The brief
Lecto Lab sells reading tools to schools. They had one Spanish marketing site and a habit of calling it LatAm. Hreflang tagged every Spanish URL as every Spanish-speaking country. Publisher outreach used a single spreadsheet. Mentions from Mexico landed on a URL that also claimed to be for Colombia. Search systems were being told the company had not chosen.
School procurement is local. Vocabulary is local. The Mexican results page for a classroom-reading query is not the Colombian one. Mixing the citations made both graphs look accidental.
The commercial team wanted one report. They got two, on purpose. A blended “Spanish America” chart would have hidden which market was actually moving.
What we did
On-site: two country hubs, honest hreflang, no x-default games. Outreach did not start until those URLs existed. Mentions have to point at the page meant to rank for that audience.
Two prospect lists. Mexican education titles, teacher magazines, and a few national papers with an education desk. Colombian equivalents, built from scratch, not translated. A host that was useful in Mexico was not copied into the Colombian sheet “to save time.”
Pitches talked about classroom practice in that country: devices, curriculum notes, procurement cycles. A recycled English case study was not the angle. Teachers can tell.
Metrics were compared within a market, never as a blended DR average across both. Blended averages are how you congratulate yourself for the wrong country.
Sessions across the engagement
Jan–Aug 2025. Organic is the destination URLs in this brief; referrals are clicks from accepted placements.
Organic sessions Referral sessions
What actually went live
Accepted URLs only. Declined prospects are not dressed up as a pipeline.
- Mexico placements22
- Colombia placements18
- Regional (rare)3
What moved
Mexico hub organic sessions rose 44 percent. Colombia rose 38 percent. Both from a low base. School-year seasonality is visible in the monthly chart; we did not flatten it into a straight line to make the slide prettier.
Referring domains that actually helped were country-matched. A handful of regional Spanish titles were used only when the article was about a practice that genuinely crosses the border. Default remained no.
Lecto Lab’s board now sees two markets. The previous single “LATAM SEO” slide is gone. That governance change will outlast any one placement.
School calendars are not a rounding error. Mexican and Colombian terms do not peak in the same week. We reported against those calendars rather than against our invoice month. A flat line would have hidden that Mexico moved earlier in the spring and Colombia moved when classrooms actually came back. If you blend the two countries to make a smoother chart, you are back in the folder trick, only this time in the dashboard.
Hreflang, once it was honest, stopped being a superstition. Mentions pointed at the hub that matched the publisher’s country. A Mexican education magazine citing the Colombian URL would have been a mistake we refused to make even when the magazine was easier to close. Easy closes are how mixed profiles happen. The placement bars in this write-up are counted separately for a reason: so nobody can average them into a comforting “LatAm total.”
The control URL — a shared Spanish blog not tied to a country — barely moved. That is useful. It suggests the lifts on the hubs were not a general brand tide. They followed the citations we actually placed. A rising tide is a pleasant story. A control URL that stays put is better evidence.
Selected queries
Positions from the market named in the brief. A dash means the URL was not ranking in the top 100 at the start.
| Query | Before | After | Move |
|---|---|---|---|
| comprensión lectora primaria MX | #20 | #8 | +12 |
| software lectura aula México | #27 | #11 | +16 |
| comprensión lectora Colombia | #18 | #7 | +11 |
| herramientas lectura colegios CO | #24 | #10 | +14 |
What we declined
- A single pan-regional guest-post network.
- Reusing Mexican copy in Colombian pitches.
- Hreflang-as-decoration on one URL.
Peru is a conversation for next year, as a third brief, if the hub exists. It will not be added as a row on the Mexican sheet.
Two markets, two graphs, two reports. That sentence is the entire engagement, repeated until it sticks. A product can be shared. A publisher list cannot. The donut in this write-up is split on purpose: so a blended “Spanish America” slice never gets to look like a strategy.
The shared blog used as a control is the unglamorous proof. If both country hubs had simply ridden a brand wave, that URL would have moved with them. It did not.
Written 29 August 2025. Figures are from the campaign reporting for Lecto Lab.