Perplexity said on 4 November that its agent platform, Computer, ran Meta ads for seafood seller Get Maine Lobster that brought in 3.5 times the revenue of the company’s usual ads while spending 2.1 times as much. In a customer story, Perplexity says the campaign had 63% higher return on ad spend and 14% lower customer acquisition cost.
The test ran on a $30,000 budget that Get Maine Lobster received from Perplexity, the customer story says. The figures are Perplexity’s, from one customer.
Business Pill · Agentic Commerce
An evergreen Business Engineer explainer on agentic commerce. It teaches the general idea only and does not discuss Perplexity or Get Maine Lobster.
The key insight: Perplexity says its Computer agent brought Get Maine Lobster 3.5 times the revenue of its usual Meta ads on 2.1 times the spend, about 1.67 times the revenue per ad dollar by our arithmetic, in a test run on a $30,000 budget that Perplexity provided.
Perplexity posted on X at 17:39 UTC: “Computer managed $30k in Meta ad spend for Get Maine Lobster, outperforming their agency”. Chief executive Aravind Srinivas followed at 17:47 UTC: “Perplexity’s agents are running ads better than ad agencies for business owners”.
The customer story describes the comparison differently. It says Get Maine Lobster “ran both the Computer-generated Meta campaign, and its typical non-AI generated ads”, and that “Computer’s campaign generated 63% higher ROAS than the non-AI one while spending 2.1x as much, bringing in 3.5x the revenue”. The story does not name an ad agency.
By our arithmetic, 3.5 times the revenue on 2.1 times the spend is about 1.67 times the revenue per ad dollar, consistent with the stated 63% after rounding.
Perplexity’s claim for its Computer agent’s Meta campaign for Get Maine Lobster against the company’s usual non-AI ads: 2.1 times the spend and 3.5 times the revenue, with 63% higher return on ad spend. Data: Perplexity customer story, 4 November 2026; one customer, company claims, $30,000 test budget from Perplexity.
What Was Posted on X
Computer managed $30k in Meta ad spend for Get Maine Lobster, outperforming their agency.
Compared to their usual ads, Computer delivered 3.5x revenue with 63% higher ROAS. It automated creative analysis, ad production, and media buying.
Perplexity’s post says Computer managed $30k in Meta ad spend for Get Maine Lobster, outperforming their agency, with 3.5x revenue and 63% higher ROAS compared with the company’s usual ads.
Perplexity’s agents are running ads better than ad agencies for business owners.
Perplexity chief executive Aravind Srinivas wrote that Perplexity’s agents are running ads better than ad agencies for business owners. The customer story behind the claim covers one company.
How the Test Ran
According to the story, “Get Maine Lobster received a $30,000 test budget from Perplexity to test Mark’s hypothesis”, referring to founder Mark Murrell. Each morning, it says, Computer analyzes Meta and Shopify data, then generates ad ideas, writes creative briefs and makes static ads that it “sends them to Mark in Slack for review”; it connects with Higgsfield AI for video ads.
Perplexity says that “over the first seven-day campaign, he went from approving 20% of Computer’s ads, to approving 90% with no feedback”. Murrell is quoted: “All I have to do is approve campaigns in Slack, and they ship”.
As we read the customer story, the agent took over the production side of advertising, analysis, briefs, creative and buying, while the owner kept an approval step in Slack; the approval rate rising from 20% to 90% in a week is Perplexity’s measure of trust in that loop.
As we read the two claims side by side, the X post compares the agent with an agency, while the customer story compares it with the company’s own non-AI ads; one customer and a vendor-funded budget cannot settle which comparison generalises.
Mark Murrell, CEO, Get Maine Lobster (in Perplexity’s customer story)
“All I have to do is approve campaigns in Slack, and they ship.”
Three Implications
FOR SMALL MERCHANTS Perplexity says Computer connects to Shopify, Attentive, Wicked Reports, GA4 and Xero, and generates a daily ad report sent in Slack.
FOR AD AGENCIES Srinivas says Perplexity’s agents are running ads better than agencies; the evidence published so far is one customer’s test.
FOR AGENTIC COMMERCE This is an agent working for the seller; on the buyer side, Uber counted 4 AI-agent rides in a 320M-trip week.
The framework says: “SaaS is charged for access to tools. AGaaS charges for the execution of outcomes.”
As we read it, Perplexity is pitching Computer against the agency, a service that sells outcomes, rather than against ad software; its evidence so far is one customer’s revenue and return on ad spend. The framework predates this case study and does not discuss Perplexity.
The results come from one customer, in a pilot that Perplexity invited the company to join and funded, and are reported by Perplexity. The story does not give the dates, the length of the side-by-side comparison, the dollar revenue or the spend on the non-AI ads, and it does not identify the agency the X post refers to. We have not seen independent data.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Perplexity says its Computer agent, given a $30,000 Perplexity-funded Meta budget, brought Get Maine Lobster 3.5 times the revenue of its usual ads on 2.1 times the spend, with 63% higher return on ad spend. It is a single-customer claim, and the customer story compares the agent with the company’s own non-AI ads rather than a named agency.
A note on sourcing. All results are Perplexity’s claims for one customer, from its 4 November 2026 customer story on Get Maine Lobster and its post on X; the $30,000 test budget came from Perplexity, the story says. Srinivas’s remark is from his post on X. The revenue-per-dollar ratio is our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
