Squarespace Is the Named Website Partner for Perplexity's Comet

The announcement landed on the same day Perplexity made Comet free to the public worldwide³. Comet had previously been gated behind Perplexity Max at $200 per month⁸. The pairing of the partnership and the free release was deliberate: Squarespace becomes the default destination for Comet users who want to convert AI research into a live business presence. Paul Gubbay, Squarespace's Chief Product Officer, described the scope as “our partnership with Perplexity helps guide anyone creating a website towards our best in class platform”¹. The Perplexity Chief Business Officer Dmitry Shevelenko framed the same announcement from his side: “This partnership ensures that as Comet transforms how people explore the internet, the businesses they discover are equipped to deliver exceptional experiences”¹.

Two things to notice. First, the partnership announcement does not say Squarespace-built sites get preferential ranking in Perplexity search results. Second, the companion Squarespace blog post references Squarespace Beacon AI as “coming soon”² — the technical specifics of how Comet integrates with Squarespace-built sites are not fully documented as of mid-2026. The honest framing for a Squarespace owner: this is a routing partnership inside Comet's browsing surface, not a ranking promise inside Perplexity's answer engine.

Perplexity's crawler documentation states each role plainly⁴. PerplexityBot: “designed to surface and link websites in search results on Perplexity. It is not used to crawl content for AI foundation models.” User-agent string identifies as PerplexityBot/1.0. Respects robots.txt; Perplexity recommends allowing it. Perplexity-User: “supports user actions within Perplexity. When users ask Perplexity a question, it might visit a web page to help provide an accurate answer.” User-agent identifies as Perplexity-User/1.0, and the docs note it “generally ignores robots.txt rules” because user-initiated fetches drive the visit.

The robots.txt asymmetry mirrors the one OpenAI documents for ChatGPT-User⁹. The reasoning is the same: when a person asks an AI agent a live question, the visit is no different in spirit from the person typing the URL into their own browser, and robots.txt is not generally interpreted as a stop sign for human navigation. The practical consequence for Squarespace owners: even if every AI block were enabled, Perplexity-User would still reach the site. The bot that does honor robots.txt is PerplexityBot, and Squarespace's panel does not list PerplexityBot among its 26 controlled user-agents.

Perplexity's interface decision is the part that changes the optimization math. Where ChatGPT cites inside a conversational paragraph, Perplexity puts the source cards at the top of the answer with consistent visual treatment. A user scanning a Perplexity response sees the citations first, the synthesis second. That weighting means a Squarespace site cited as one of three to seven cards above a Perplexity answer can receive measurable click-through from queries no Google ranking would have caught.

The content shape that gets cited has been documented across multiple 2026 GEO studies⁷. Three patterns appear consistently. First, Perplexity favors comparison and listicle content for product-research queries because each list item is a self-contained extraction unit. Second, content with year-specific signals (“2026” in the title and opening paragraph) earns higher citation rates than undated content. Third, pages that cite named primary sources inline are weighted more heavily than pages making the same claims without attribution — a hallucination guardrail the engine carries by design.

The citation-surface difference is visible to any user who runs the same query in both engines. A Perplexity query like “best Squarespace plan for a small photography business” returns a numbered set of source cards (typically three to seven) above a synthesized answer³. The same query in ChatGPT returns prose with inline links to one or two sources, sometimes a “More sources” expandable footer, and rarely with the visual weight Perplexity gives citations. The display difference changes user behavior: Perplexity readers click citation cards; ChatGPT readers click occasionally to verify a specific claim.

The content-shape difference is the part that matters for what to write. Search Engine Land's 2026 GEO guide⁷ documents that Perplexity weights comparison-format content heavily for product-research and decision-support queries. The mechanism is mechanical: a comparison article with five H3 sections (one per option) gives the engine five extraction units instead of one, and each unit can stand alone in a citation card. ChatGPT's narrative-citation surface does not benefit from that fragmentation as much — it lifts longer passages and prefers continuous reasoning. The implication for a Squarespace editorial calendar: lead with comparison content for Perplexity, lead with explainer content for ChatGPT, and the two together cover both engines.

 # A long-form "best Squarespace plan" explainer article For ChatGPT: 1 page = 1 narrative extraction unit For Perplexity: 1 page = 1 narrative extraction unit # The same article rewritten as a 5-plan comparison with H3-per-plan For ChatGPT: 1 page = 1 narrative extraction unit + 5 weaker fragments For Perplexity: 1 page = 5 strong extraction units, one per plan 

The Perplexity-vs-ChatGPT difference does not require running two content programs. A single comparison article structured with H3-per-item, a bolded one-sentence verdict at the top of each, and named sources inline gets cited well on both engines. The Perplexity edge sits in how much more weight that structure earns on the citation cards. The comparison-content leaf documents the format in full; the ChatGPT cluster covers the narrative-passage rules for that engine.

The reason to install in this order is mechanical. Crawler access is the floor — without it, the other three steps produce content the engine cannot read. Citation hygiene comes next because it gates the engine's confidence to quote. Comparison structure compounds the previous two by giving the engine multiple extraction units per page. Measurement closes the loop and produces the signal you need to know which content earned the citations. Skip a step and the next one underperforms.

The verification step is the under-done part of most installs. Open yoursite.com/robots.txt in a private window and look for any User-agent: PerplexityBot or User-agent: Perplexity-User blocks. On a default Squarespace site, neither should appear — the platform does not list these bots on its panel-controlled checkbox⁶. If you see Disallow rules, custom code added them somewhere in your site configuration, and the unblock leaf walks the removal path.

The mechanism is the same hallucination guardrail every generative engine carries: a page with named-source attribution gives the model a defensible answer if a user asks where the claim came from. Perplexity makes this visible — the source card above the answer is the engine's commitment to its quote. A page that cites “Squarespace's help center documentation, 2026-Q1” with a primary-source link is easier for the engine to lift from than a page citing nothing. The full pattern lives in the citation-hygiene leaf.

The 7.1 patterns that ship this format natively are the Markdown Block (fastest for plain lists), the Text Block with manual H3 + bold lead + body (best for editorial control), and the Summary Block with manual configuration (cleanest visual cards). The block-by-block walkthrough with screenshots lives in the comparison-content leaf.

The tracking template, the queries we recommend by industry, and the GA4 channel grouping for the small amount of tagged Perplexity traffic that does carry referrers all live in the 15-point checklist at the bottom of this cluster. Expect 6-12 weeks before steady citation growth shows up in the tracking log, longer for engines with slower freshness windows.

Cloudflare's investigation⁵ documented two distinct crawling patterns. The declared crawlers (PerplexityBot and Perplexity-User) generated roughly 20-25 million daily requests across the Cloudflare network. The stealth crawler — a generic Chrome browser identifier on macOS, rotating through IPs outside Perplexity's published range — generated an additional 3 to 6 million daily requests. The stealth traffic correlated with sites that had blocked the declared crawlers, which is the pattern Cloudflare flagged as evasive.

Cloudflare took two actions in response: it removed Perplexity from its verified-bot allowlist, and it shipped detection heuristics in its managed rules to block the stealth pattern for customers who want it blocked⁵. For a Squarespace owner whose site goal is Perplexity citations, the implications are mostly directional: the engine is going to visit your site one way or another, and the question is whether it visits in a way that produces a clean robots.txt-friendly index (PerplexityBot, the recommended state) or a way that bypasses your stated preferences (the stealth pattern, which Cloudflare can now block separately).