Meta Robyn

Named in 16 judge labels across 1 category by 5 of 6 models in the September 2026 Edition. 1 first choice, 6 negative labels. Every number here is derived from the raw labels under vendor table v2026-09-08.5.

Best standing

4% in Attribution & MMM

Rank 7 of 82 products named in attribution and marketing mix modeling, criticized challenger. 0 of 6 models made it the first choice on the direct prompt; 38% of its 16 labels there were negative.

Standing by category

Every category where a model named Meta Robyn. 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	4%	7 of 82	38%	16	criticized challenger

By model

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

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

What the models said for it

Verbatim evidence the judge attached to positive labels.

“Choose Robyn if most of your spend is on Meta/social” Claude Opus 4.8 · Attribution & MMM · budget prompt · first choice

“the three real open-source options are Google Meridian..., Meta Robyn (R-based, supports calibration against lift tests)” Claude Opus 5 · Attribution & MMM · budget prompt · alternative

“Meridian is generally the stronger starting point for a new Bayesian, geo-level program.” GPT-5.6 Sol · Attribution & MMM · comparative prompt · alternative

“Robyn likewise documents its media, organic and contextual-variable treatment.” GPT-5.6 Sol · Attribution & MMM · scale prompt · alternative

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.

“Robyn’s documentation emphasizes analytical expertise, modeling assumptions, and experiment calibration. Neither framework removes the need for competent implementation.” GPT-6 Astra · Attribution & MMM · negative prompt · soft negative

“signals that Meta has been winding Robyn down while Google pushes Meridian harder — which is also a maintenance risk” Claude Opus 5 · Attribution & MMM · scale prompt · soft negative

“Traditional MMM tools (like Meta's Robyn or Nielsen) require thousands of conversions a month to find patterns” Gemini 3.1 Pro · Attribution & MMM · direct prompt · soft negative

“describes Robyn as semi-automated and notes that automation can create modeling risk” GPT-5.6 Sol · Attribution & MMM · negative prompt · soft negative

Names read as Meta Robyn

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

Meta's Robyn 3Robyn 3Google 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 Meta Robyn, 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.