# Work-Bench leads Petra Labs' $5.2 million seed for AI search attribution > Source: > Published: 2026-07-22 20:25:48+00:00 [Sami Akkawi](https://www.linkedin.com/posts/sami-akkawi-815b84130_the-way-people-discover-and-learn-is-shifting-activity-7384995865263845376-Bo0K?ref=runtimewire), [Rishi Penmetcha](https://www.rishipenmetcha.com/?ref=runtimewire) and [Ramie Khoury](https://happenstance.ai/u/ramie-khoury?ref=runtimewire) announced a $5.2 million seed round for [Petra Labs](https://www.petralabs.com/?ref=runtimewire) on Wednesday, putting new capital behind their attempt to connect brands' visibility in AI-generated answers to traffic, pipeline and revenue. [Work-Bench](https://www.work-bench.com/?ref=runtimewire) led the round, with participation from [Afore](https://www.afore.vc/?ref=runtimewire), [Pathlight](https://pathlight.vc/?ref=runtimewire) and unnamed strategic angels, according to the [July 22 Business Wire announcement](https://www.businesswire.com/news/home/20260722868575/en/?ref=runtimewire). Petra Labs did not disclose its valuation, customer count, revenue or the names of its angel investors. The three founders started the New York-based Petra Labs in 2025 around a problem that sits downstream from the rush into answer engine optimization, or AEO. Marketing teams can buy tools that track whether ChatGPT, Claude, Gemini and other AI systems mention a brand. They still have to decide which measurements matter, how to change the answers and whether any resulting visibility produces business. Akkawi arrived at that problem after spending five years at Strategy&, where Petra Labs says he advised private equity firms on over 50 acquisitions involving marketing data, technology and services. Penmetcha had built large-scale data systems and AI agents at workforce marketplace Traba. Khoury led Traba's growth team, working across enterprise sales and operations. Their backgrounds explain Petra Labs' operating model. Akkawi brings the buyer and marketing-services view; Penmetcha owns the data layer; Khoury brings the enterprise execution required to sell a high-touch product. Petra Labs is applying that mix to a young market where the data remains noisy and the work extends well beyond installing software. ### Selling execution with the software Petra Labs tracks prompts, brand presence, ranking, citations, sentiment and competitors across AI platforms. Its [platform description](https://www.petralabs.com/intelligence/detailed-overview-of-petra-labs-platform?ref=runtimewire) says Petra Labs repeatedly runs natural-language prompts because large language model outputs vary between attempts. Petra Labs then connects that measurement layer with customers' analytics and revenue systems through custom attribution models. The second half of the product is human labor. Petra Labs embeds AEO specialists who select prompts, interpret citation data and carry out work across owned content, media coverage and social channels. Akkawi described the goal in the funding announcement as owning customers' AI search outcomes from measurement through execution. That services component is central to Petra Labs' economics. On its [pricing page](https://www.petralabs.com/intelligence/petra-labs-pricing?ref=runtimewire), Petra Labs says most customers pay between $15,000 and $50,000 per month, with engagements customized around industry, competitive intensity, internal systems and the amount of execution required. Those prices put Petra Labs closer to an embedded enterprise partner than a self-service marketing dashboard. They also create a scaling test. Services can help a young vendor learn quickly, secure larger contracts and produce the outcomes customers actually want. Petra Labs will still have to show that software and automation can absorb a growing share of delivery as the customer base expands. Petra Labs identifies [Zapier as a customer](https://www.petralabs.com/customers/an-end-to-end-aeo-partnership-with-zapier-from-strategy-to-attribution?ref=runtimewire). Petra Labs has not disclosed the number or size of its other customer contracts. ### AI visibility is a moving target The technical problem gives Petra Labs room to sell expertise. AI answers can change across repeated runs, prompt wording, user accounts and time. An [April 2026 research paper](https://arxiv.org/abs/2604.07585?ref=runtimewire) found that one-off observations are unreliable and argued that AI search visibility should be measured as a distribution built from repeated samples. Attribution is harder again. A buyer may encounter a brand in an AI answer, return later through a direct visit or conventional search, and never create a clean AI referral event. Petra Labs says its custom models combine AI visibility, site traffic and customers' internal data to reconstruct that path. The funding release does not provide independent results showing how accurately those models separate AI search's contribution from other marketing activity. That missing proof is the core commercial hurdle for the category. AEO budgets will eventually face the same scrutiny as paid search, content and public relations. Visibility scores can start a sales conversation, but finance and marketing leaders will keep spending only when vendors can connect the work to pipeline or revenue with defensible measurement. ### Work-Bench backs an enterprise delivery model Work-Bench's lead investment fits Petra Labs' decision to build for large organizations from the start. Jessica Lin, Work-Bench co-founder and general partner, said in the release that Petra Labs stood out by connecting AI visibility insights to revenue while taking responsibility for execution. Petra Labs is entering a market that has already attracted much larger checks. New York-based [Profound raised a $96 million Series C](https://www.tryprofound.com/blog/profound-raises-96m-series-c?ref=runtimewire) at a $1 billion valuation in February. Berlin-based [Peec AI raised a $21 million Series A](https://peec.ai/blog/we-raised-21m-series-a-to-help-brands-win-in-ai-search?ref=runtimewire) in November 2025. [Evertune raised a $15 million Series A](https://www.evertune.ai/resources/insights-on-ai/evertune-raises-15-million-series-a-to-scale-its-ai-marketing-and-discovery-platform?ref=runtimewire) in August 2025, and Sitecore [acquired Scrunch](https://www.sitecore.com/company/newsroom/press-releases/2026/06/sitecore-acquires-scrunch-to-help-brands-influence-discovery--and-buying-decisions?ref=runtimewire) in June. Most competitors began with monitoring and analytics. Petra Labs is betting that enterprise buyers will pay a higher contract price for one accountable vendor to measure, execute and attribute the work. The $5.2 million seed gives Akkawi, Penmetcha and Khoury capital to turn that delivery model into a repeatable product before larger software platforms bundle similar capabilities into existing marketing suites.