What’s funding like in the AI Shopping market right now?

All funding deals in the AI shopping market over the last 24 months

Below is the table listing all the deals. You can find our methodology at the end of this page.

If you want a deeper understanding of the market and its current dynamics, get our report covering the AI Shopping Market.

Is AI shopping funding actually weaker right now?

AI shopping funding is weaker in dollars today, even though startups are still finding investors almost as easily as they were a year ago.

We counted 32 financings in the latest 12 months versus 31 in the previous period, with exactly 31 funded companies in each. There has been no real drop in the number of businesses receiving checks.

The size of those checks has changed much more. Disclosed capital fell from at least $291.2 million to $181.8 million, while the median round dropped 35.3%, from $4.50 million to $2.91 million. The average financing fell even harder, from at least $9.39 million to $5.68 million.

Funding has also become more regular. Every rolling monthly window in the latest period contained at least one deal, compared with two empty windows previously. AI shopping is still producing a steady flow of financings; it is simply doing so with far less money at the upper end.

The market is therefore fairly specific right now: funding breadth has held up, while capital intensity has weakened considerably.

Metric	Latest 12 months	Previous 12 months	Change
Deals	32	31	+3.2%
Funded companies	31	31	0%
Total disclosed capital	$181.8M	≥$291.2M	≤-37.6%
Median round	$2.91M	$4.50M	-35.3%
Average round	$5.68M	≥$9.39M	≤-39.5%
First financings	12	13	-7.7%
Follow-on financings	20	18	+11.1%
Verified Series B deals	0	2	-2
Monthly windows with zero deals	0	2	-2

Is one giant AI shopping round making the funding decline look worse than it really is?

Last year's Gensmo financing makes the AI shopping funding decline look more dramatic, but removing large rounds still leaves a clear slowdown.

Gensmo disclosed more than $60 million in funding during the previous period. Because the exact amount was never published, the previous-period total of $291.2 million is itself a floor.

We ran the comparison again after excluding every deal above $50 million. Current funding stays at $181.8 million because there are no recent rounds above that level. Previous funding falls to $231.2 million. The gap is still 21.4%.

The same exercise also shows how much less dependent the market has become on huge checks. The previous period included Gensmo above $60 million, FERMÀT at $45 million, Cimulate at $28 million and Vizit at $25 million. Phia's $35 million Series A is the only recent financing above $20 million.

Concentration fell along with round size. The top three recent deals represent 36.9% of capital, compared with at least 45.7% previously. The top five share declined from at least 59.8% to 50.9%.

Gensmo clearly amplified last year's total. The underlying slowdown is still there once Gensmo is taken out.

Are AI shopping startups still getting funded?

Yes. AI shopping startups are still getting funded at almost exactly the same rate today, which is why the drop in capital should not be confused with founders suddenly losing access to venture money.

Thirty-one companies received financing in each period. Deal count even edged higher from 31 to 32, although InsightX accounts for the extra event because two closes of the same Series A appear separately in our dataset.

The monthly pattern is also healthier than the annual capital number suggests. We averaged 2.67 deals per monthly window recently versus 2.58 before, with the same median of 2.5. More importantly, recent activity never disappeared for a full monthly window.

It is a slightly odd funding environment. Investors are still meeting companies, leading rounds and creating new funded startups. What has become difficult is raising a large round.

For founders, that distinction is useful. AI shopping has not become a dead venture category. The market currently supports plenty of Seed activity and some meaningful Series A rounds, while the growth-capital layer has become much thinner.

How much smaller are AI shopping rounds now?

AI shopping rounds have become much smaller, especially once a company needs more than $20 million.

Almost two-thirds of recent deals were below $5 million, up from 58.1% in the previous period. Those small financings now account for 23.5% of total capital, twice their previous 11.6% share.

The real center of the market has moved into the $5 million to $20 million range. That band now absorbs 57.3% of all funding, compared with 34.2% a year earlier.

Above $20 million, the picture changes sharply. Deals between $20 million and $50 million fell from 9.7% of transactions to 3.1%, while their share of capital dropped from 33.7% to 19.3%. The $50 million-plus band disappeared entirely.

That is why the $2.91 million median tells us more than the headline funding total. Smaller rounds are becoming a larger part of the market, while the layer that used to produce $25 million, $30 million or $45 million financings has almost vanished.

Deal size	Recent deal share	Previous deal share	Recent capital share	Previous capital share
Below $5M	65.6%	58.1%	23.5%	11.6%
$5M to <$20M	31.3%	29.0%	57.3%	34.2%
$20M to <$50M	3.1%	9.7%	19.3%	33.7%
$50M to <$100M	0%	3.2%	0%	≥20.6%
Deals above $20M	1	4	19.3%	≥54.3%
Deals above $50M	0	1	0%	≥20.6%
Largest round	$35M	>$60M	19.3% of capital	≥20.6% of capital

Are investors still backing new AI shopping startups?

AI shopping investors are still backing new startups, although first-time financings have become a little smaller and slightly less common.

We counted 12 first financings recently versus 13 previously. Their share of total deals moved from 41.9% to 37.5%, while the median initial round fell from $3.2 million to $2 million.

The enormous difference in first-financing capital needs context. Recent first rounds raised $39.9 million in total, compared with at least $108.7 million previously. Gensmo alone contributed more than $60 million to the earlier figure. Without that one financing, the previous total falls to roughly $48.7 million, much closer to today's $39.9 million.

Follow-ons moved in the other direction. We counted 20 recently versus 18 before, although their median size also dropped, from $4.86 million to $3.55 million.

Several companies did manage major step-ups. ReFiBuy went from $1.7 million to $13.6 million, Phia from $8 million to $35 million, shoppin' from $1 million to $3 million and Spangle from $6 million to $15 million.

Investors are still creating new funded companies and backing existing ones again. The tougher part comes when those companies try to turn progress into a much larger growth round.

Metric	Latest 12 months	Previous 12 months
First financings	12	13
First-financing share	37.5%	41.9%
First-financing capital	$39.9M	≥$108.7M
Median first financing	$2.0M	$3.2M
Follow-ons	20	18
Follow-on share	62.5%	58.1%
Follow-on capital	$141.8M	$182.4M
Median follow-on	$3.55M	$4.86M

Has growth-stage AI shopping funding dried up?

Growth-stage AI shopping funding has become very scarce for now: we found no verified Series B financing in the latest 12 months.

The previous period had two Series B rounds, FERMÀT at $45 million and Vizit at $25 million. Together they contributed $70 million, roughly a quarter of the previous period's disclosed capital.

Seed activity, meanwhile, stayed at exactly 24 deals. Series A slipped from five rounds to four, and the median Series A fell from $16 million to $9.43 million.

Four recent financings have an unknown stage, so we cannot conclude that every growth-stage transaction has disappeared. What we can say with confidence is that no current deal in our research has a verified Series B-or-later label.

That missing layer explains a lot of the current market. Early-stage companies continue to raise, and Phia still managed a substantial $35 million Series A. The usual progression into $20 million-plus growth rounds has become much harder to see.

Which parts of AI shopping are investors most excited about now?

AI Recommendations and Commerce Decision Support have gained the clearest momentum in AI shopping funding lately.

AI Recommendations increased from 9.7% to 12.5% of deals, while its share of capital jumped from 4.3% to 16.6%. The median financing more than doubled from $3.20 million to $6.93 million. Deal share, dollar share and typical round size all moved upward together, which is a stronger pattern than any one of those measures on its own.

Commerce Decision Support expanded from one transaction to five. Its capital share went from 4.8% to 19.1%, helped by companies such as eComID, CATCHES, Made With Intent, Replenit and SECONDSENSE. The comparison starts from a tiny previous base, so the percentage jump looks more spectacular than the sample really is.

AI Shopping Assistants followed a different path. Their share of deals dropped from 32.3% to 25%, while the median round climbed from $1.07 million to $3.61 million. Fewer assistant companies got checks, but the companies that did were generally raising more money.

AI Merchandising Tools still produce plenty of deals. Their funding, however, has shifted heavily toward smaller rounds.

Category	Deal share now	Deal share before	Capital share now	Capital share before	Median now	Median before
AI Merchandising Tools	31.3%	29.0%	19.1%	37.2%	$2.0M	$5.22M
AI Shopping Assistants	25.0%	32.3%	34.9%	33.8%	$3.61M	$1.07M
AI Recommendations	12.5%	9.7%	16.6%	4.3%	$6.93M	$3.20M
AI Product Search	15.6%	16.1%	10.3%	13.9%	$3.0M	$4.5M
Commerce Decision Support	15.6%	3.2%	19.1%	4.8%	$3.25M	$14.0M
Product Advice Engines	0%	9.7%	0%	6.1%	-	$1.24M

Why are AI recommendations and shopping decision tools getting more funding?

AI shopping investors are paying more attention to systems that can understand what a shopper wants and decide what to show next.

Malachyte raised $10 million around real-time behavior intelligence and personalization. The team came from Spotify's personalization work, and the product uses live shopper behavior to rank and recommend products based on current intent.

Spangle raised a $15 million Series A for a related problem. Its software adapts recommendations and shopping experiences as customers interact with a site rather than relying only on static segmentation.

eComID approaches the decision layer through shopper context. Its $17 million round came after the company said it was working with more than 60 brands and reaching around 20 million shoppers each month. The Shopping Passport carries information such as size, fit, taste and preferences between participating shopping experiences.

The connection is fairly intuitive. Retailers already have enormous catalogs and increasingly capable AI models. The valuable problem now is deciding which handful of products should appear for a particular person, in a particular context, at a particular moment.

That helps explain why recommendation funding looks stronger today. These products sit very close to the purchase decision, where better AI can potentially be tied directly to conversion, basket size and returns.

Is AI merchandising falling behind?

AI merchandising remains one of the busiest parts of the market, but investors are currently funding it with much smaller checks.

Merchandising represented 31.3% of recent deals versus 29% previously, yet its share of capital almost halved from 37.2% to 19.1%. The category median fell from $5.22 million to $2 million.

The newer companies also look somewhat different from the previous generation. ReFiBuy raised $13.6 million to help brands enrich and optimize product catalogs for AI shopping agents. Era raised $1.4 million around making products discoverable, trusted and purchasable inside AI assistants. Nudge is working on SKU-level visibility and catalog enrichment, while 40rty.ai focuses on making Shopify product data easier for AI shopping agents to understand and select.

Channel3 sits nearby with a $6 million Seed round for a universal AI-ready product graph. The company said its catalog already covered more than 50 million products when the round was announced.

So the merchandising category is increasingly becoming part of AI-commerce infrastructure. Catalog quality, product attributes, availability and context now determine whether an agent can understand and recommend a product at all.

That transition is active today, but most of it is still being financed at Seed scale.

Is Europe becoming a serious AI shopping funding hub?

Europe has become a genuine second center for AI shopping funding, with the number of European deals jumping from five to 12.

Europe's share of transactions increased from 16.1% to 37.5%, while its share of disclosed capital rose from just 3.1% to 29.3%. In dollar terms, European funding went from about $9.2 million to $53.3 million.

The geographic breadth makes the shift more convincing. Sweden produced eComID's $17 million round. The UK contributed CATCHES at $10 million, alongside Made With Intent and Wassist. France produced Lemrock and Dialog. Other recent deals came from Poland, Spain, Italy, the Czech Republic and Turkey.

North America still takes the majority of capital at 63.7%, so Europe has not displaced it. The gap has simply narrowed a lot.

The European change is one of the strongest structural shifts in the dataset because both company count and dollars moved together. This looks like a broader startup base, not one unusually large European transaction distorting the picture.

Are AI shopping investors pulling away from the market?

AI shopping investors are still showing up in roughly the same numbers, even as they write smaller checks.

Twenty-nine of the 32 recent deals disclosed a lead investor, almost identical to 28 of 31 previously. We identified 35 unique disclosed lead investors in the latest period versus 33 in the earlier one.

Two recent investors also appeared as lead more than once: Galion.exe and NewRoad Capital Partners. No investor did so in the previous window.

That investor breadth fits what we see elsewhere in the market. There are plenty of firms willing to finance an AI shopping company. The missing money is concentrated much further up the round-size curve.

In practice, the current environment is selective. A startup can still persuade investors that AI shopping is worth funding. Convincing them to deploy $25 million, $40 million or $60 million has become much rarer.

Are strategic investors becoming more important in AI shopping?

Strategic investors are becoming a little more visible in AI shopping, especially where the investor already understands the commercial problem.

Strategic participation rose from three deals to four, or from 9.7% to 12.5% of transactions. The increase is modest, so it does not look like a structural takeover of the funding market.

The individual relationships are more interesting than the percentage. Stadium joined eComID's $17 million round after already being one of the company's early retail customers. Sia invested in Lemrock as the consulting group expanded its work around enterprise agentic commerce. Gradient Ventures participated in Malachyte, while Rakuten Capital backed InsightX.

Stadium is probably the cleanest example of what strategic money can mean here. eComID had already moved from a pitch to a deployed product inside Stadium before the retailer became an investor.

That kind of customer-investor overlap tells us more than a corporate logo on a cap table. At least some strategic capital is following firsthand exposure to the technology.

Do AI shopping startups need more proof to raise money now?

The biggest AI shopping rounds today increasingly go to companies that can show real usage, customers or deployment, although the public data is not complete enough to turn that into a rule for every deal.

Phia is the clearest case. When the company announced its $35 million Series A, it said the app had passed one million users, partnered with more than 6,200 retail brands and grown revenue 11-fold since launch. Those are unusually concrete traction metrics for a company barely ten months after launch.

eComID also entered its $17 million round with substantial adoption. The company reported more than 60 participating brands and roughly 20 million monthly shoppers, while Stadium had already become a customer.

ReFiBuy provides another useful progression. Its first $1.7 million round came while the company was newly launched. When it returned for $13.6 million, ReFiBuy said its product was already in production with brands and retailers across beauty, fashion, electronics, consumer goods and home.

The pattern becomes less clear once we move beyond these larger transactions because public milestone disclosure is inconsistent. Many companies release a round amount and investor names while saying little about revenue, customers or usage.

At the top end, though, the pattern is hard to miss: strong recent rounds often come with tangible proof.

What are AI shopping companies raising money to do now?

The better-funded AI shopping companies are increasingly raising money to deploy their products more widely, improve the underlying infrastructure and win distribution.

Phia said its financing would support machine-learning infrastructure, product intelligence and personalization as the shopping agent scaled. eComID plans to enter more markets, expand its Shopping Passport, deepen sizing and discovery capabilities and build the team and infrastructure around it.

ReFiBuy said its $13.6 million Seed would expand its Commerce Intelligence Engine and accelerate go-to-market. The company has since kept pushing that infrastructure outward, including a recent integration that connects enriched product catalogs with the Claude Commerce Agent.

Malachyte's funding also targets distribution and team expansion around its real-time personalization technology.

Some of today's better-funded companies have already moved beyond the basic question of whether an AI shopping product can work. They are dealing with the next problems: better data, more integrations, stronger distribution and deployment across much larger customer bases.

We cannot calculate a market-wide shift from R&D to commercialization because use-of-funds information is not consistently disclosed. The company-level evidence still shows that scaling has become a much more visible part of the story around the largest recent rounds.

Is agentic commerce actually changing what AI shopping investors fund?

Agentic commerce is already changing the kinds of AI shopping companies getting funded, especially around product data, discovery and the infrastructure that connects merchants to AI assistants.

The shift became much more concrete when OpenAI launched the Agentic Commerce Protocol and Instant Checkout. OpenAI later expanded its shopping stack around richer product discovery and formalized merchant feeds through which businesses can submit product information for OpenAI services.

Google has been building in the same direction through Universal Commerce Protocol, an open standard spanning discovery, buying and post-purchase interactions. UCP is designed to connect AI shopping experiences directly with retailer systems.

Those platform changes solve an important part of the timing question. Startups such as ReFiBuy, Channel3, Era, Nudge, Lemrock and 40rty.ai now have a much easier commercial story to tell. Brands increasingly need product information that machines can interpret, compare, recommend and transact against.

ReFiBuy is especially useful to watch because the company has continued announcing integrations after its financing. Its latest work connects enriched catalogs to the Claude Commerce Agent and feeds shopper conversation context back into the catalog-optimization process.

Agentic commerce probably does not explain why total AI shopping funding fell. It does help explain why capital is rotating toward catalog intelligence, intent, recommendation systems and infrastructure that makes products usable inside AI interfaces.

Are the strongest AI shopping startups raising money faster?

Some of the strongest AI shopping startups are coming back for much larger rounds very quickly, but the sample is still too small to call this a broad market trend.

Among follow-on deals where we could establish the previous financing date reliably, the recent median time between rounds was 9.5 months versus 13.3 months in the previous period. Two-thirds of the observable recent group returned within a year.

The step-ups can be large. ReFiBuy raised about eight times its previous round, Phia roughly 4.4 times, shoppin' three times and Spangle 2.5 times.

There is plenty of variation underneath that headline. Velou's latest financing was smaller than its preceding round, and InsightX's two separate Series A closes produce another sub-1x comparison.

Only six recent follow-ons and three previous ones have sufficiently complete prior-round information for this calculation. That is nowhere near enough to claim that the entire AI shopping market is fundraising faster.

The useful takeaway is narrower: a small group of companies is moving through the funding cycle exceptionally quickly when investors see enough progress to justify a major step-up.

What really changed over the last 12 months?

The AI shopping funding market currently looks shallower and more selective rather than broadly weaker. Companies are still getting funded, investors are still participating and new startups continue to appear. The big change is what happens after that first check: rounds are smaller, verified Series B activity has disappeared, and far less money reaches the upper end of the market.

At the same time, the technology being financed has moved. Recommendations, real-time shopper intent, decision support, AI-ready product data and agentic-commerce infrastructure are taking a larger role. Europe has also become a much more credible second funding hub.

Rank	Structural change	What it tells us
1	AI shopping rounds got much smaller	The median fell 35.3%, and the decline remains 21.4% even after removing >$50M rounds.
2	Funding breadth held up	We found 32 deals and 31 funded companies versus 31 and 31 previously.
3	Growth capital became scarce	No verified recent Series B appears in our research, compared with two worth $70M previously.
4	New entrants still get funded	First financings slipped only from 13 to 12, although their median fell from $3.2M to $2M.
5	Money is rotating toward the decision layer	Recommendations and Commerce Decision Support gained substantial capital share.
6	Agentic commerce is reshaping the product stack	Catalog intelligence and machine-readable product infrastructure now have a much clearer commercial role.
7	Europe became a real second cluster	European deal count rose from 5 to 12 and capital increased from about $9.2M to $53.3M.

The misleading headline would be that AI shopping funding has collapsed because total capital fell by more than a third. That skips the most interesting part of the story. The same number of companies are still getting funded, while investors have become much less willing to provide large growth checks. At the same time, the companies attracting attention are changing as AI assistants become real shopping and distribution interfaces.