The dotted guess still works more often than not, but the failure rate climbs.
- Marketing and PR contacts are the hardest group to reach. Vendor records here are noisy, and many addresses live on agency domains rather than on the corporate domain. A guess on the parent domain will bounce in those cases.
- Customer support and call center staff sometimes sit on a different subdomain. The dotted shape still works, but a verifier check is what tells you whether the box is alive on any given day.
If you are reaching out to anyone in marketing or to anyone whose title mentions director, treat the first guess as a probe and assume the second guess will need a verifier behind it. The other departments are more predictable, and you can move faster through them.
A short checklist before you load a campaign
Before you queue a list and hit send, run these checks once. They take ten minutes and they save you from a sender-score crash later.
- Confirm which domain you are actually sending against. This retailer has more than one. The main domain is
magazineluiza.com.br, but subsidiaries and partner brands exist. Sending to the wrong domain is the single biggest reason a campaign looks broken when the data was the actual problem. - Strip the role-based addresses. Anything starting with
contato@,imprensa@,sac@orfaleconosco@will not reach a person and will hurt your bounce rate. Remove them from the list before you run the verifier. - Run a verifier pass on the cleaned list. Free tiers cover small lists. Paid tiers cost less than a single campaign reset.
- Sort by guessed pattern and send the dotted-pattern list first. That is your highest-confidence batch. If something is misconfigured, you want to find it on the strongest segment, not the weakest.
- Hold the long tail for a separate campaign window. The 40% of inboxes outside the top five shapes deserve their own sending window, not a forced march through the same list.
There is usually a catch, and pretending otherwise is how you spend a Tuesday chasing inboxes that do not exist. The catch here is that the long tail is where the high-value contacts often sit, and the long tail is also where guessing alone will not get you there.
Trade-offs
The dotted shape is the strongest single guess, but it is not a guarantee. Roughly 40% of inboxes fall outside the top five shapes, and that long tail is full of edge cases (executives, contractors, marketing aliases, people who set their own addresses). Lookup services exist precisely because guessing alone cannot reach them.
Vendor stats disagree by tens of percentage points, which means any confidence claim above 90% on a sales page is marketing copy, not evidence. Use vendors as a starting point, not as a source of truth. If two vendors disagree, neither is right and you have to verify against the mail host.
Verifiers cost money at scale and they slow down campaigns. Skipping them costs more, but the choice is yours. The trade-off is speed versus sender score, and sender score is the asset that compounds over time.
There is also a softer trade-off I want to name. Cold email at scale puts you in the same room as people who do not want to hear from you. The patterns and verifiers help you reach the right inboxes. They do not help you decide whether reaching out was the right call in the first place. That is a separate decision and it deserves its own check before you build the list.
What I would tell past me
Start with the dotted shape, run a verifier on every address, and read vendor stats as a starting point. The Brazilian retail space is large enough that any single source is incomplete. If you are running more than a handful of sends a week, pay for the tools. The cheapest plan is rarely cheap when your sender score drops, and the recovery time is measured in weeks, not days.
The right test is whether your bounce rate stays under 3%. The right answer to a bounced inbox is a verifier, not a fourth guess at a shape.