Research · 4 August 2026
83% of Miami immigration law firms do not serve a real Spanish version.
Of 817 firms whose websites we could reach and score, 65% serve no Spanish at all and a further 17% serve it through a machine-translation widget. Measured 4 August 2026.
What the study found
Of 817 Miami-Dade immigration law firm websites reachable and scored on 4 August 2026, 83% did not serve a real Spanish version — 65% served none at all, 17% used a machine-translation widget. Across all 1,624 sites scored, the median compliance score was 0 out of 100 and 36% scored below 60 overall.
How this was measured, and what it is not
On 4 August 2026 we fetched 1,938 distinct root domains collected by our own Miami-Dade prospecting campaigns and scored each reachable homepage with our six-pillar engine. 1,624 were scored. 297 were unreachable, 14 blocked the crawl, and 3 were listing or file URLs. Blocked and unreachable sites are counted here and never scored or estimated.
This is a validation over URLs our campaigns collected. It is not a random sample of Miami businesses, and it should not be read as one. The population over-represents the sectors we prospect and carries no survey weighting, so the figures describe this set of 1,624 websites and nothing wider.
The six pillars are SEO and answer-readiness, security, speed, conversion, compliance, and usability. Speed here is a proxy computed from the response and the document, not a lab measurement. Compliance counts the presence of a privacy policy, terms, and a cookie or storage notice — it is not a legal review. We did not test accessibility conformance and make no claim about it.
The scoring engine is our own and it is deterministic: the same page scored twice returns the same number. Sector is taken from campaign provenance. 88 of the scored domains carry more than one sector, so every per-vertical figure below is computed over single-sector domains only — 1,536 of the 1,624.
Spanish, by vertical
Three states are detectable from a page: no Spanish version, a machine-translation widget in the source, or a real Spanish version served on its own URLs.
Immigration law (n=817): 65% none, 17% widget, 16% real. Trade logistics, customs brokers and freight forwarders (n=584): 60% none, 9% widget, 29% real. Commercial contracting (n=72): 87% none, 4% widget, 8% real.
What detection supports is exactly this and no more. We can see whether a Spanish version exists and whether a translation widget is present. We cannot tell whether Spanish that looks hand-written was machine-translated, and we do not claim to.
The scores
Across 1,624 scored sites the mean overall score was 62 and the median 63. 36% scored below 60. 9% scored below 40. 6% scored 80 or above. 117 sites (7%) carried at least one critical finding.
By pillar, the means were: usability 86, conversion 75, speed 72, SEO and answer-readiness 66, security 53, compliance 20. Compliance is the outlier and the median tells it better than the mean: 890 of 1,624 sites — 54% — scored zero, meaning no privacy policy, terms, or storage notice was found at all.
By vertical, median scores were close: immigration law 64 (n=817), trade logistics 62 (n=584), commercial contracting 61 (n=72). The share scoring below 60 differed more: 32%, 42% and 41% respectively.
Machine-readability
Most sites are legible to a crawler and few are legible to an assistant. Of the 1,624 scored: 91% publish robots.txt, 89% publish a sitemap, and 96% permit AI crawlers. 33% publish llms.txt. On transport security, 36% send HSTS and 20% send a content-security-policy header.
Spanish, by vertical
How each site serves Spanish, as a share of the sites we could reach and score in that vertical.
- No Spanish version
- Machine-translation widget
- Real Spanish version
Single-sector domains only, 1,536 of 1,624 scored. Rows may not total 100 because each is rounded independently. We can detect whether a Spanish version exists and whether a translation widget is present; we cannot tell whether hand-written Spanish was machine-translated, and do not claim to.
View as a table
| Spanish, by vertical | No Spanish version | Machine-translation widget | Real Spanish version | Sample |
|---|---|---|---|---|
| Immigration law | 65% | 17% | 16% | n=817 |
| Trade logistics | 60% | 9% | 29% | n=584 |
| Commercial contracting | 87% | 4% | 8% | n=72 |
Mean score by pillar
Out of 100, across all 1,624 sites scored.
Compliance is the outlier, and its median tells it better than its mean: 890 of 1,624 sites scored zero. Speed is a proxy computed from the response and the document, not a lab measurement.
View as a table
| Mean score by pillar | Value |
|---|---|
| Usability | 86 |
| Conversion | 75 |
| Speed | 72 |
| SEO and answer-readiness | 66 |
| Security | 53 |
| Compliance | 20 |
Where the scores fall
Share of the 1,624 scored sites in each band.
Mean 62, median 63. 36% scored below 60.
View as a table
| Where the scores fall | Value |
|---|---|
| Below 40 | 9 |
| 40 to 59 | 27 |
| 60 to 79 | 56 |
| 80 and above | 6 |
What this does not show
It does not show that a low score costs a business customers, revenue, or rank. Nothing here measures any of those, and no such link is claimed.
It does not describe Miami businesses in general. It describes 1,624 websites our campaigns had already collected.
It does not assess accessibility conformance, and it is not a legal review of any policy found or missing. Sites that blocked our crawl are counted, not estimated.
The same six pillars, run on your own site.
One page, your numbers, and the method in full. Free, and no obligation.