A Seed Round for the Invisible Shelf
On July 3, a New York startup called Nudge announced a 1.1 million dollar pre-seed round and, with it, a product aimed at one of the least understood problems in retail: whether AI shopping agents recommend your product at all. Founded by Kanishka Thakur and Gaurav Rawat, Nudge is building what it calls commerce infrastructure for AI shopping, a layer that helps consumer brands get surfaced, cited, and ultimately bought inside assistants like ChatGPT, Claude, Gemini, and Perplexity. The round is small, but the thesis is large: as shoppers increasingly start their journeys in a chat window, the brands that cannot be read by an agent simply cease to exist for a growing slice of demand.
The investor list signals the theme is drawing serious attention despite the modest check size. The round was led by s16vc, a fund backed by the founders of Miro, Datadog, and Intercom, with participation from Antler and operators from Shopify, Nutanix, and Postman. "In rapidly evolving markets, founder velocity is often the most important predictor of success," said Aleks Shamis, co-founder and managing partner at s16vc, framing the bet as much on the team's speed as on the specific product. That framing fits a category still being defined, where the winners will be decided less by today's feature set than by who iterates fastest as the assistants and their commerce protocols keep changing.
Measuring Visibility Nobody Could See
Nudge's first job is measurement, because most brands have no idea how they appear inside AI answers. The platform tracks visibility at the individual stock-keeping unit level, monitoring how a brand and each of its products are ranked, recommended, and cited across ChatGPT, Claude, Gemini, Perplexity, and AI Overviews. This is the agentic successor to search-engine optimization, and it exposes a blind spot that traditional analytics never touched. A brand can dominate classic search results and still be invisible to an assistant that summarizes the web into a single recommendation, and until now there was no reliable way to even see that gap, let alone close it.
"The AI platform becomes the salesperson. The challenge is making sure the product page still matches what the user was actually looking for," said Thakur, describing the disconnect the company is trying to fix. That sentence captures why measurement alone is not enough. Knowing you are absent from an assistant's answer is useful only if you can act on it, and Nudge pairs its visibility tracking with catalog enrichment that restructures product data so agents can actually parse it. The company maps schema and conversational attributes and pushes fixes directly into storefronts running Shopify, WooCommerce, and Salesforce Commerce Cloud, turning a diagnostic into a remediation.
From Recommendation to Purchase
The third layer is where Nudge tries to close the loop from discovery to revenue. Getting recommended is only half the battle, and the company builds what it calls shoppable funnels that catch shoppers arriving from AI discovery and match the experience to their expressed intent, inside the brand's existing storefront rather than a rented marketplace. "Being recommended is the start. Being bought is the win," as Thakur puts it, and the distinction is the crux of the pitch. Plenty of tools promise to boost mentions inside assistants; far fewer connect those mentions to a checkout in a way a brand can measure and control on its own property.
Nudge claims early customers have seen meaningful movement, reporting improvements of up to four times in AI visibility and order-volume increases of roughly 24 percent. Those figures come from the company and cover a small, self-selected set of early users, so we would not treat them as a category benchmark. Still, the direction is credible. If assistants are steering an expanding share of discovery, then even modest gains in how an agent reads and recommends a catalog can translate into real orders. The interesting question is durability: whether these lifts persist as every competitor adopts similar tooling and the assistants keep changing how they rank and cite products.
Betting on Protocols, Not Platforms
Nudge is making a deliberate architectural bet by aligning its catalog work with emerging commerce standards rather than any single assistant. The company structures product data to match protocols like the Agentic Commerce Protocol and the Universal Commerce Protocol, the same standards that larger players such as Google, Shopify, and the card networks have been pushing. That choice matters because it hedges against platform risk. A brand that optimizes only for today's version of ChatGPT is exposed the moment the model changes; a brand whose catalog speaks the underlying protocols is positioned to appear wherever those protocols are honored.
"The search part becomes table stakes. The bigger opportunity is controlling the actual commerce experience from discovery to purchase," Thakur argues, and the protocol alignment is how Nudge intends to reach for that larger prize. We think the framing is right even if the execution is unproven. The value in agentic commerce will not accrue to whoever games a single assistant's ranking this quarter, but to whoever owns the durable, machine-readable representation of a brand's catalog across every surface. That is a harder, more infrastructural problem than answer-engine optimization, and it is the one Nudge is claiming, ambitiously, for a 1.1 million dollar round.
Why Investors and Brands Are Paying Attention
The bet on the team is explicit, and it reflects how early this market still is. With a check this size, investors are not underwriting a proven revenue engine but a wager that a fast-moving founding team can define a category before larger incumbents wake up to it. That is a high-variance strategy. The same agentic-visibility problem Nudge is chasing has attracted a crowd of startups, and the commerce clouds and assistant makers could fold much of this functionality into their own platforms. Nudge's defense is focus and speed, betting it can go deeper on brand-side commerce control than a generalist platform will bother to.
For enterprise brand and ecommerce leaders, the takeaway is not whether to buy Nudge specifically, but to recognize the gap it is naming. Most organizations still have no instrumentation for how they appear inside AI assistants, no owner for that surface, and no process for fixing catalog data so agents can read it. Those are governance and data questions that will outlast any single vendor. Nudge's raise is a useful prompt to treat agentic visibility as a real channel with a budget and an owner, rather than a curiosity, because the shelf may be invisible but, as this round argues, the revenue moving across it is not.
