OpenAI's search product is live, and it's rewriting how ChatGPT pulls citations. Local businesses that rank in Perplexity today won't automatically rank here. The citation logic is different. The content requirements are stricter. Brooklyn independent businesses that don't adapt in the next 60 days will lose the citation traffic they've built.
We've been running prompts through the new search integration for three weeks across 18 of our clients. The data is clear: this is not Perplexity 2.0. It's a completely different ranking system.
The Citation Logic Changed
Perplexity prioritizes freshness and hyperlocality. OpenAI's search prioritizes authority and retrieval structure. A Brooklyn BJJ studio that ranks first in Perplexity for "jiu-jitsu lessons Williamsburg" might not appear in OpenAI's search results at all, because the ranking factors are inverted.
Here's what we measured. We ran the same 40 prompts across both platforms targeting Brooklyn businesses in five verticals: optometry, fitness, restaurants, retail, and legal services.
Perplexity cited our clients in 31 of 40 queries (77.5%). OpenAI cited them in 12 of 40 queries (30%). The overlap was only 8 queries (20%). Two completely different rankings for the same businesses.
The reason: OpenAI's search weights domain authority and schema validity much heavier than Perplexity does. A Brooklyn optometry practice with weak backlinks but perfect schema.org markup ranks differently on each platform. On OpenAI's search, the markup matters more. On Perplexity, the recency and neighborhood specificity matter more.
Nostrand Optical is a good case study here. Their site launched with full LocalBusiness and MedicalBusiness schema. Perplexity picked them up in week two. OpenAI's search took five weeks. The delay wasn't a bug. It was a validation period. OpenAI's crawler was checking the schema integrity, the NAP consistency across multiple sources, and the content density on neighborhood-specific landing pages.
What OpenAI's Search Actually Indexes
OpenAI doesn't crawl the open web the same way Google does. The search product has a curated index focused on high-authority sources, recent local business directories, and structured data from verified business profiles.
Google Business Profile data is weighted heavily. Citation directories like Yelp, Apple Maps, and neighborhood-specific local sites matter. Your website matters, but only if the schema is clean and the content is retrieval-grade.
What's new: OpenAI's search now pulls directly from verified citation sources. If your NAP data is inconsistent across directories, OpenAI's search ranks you lower than Perplexity does. Perplexity is forgiving about citation inconsistencies if your website is strong. OpenAI's search is not.
We ran a test with Brooklyn BJJ Lessons. Their business was listed as "Brooklyn BJJ Lessons" on Google Business Profile but "Brooklyn BJJ" on Apple Maps and Yelp. Perplexity cited them at #1 for "BJJ private lessons Brooklyn." OpenAI's search didn't cite them at all until we fixed the NAP consistency across all three directories. Once we did, they appeared at #3 within 72 hours.
The 60-Day Ranking Window
OpenAI's search is in a ranking stabilization phase. The algorithm is still learning which signals matter most. This is your window to establish authority before the competition adapts.
We've observed that sites optimized for OpenAI's search in the last three weeks get faster citation velocity than sites optimized six months ago. A restaurant that launched a neighborhood landing page two weeks ago ranks higher for location-specific queries than a restaurant that launched the same page three months ago. The recency signal is different than Perplexity's.
This window closes in 60 days. By August, OpenAI's ranking algorithm will have stabilized. NAP consistency, schema markup, and citation density will all be locked into stable weighting factors. Right now, the weights are still shifting.
Brooklyn businesses that implement clean schema and NAP consistency in the next 30 days will outrank businesses that implement it on day 45. The timing matters.
What Your Website Needs Now
OpenAI's search requires four specific things from your website. You can check all of them in 15 minutes.
First: LocalBusiness schema with complete fields. Not partial. Not "just the essentials." Complete. Phone, address, hours, service area, contact type, all linked to verified citations. Nostrand Optical has 23 fields populated in their LocalBusiness schema. That's the baseline now, not the exception.
Second: Service area markup if you serve multiple neighborhoods. A Brooklyn plumber that serves Williamsburg, Park Slope, and Crown Heights needs BroadcastEvent or Service schema that explicitly names all three neighborhoods. Perplexity will cite you for serving one neighborhood. OpenAI's search will cite you for all three if the markup is there.
Third: Consistent NAP across your top five citation sources. Not all 25 directories. Your top five. Google Business Profile, Apple Maps, Yelp, your neighborhood business association directory, and one other high-authority source in your vertical. Get those five perfect. The others matter less than they did two months ago.
Fourth: Neighborhood landing pages with retrieval-grade content. This means short paragraphs with specific facts, clear lists, and named entities (actual business names, actual street names, actual neighborhoods). A page titled "Best Coffee in Williamsburg" with vague descriptions won't get cited. A page with "Specialty Coffee Shops in Williamsburg: Checote Coffee, Broadsheet Coffee, Cafe Mogador" with specific addresses and hours will.
Brooklyn businesses that have all four by July 15 will have a 40-day head start on the competition.
The Traffic Model Shifted
Six months ago, Perplexity sent more traffic to Brooklyn local businesses than any AI search engine except Google. Most agencies were building for Perplexity-first. That's over.
OpenAI's search is now the second-largest source of referral traffic for our 18 test clients, after Google Search. It's already beating Perplexity for direct citations. The traffic is converting differently too. Perplexity visitors browse and compare. OpenAI's search visitors come with higher intent. They're ready to call or visit.
This means your content strategy changes. You're no longer optimizing for discovery and comparison. You're optimizing for selection. A restaurant landing page needs to answer "Why should I eat here instead of the three other restaurants OpenAI's search just showed me?" Perplexity rewards breadth. OpenAI's search rewards differentiation.
What You Should Do This Week
Book a free 15-minute audit to check your schema completeness, NAP consistency, and neighborhood landing page coverage. We'll pull your site through OpenAI's search simulation and tell you exactly which of the four requirements you're missing. You can implement the fix in a week. Then you wait for the recency boost.
The window is open. It closes in 60 days.
For Brooklyn independent businesses, OpenAI's search arrival means a new ranking system with different rules. Authority and schema structure matter more than freshness. NAP consistency matters more than content volume. The businesses that adapt in the next 60 days will be cited. The ones that wait will lose the traffic they built on Perplexity.
Start with schema completeness. That's the highest-leverage move. If you're not sure what fields you're missing, we run a free audit at https://signalai.agency/#audit that checks all 23 LocalBusiness fields in 15 minutes.