Nike Just Swept Allbirds 3-0 in the AI Answer Layer

Allbirds was supposed to be the footwear brand that broke the mold. Merino wool sneakers, a sustainability story that landed hard with millennials, a 2020 IPO that valued the company north of $4B. For a stretch there, if you were writing a case study on DTC done right, Allbirds was Chapter One.

Then the stock tanked, the stores closed, and the brand narrative flipped. Allbirds today is a turnaround story trying to remember what it used to be.

Nike is Nike.

We wanted to see how that plays out inside the AI answer layer, so we ran the two through Foxish's /vs framework: five buyer-shaped prompts, three models (ChatGPT, Gemini, Perplexity), and a tally of which brand each model actually recommended when a real shopper would be listening. You can see the full head-to-head at foxish.ai/vs/allbirds-vs-nike.

The verdict: NIKE — DOMINATION. Nike swept the board 3-0.

The Numbers Aren't Subtle

Here's what each model said when asked buyer-intent questions like "best footwear to buy right now":

Model	Allbirds mentions	Nike mentions	Winner
ChatGPT	1	5	Nike
Gemini	0	3	Nike
Perplexity	4	5	Nike
Total	5	13	Nike 3-0

Two things jump out. Gemini shut Allbirds out entirely. Zero mentions across five prompts. Allbirds doesn't exist inside Google's own AI when the conversation turns to "what should I buy." That's not a soft loss. That's the brand being invisible to the model most likely to be embedded in future Google surfaces. Perplexity was the closest fight — and Allbirds still lost. 4 vs 5. Perplexity's answer style leans more editorial and it cites sources aggressively, which tends to reward brands with strong review-site coverage. Allbirds has that. But Nike still edged it out because Nike has that and everything else.

ChatGPT sat in the middle. One mention of Allbirds against five of Nike is essentially a token appearance. The model knows Allbirds exists. It just doesn't think of Allbirds first, second, or third when someone asks a purchase question.

Being known isn't the same as being recommended. Allbirds is known. Nike is recommended.

Why the Models Picked the Way They Did

This is the interesting part. The result isn't surprising in the "big brand beats small brand" sense. Of course Nike has more brand equity. What's worth unpacking is the specific signals each model is weighting to arrive at that answer.

A few plausible mechanisms:

Training corpus mass. Nike shows up in orders of magnitude more training text than Allbirds does. Sneakerhead forums, sports journalism, product review sites, retail listings, Wikipedia pages for individual shoe models, decades of ad copy, dozens of collab writeups. Allbirds gets DTC-era coverage from 2016-2021 and then a lot of "what happened to Allbirds" pieces. The volume gap alone would tilt any statistical model toward Nike. Product breadth as a recommendation multiplier. Ask "best running shoe," "best walking shoe," "best gym shoe," "best casual sneaker" — Nike has an answer for each. Allbirds has essentially one product category (lifestyle/casual) with a running line that never got the reviewer coverage Nike's did. When a model tries to fit a query to a recommendation, breadth wins. Recency of positive coverage. LLMs are sensitive to sentiment in recent training data. Allbirds' recent press cycle has been about store closures, layoffs, and turnaround plans. Nike's had headwinds too, but the coverage skews product-and-culture rather than existential. If a model has ingested 200 articles about Allbirds in the last 18 months and 60% of them use words like "struggle," "decline," and "loss," that colors how the brand surfaces in a "best of" list. Structured data on retailer and review sites. Nike products live in structured data on Zappos, Foot Locker, Dick's, Runner's World, Wirecutter, and hundreds of other sites with schema-marked product info. Perplexity in particular loves this. Allbirds sells largely direct, which is great for margin and terrible for AI distribution. The Perplexity anomaly. Allbirds actually held its own on Perplexity (4 mentions to Nike's 5). That tracks with how Perplexity works. It cites live web sources, and Allbirds has strong presence in sustainability-focused review sites, "best eco-friendly sneaker" listicles, and press coverage on materials. When Perplexity leans on that corpus, Allbirds surfaces. When ChatGPT leans on its training weights and Gemini leans on Google's ranking signals, Allbirds gets buried.

That last point is the most actionable one in the whole analysis.

What Allbirds Should Actually Do

The temptation, if you're on the Allbirds marketing team, is to read this and reach for a brand campaign. Don't. This isn't a brand awareness problem. It's a retrieval and ranking problem, and it needs retrieval and ranking answers.

Here's what would actually move the needle:

Own a specific query, not a category. Allbirds will not out-mention Nike on "best sneakers." Ever. But "best merino wool shoes," "most sustainable sneakers 2025," "best machine-washable sneakers," "shoes made from recycled materials" — those are winnable. The play is to identify the 20-30 queries where Allbirds should be the top answer and engineer content, PR, and retailer coverage around them. Right now Allbirds is competing everywhere and winning nowhere. Get into third-party structured product data. The DTC-purity thing hurt Allbirds twice. Once at retail, and now inside the AI answer layer. Getting products listed with rich schema on Zappos, REI, and outdoor-focused retailers isn't a channel decision anymore. It's a visibility decision. Perplexity in particular indexes those pages aggressively. Fund the review corpus. Wirecutter, RTINGS-equivalents for footwear, Runner's World, Outside Magazine — brands need to be in the comparison content that models cite. Getting a pair of Tree Runners into every "best walking shoe" roundup is a 12-month PR campaign, but it's also the difference between "Allbirds gets 0 mentions on Gemini" and "Allbirds gets 2 mentions on Gemini." Each one of those roundups compounds because models cite each other's citations. Publish structured comparison content on-domain. Allbirds' own site should have a page called something like "Allbirds vs Nike: How We Compare on Sustainability, Comfort, and Price" with a proper comparison table, schema markup, and honest framing. Models cite brand-owned comparison pages more than most marketers realize, especially when the comparison includes concrete specs. Nike won't publish that page. Allbirds should. Feed the Q&A layer. Product pages need FAQ blocks with schema. Every question a buyer might ask — "Are Allbirds good for wide feet?" "Can you run in Allbirds?" "How long do Allbirds last?" — should have a structured, answerable block on the product page. ChatGPT and Perplexity both scrape these aggressively when forming responses.

None of this is glamorous. It's not a Super Bowl ad. But the marginal dollar spent on schema markup and comparison content will beat the marginal dollar spent on brand right now, because the retrieval problem is bigger than the awareness problem.

Allbirds doesn't need more people to know it exists. It needs more machines to recommend it when they do.

The Broader Lesson for Footwear

Allbirds vs Nike isn't the only fight worth running in this category. Hoka, On, Vionic, Rothy's, Cariuma, Vessi — every challenger brand in footwear is running the same play against the same incumbents (Nike, Adidas, New Balance) and getting the same answer from the models.

The pattern:

Incumbents dominate broad queries. "Best sneakers," "best running shoes," "best athletic wear" — Nike, Adidas, and New Balance sweep. This is not going to change with anything short of a decade of category shift.
Challengers can win narrow queries — if they invest. On Running actually punches well above its size on "best road running shoes" and "cushioned running shoes" because Hoka and On collectively rewrote that vertical's coverage. Allbirds could do the same in sustainable/casual. Cariuma is trying it in canvas.
Perplexity is the challenger's best model. Every challenger brand we've tested does relatively better on Perplexity than on ChatGPT or Gemini. If you're a smaller brand, Perplexity is where your visibility work has the highest leverage today. It's also where a lot of high-intent research traffic is moving.
Gemini is the incumbent's home turf. Because Gemini is downstream of Google's ranking signals, and Google's ranking signals reward domain authority and backlink profiles that took decades to build. Beating an incumbent on Gemini is the hardest game in AI visibility right now.

If you're a footwear brand and you're not measuring which model recommends you against which competitor, you're flying blind on a channel that already drives real purchase behavior. And the delta between models isn't a rounding error. Allbirds got 0 mentions on Gemini and 4 on Perplexity for the same set of questions. That's not variance. That's a strategic signal about where to focus.

What This Means for Your Brand

The Allbirds/Nike result isn't a footwear story. It's the same story playing out in every category we run these fights on: skincare, SaaS, home goods, supplements, financial services. The incumbent wins broad. The challenger wins narrow — but only if the challenger has done the AEO work to earn it.

Most challenger brands haven't. They're still spending on Google Ads and thinking of AI as a future channel. Meanwhile their customers are already asking ChatGPT what to buy, and the answer isn't them.

The good news: this is a fixable problem. The signals models weight are knowable. The content patterns that surface in AI answers are reproducible. The comparison pages, the schema markup, the review corpus, the structured Q&A — none of it is exotic. It just has to actually get done, and it has to be measured against real model output, not vibes.

Which is where Foxish comes in.

Run Your Own Fight

We built Foxish to measure exactly this: how often ChatGPT, Gemini, Perplexity, and Google's AI Overviews mention your brand versus your competitors, on the buyer-intent queries that actually drive purchases. The /vs tool lets you run a head-to-head like the one above on any two brands. The full platform tracks your visibility over time, benchmarks you against your category, and shows you which prompts you're winning and losing.

You can run your own brand fight at foxish.ai/vs, or get a full visibility scan on your own brand — free, no credit card — at Start a free scan →.

Nike won this one. The next fight's yours to run.