Skip to main content
PandaCodeGen.
Back to Insights

Research

ChatGPT out-earned our Google Ads by 6.7x, March to August 2026

Over 1 March to 18 August 2026, a store we operate took 760 orders worth $271,620.61. ChatGPT accounted for $11,614.78 of it and Google Ads for $1,733.35. AI assistants together produced 63 orders and $16,783.76, which is 6.2% of all revenue.

Panda Patches is founder-affiliated. Owned and operated by a PandaCodeGen co-founder. We publish the relationship because it changes how you should read the numbers. Figures read on 2026-08-18.

Hassan Jamal·Published ·Updated ·11 min

How we recorded AI discovery

Referrer-only attribution can miss AI-assisted discovery. Referrers and campaign parameters vary by link and client. A later direct visit or brand search may not retain the earlier AI interaction. This dataset cannot measure how often that happens elsewhere, so missing attribution alone does not prove a large hidden AI audience.

Panda Patches records surviving referrer signals and customers' answers to “How did you find us?” in its quote and CRM workflow. Those inputs feed the recorded source value used for these orders. A customer-declared discovery source is useful evidence, but it is not the same as a captured referral click.

A practical tracking step

Add a discovery question to your enquiry form and keep the answer alongside any captured referrer or campaign signal. Review the records over a stated period. Missing referral data alone does not show whether a customer first encountered the business in an AI answer.

These figures describe recorded discovery sources for one store, not how much AI traffic other businesses receive.

Revenue by source, 1 March to 18 August 2026

Every row Panda Patches recorded, including the ones that produced nothing. Sorted by revenue. AI assistants are marked so you can find them without hunting.

Orders and revenue by acquisition source, 1 March to 18 August 2026
SourceOrdersRevenueShare
Repeat Order128$61,020.5122.5%
Google152$49,083.8418.1%
Facebook129$34,201.8512.6%
Facebook Ad106$28,068.9110.3%
Other4$24,480.009.0%
Direct49$18,021.666.6%
ChatGPT (AI assistant)51$11,614.784.3%
Instagram39$9,995.003.7%
Tawk.to27$8,960.003.3%
WhatsApp25$7,760.002.9%
RingCentral10$3,185.001.2%
Unknown7$3,160.001.2%
Claude (AI assistant)4$2,520.000.9%
Referral4$1,850.000.7%
Google Ad5$1,733.350.6%
Brave2$1,499.770.6%
Perplexity (AI assistant)2$990.000.4%
Google AI Overview (AI assistant)1$850.000.3%
DuckDuckGo4$732.760.3%
Gemini (AI assistant)4$673.980.2%
WEBSITE2$435.040.2%
Checkout2$346.900.1%
Bing2$302.260.1%
Meta AI (AI assistant)1$135.000.0%
TikTok0$0.000.0%
YouTube0$0.000.0%
LinkedIn0$0.000.0%
Total760$271,620.61100%

Both totals exclude Web Checkout (48 orders, $10,295.11), which the Panda Patches operations platform treats as a fulfilment route rather than an acquisition source and already excludes from its own revenue total. Excluding it from the order count too keeps both figures describing the same population; mixing them was an error caught by summing the rows against the declared totals.

Re-read on 19 November 2026, after the store started taking payment-link orders' source from the customer's earlier quote, chat or order: one order in this window (9 August, $114) moved from Direct to Google Ads. On that reading Google Ads is 6 orders and $1,847.35, and ChatGPT out-earned it by 6.3×.

What each AI assistant produced

ChatGPT did the volume, but it is not the whole picture. Six assistants appear separately, including Google AI Overview, which most analytics setups cannot distinguish from ordinary Google organic traffic at all.

Orders, revenue and average order value by AI assistant
AssistantOrdersRevenueAvg order
ChatGPT51$11,614.78$227.74
Claude4$2,520.00$630.00
Perplexity2$990.00$495.00
Google AI Overview1$850.00$850.00
Gemini4$673.98$168.50
Meta AI1$135.00$135.00

The average order across the whole store is $357.40. ChatGPT sits below that at $227.74, so it brings more orders of smaller size. Claude sits well above it at $630.00. That Claude figure rests on 4 orders and proves nothing. It is here because leaving it out would be selective, not because four orders support a conclusion. Treat it as a thing to watch, not a finding.

The channels that produced nothing

TikTok, YouTube, LinkedIn each recorded zero orders and zero revenue across the whole window. Those are real zeroes, published as zeroes.

We include them because a table showing only the sources that worked is an advertisement. The useful version shows what was tried and did not pay, which for this store is the entire short-form video category over six months.

Separate Google AI visibility evidence

Google reports visibility, not the source of these orders. Google Search Console, Performance, Generative AI features (Beta) for the same property recorded 48,900 impressions over 3 months to 18 August 2026, rising from roughly 400 a day in mid-May to 1121 on 2026-08-12.

The store's referrer and customer-declared records describe attributed orders and revenue. Google's report describes appearances in its own AI search features. These are separate measurements: the impressions do not verify ChatGPT order attribution, the revenue ratio or the effect of any particular website change. Impressions in Google's generative surfaces, not clicks, orders or revenue. They do not verify which assistant a customer used or the revenue attributed to it.

Month by month since the study

From June to October 2026, Panda Patches averaged $8,139.02 a month in revenue attributed to AI assistants. Individual months ranged from $3,970.00 to $12,296.49. There was no advertising spend on those AI referrals; the website, content and maintenance work behind them still took investment. Attribution works the same way as in the study above: captured referral signals plus customers' answers to “How did you find us?”.

Revenue attributed to AI assistants by month, as recorded on 19 November 2026
MonthAI ordersAI revenueShare of all revenue
March 20261$170.000.5%
April 20261$180.000.3%
May 202610$2,223.988.1%
June 202617$5,075.208.7%
July 202615$3,970.008.7%
August 202640$10,027.8413.8%
September 202625$12,296.4915.5%
October 202629$9,325.5511.3%
November 2026 (1–19, partial)15$4,246.599.8%

These rows were read from the store's portal on 19 November 2026, one calendar month at a time. An order counts under the source recorded on that order, so a returning customer who came back through an AI assistant counts as AI for that order. Read on that date, the portal shows $15,803.76 in AI-attributed revenue for 1 March to 18 August 2026, against the $16,783.76 read on 18 August 2026 and used above. Between readings, some orders' recorded source changed: on 19 November 2026 the store began taking payment-link orders' source from the customer's earlier quote, chat or order, and staff corrected a few sources by hand. We show both readings with their dates. The monthly average uses the calendar-month figures recorded on 19 November 2026. September's total is $2,950 lower than on 27 October: two order amounts were corrected (one from $2,540 to $120) and one order was refunded.

Separately, Google Search Console recorded about 78,900 impressions in Google's AI Overviews and AI Mode from 26 July to 25 October 2026. An impression is a link shown in an AI answer, not a visit or an order, and this window is later than the one reported above.

What a customer actually typed

One customer explained where she had come from without being asked. She had gone to ChatGPT and asked for patch companies with no or low minimums, and the store came up.

The query is the useful part. She did not search the category, she searched a constraint. Nobody wins that phrasing by publishing a page about custom patches; you win it by stating your minimum order quantity somewhere a model can read it. Every product decision the business had already made and never written down plainly is a query like that one going to a competitor. We are quoting her anonymously because permission to use her name is hers to give and had not been given when this was published.

How this compares to a larger study of brands

A 94-brand study found the same thing about order value that a single store found here. Visibility Labs analysed twelve months of Google Analytics data from 94 seven and eight-figure ecommerce brands across 2025, comparing 9.46 million non-branded organic sessions against 135,000 ChatGPT referral sessions. ChatGPT traffic converted at 1.81% versus 1.39% for non-branded organic, but its average order value was $204 against $238 for organic, a 14.3% gap, leaving revenue per session 10.3% higher rather than 31%.

Our single store shows the same direction on the part that overlaps. ChatGPT's average order here is $227.74 against a site average of $357.40. Two samples with nothing in common, one of 94 brands and one of 760 orders, both finding that AI assistants send smaller orders than the channel they are compared against. That is worth more than either finding alone, and it is the reason the AOV column is on this page at all.

This dataset records revenue by individual assistant. The cited study compares ChatGPT with organic search. Our store's records also separate Claude, Perplexity and Google AI Overview where a referral signal or customer-declared discovery identifies them. That provides detail for this business; the classification depends on the information recorded, and it does not establish that equivalent datasets exist nowhere else.

Do visitors from AI tools convert better?

It depends on the kind of business. Orbit Media looked at 97 B2B lead-generation sites and found that visitors from AI were about 3x more likely to become leads than other organic traffic (checked November 2026). Google still sent those sites about 100 times more traffic than ChatGPT. Microsoft Clarity studied more than 1,200 publisher and news sites and found that AI visitors signed up and subscribed at higher rates than visitors from search, direct or social, though AI sent less than 1% of their traffic (checked November 2026). The biggest ecommerce study we found points the other way. Across 973 online stores, ChatGPT referrals converted worse than Google search (checked November 2026). Results differ by business type, so check your own numbers before you move budget.

What this does not establish

This is one business, in one vertical, over one window. One business, one vertical (custom embroidered and printed patches), one window. These are not industry benchmarks and no claim is made that they generalise. A store selling custom patches to people who need small runs is close to an ideal case for an assistant that answers constraint-shaped questions. Your category may behave nothing like it.

What the data supports

  • Revenue attributed to each AI assistant across one store and one six-month window
  • That ChatGPT out-earned Google Ads here by 6.7x, from the same table and the same window. Read that with the denominator in view: Google Ads is 5 orders and $1,733.35 in this window, so the ratio rests on a very small paid-ads base and one more order would move it
  • That average order value from AI assistants sits below the site average, which a 94-brand study found independently
  • That referrer-only tracking undercounts AI referrals, which is why these figures are not zero

What it cannot tell you

  • Lead-to-order conversion. The lead and order records do not share a window and repeat orders inherit the original source, which produced a rate above 100% for two sources
  • Anything about your category. One business, one vertical, one window
  • That AI beat search. It did not. Google organic is $49,083.84 across 152 orders in this same table, several times ChatGPT's $11,614.78. The headline compares ChatGPT with paid ads only, because that is the line with an invoice attached
  • A clean average. Average order value of $6,120 against a site average of $357.40, almost certainly bulk or wholesale. Removing it moves the AI revenue share from 6.2% to 6.8%, so it works against the headline rather than for it.
  • Exact attribution. Someone who used ChatGPT and then searched Google may report either one

What we would do with this if it were your business

Add the how-did-you-hear question to your enquiry form today. It is the cheapest change on this list and without it none of the rest is visible. Then wait a quarter before drawing any conclusion, because a single large order will swing a small channel's share by several points.

Then write down the constraints you already satisfy. Minimum order quantity, turnaround, materials, what you will not do. Those are the terms the questions are shaped around, and most businesses keep them in a salesperson's head. Our answer engine playbook covers the structural side, and agentic browsing covers what an assistant has to be able to load before any of it counts.

Can your site tell you this?

Most cannot, and the gap is usually one form field and a place to put the answer rather than a new analytics platform. Our custom engineering work builds the reporting layer that makes this visible, on the same stack this store runs on.

Frequently asked questions

Do AI assistants like ChatGPT actually generate sales?
This store recorded sales attributed to AI discovery using captured referrers and customer-declared sources. Across 1 March to 18 August 2026, a store we operate recorded 63 orders worth $16,783.76 attributed to AI assistants, out of 760 orders and $271,620.61 total. That is 6.2% of all revenue, or 9.2% of acquisition revenue once repeat orders are excluded. Recorded revenue attributed to ChatGPT was $11,614.78. This is one business in one vertical, so it is evidence that it happens, not a benchmark for your business.
Why do most businesses report almost no AI traffic?
Referrer-only tracking can miss AI-assisted discovery. Links and clients can carry referrers or campaign parameters, while later direct visits and brand searches may lose that context. This does not establish the size of an unmeasured audience. A business measuring referrer alone will see close to zero and conclude AI sends no traffic. We capture referrer where it survives and ask the customer where it does not, which is the only reason these numbers are not zero.
How much revenue did ChatGPT generate compared to Google Ads?
ChatGPT produced $11,614.78 from 51 orders. Google Ads produced $1,733.35 from 5 orders. That is 6.7 times more revenue from ChatGPT, from a channel with no ad spend attached to it, over the same six-month window in the same store.
Which AI assistant sends the most valuable customers?
By volume, ChatGPT, with 51 of the 63 AI-attributed orders. By order value, Claude, at an average order value of $630.00 against a site average of $357.40. That Claude figure rests on four orders, which is far too small a sample to rely on, and it is reported here with the sample size attached rather than as a finding.
Do AI referrals have a lower average order value?
Yes, on the evidence available. In this store ChatGPT's average order was $227.74 against a site average of $357.40. Visibility Labs found the same direction at much larger scale, analysing 94 seven and eight-figure ecommerce brands across 2025: ChatGPT average order value of $204 against $238 for non-branded organic, a 14.3% gap. In that study ChatGPT still produced more revenue per session, 10.3% more, because it converted better. Two independent samples agreeing on the direction is stronger than either on its own.
Is there independent confirmation of this, or just your own tracking?
Not independent confirmation of the revenue. Google Search Console's Generative AI features report for the same property recorded 48,900 impressions over three months to 18 August 2026, rising from roughly 400 a day in mid-May to 1,121 on 12 August. That is a separate record of visibility in Google's AI surfaces, not clicks, orders or revenue, so it does not verify which assistant a customer used or the revenue attributed to it.

Method

Source: Custom operations platform built on Next.js, Sanity, Supabase and Square. Attribution: Referrer captured where it survives; customer-declared discovery collected through the How did you find us question in the quote and CRM workflow. Both inputs feed the recorded source value. Customer-declared discovery is distinct from a captured referral click. Since 19 November 2026, orders paid through a Sales payment link take their source from the customer's linked or earliest quote, chat or order; where none exists and the payment page carries no source, they are recorded as Unknown (payment link) rather than Direct. Window: 1 March to 18 August 2026, read on 2026-08-18. Relationship: Founder-affiliated. Owned and operated by a PandaCodeGen co-founder. Corroborating instrument: Google Search Console, Performance, Generative AI features (Beta), 3 months to 18 August 2026.

The underlying figures live in a single data file in this site's repository, and every percentage on this page is computed from it at build time rather than typed into a sentence. If a row and a total ever disagree, the build fails rather than publishing two numbers that contradict each other.