Google Meridian

Named in 21 judge labels across 1 category by 5 of 6 models in the September 2026 Edition. 4 first choices, 4 negative labels. Every number here is derived from the raw labels under vendor table v2026-09-08.5.

Best standing

15% in Attribution & MMM

Rank 3 of 82 products named in attribution and marketing mix modeling, accepted challenger. 0 of 6 models made it the first choice on the direct prompt; 19% of its 21 labels there were negative.

Standing by category

Every category where a model named Google Meridian. Share is first choices across the direct, paraphrase, budget and scale prompts; rank is within every product named in that category.

Category	Vertical	First choices	Rank	Negative rate	Labels	Quadrant
Attribution and marketing mix modeling	Marketing	15%	3 of 82	19%	21	accepted challenger

By model

How each model treated Google Meridian across every prompt where it was named. Six models, six prompts per category.

Model	First choice	Alternative	Mention	Negative	Labels
Claude Opus 5	2	3	0	1	6
Claude Opus 4.8	1	1	2	0	4
GPT-6 Astra	0	2	1	1	4
GPT-5.6 Sol	1	3	0	1	5
Gemini 3.1 Pro	1	0	0	1	2
Perplexity Sonar Pro	0	0	0	0	0

What the models said for it

Verbatim evidence the judge attached to positive labels.

“If forced to name one: Google Meridian, calibrated with geo experiments. It's free, transparent, actively maintained, and its budget optimizer is the strongest of the open-source set.” Claude Opus 5 · Attribution & MMM · budget prompt · first choice

“layering in an open-source MMM (Robyn or Meridian)... Meridian if you want Google-ecosystem integration or geo-level granularity” Claude Opus 4.8 · Attribution & MMM · budget prompt · first choice

“For most companies with a genuinely limited budget, I recommend Google Analytics 4 plus Google Meridian” GPT-5.6 Sol · Attribution & MMM · budget prompt · first choice

“represents the current gold standard for privacy-safe, top-down modeling” Gemini 3.1 Pro · Attribution & MMM · budget prompt · first choice

And against it

Verbatim evidence attached to negative labels. A warning on a product with few labels is a warning; on a product with many, it is one voice among them.

“Meridian and Robyn are free and production-grade, but they require in-house econometrics and Python capability... At 500 people, buy, unless you already have econometricians.” Claude Opus 5 · Attribution & MMM · scale prompt · soft negative

“Avoid a DIY deployment without an accountable modeler... The concern is the implementation, not that open-source software is inherently inferior.” GPT-6 Astra · Attribution & MMM · negative prompt · soft negative

“free to use but require significant engineering resources to maintain” Gemini 3.1 Pro · Attribution & MMM · comparative prompt · soft negative

“Caution level: Low as frameworks; high as turnkey solutions.” GPT-5.6 Sol · Attribution & MMM · negative prompt · soft negative

Names read as Google Meridian

What the judge wrote, as written, with how often. The vendor table decides that these count as Google Meridian; a claim can dispute any of them.

Google's Meridian 2Meridian 2Google Meridian / Meta Robyn 1

Is this your product?

Claim this page

Claiming is free and changes nothing in the data. A claimed page gets a verified contact who is told when each edition publishes; the right to propose corrections to the vendor table, meaning names the judge wrote that should or should not read as Google Meridian, applied by version and listed in the change log; and a logo and one-line description supplied by the vendor and marked as such.

It does not get any change to labels, shares or quadrants, any preview, or any say over which quotes appear. Sponsorship is separate: a sponsor funds categories or buyer dimensions and is named on what it funded, and the rules are the same for every sponsor.