In a real 39-day test for one Auckland small business, ChatGPT Ads delivered 269 clicks and seven tracked website lead events for NZ$975.79. Google Search Ads delivered 48 clicks and one primary lead-form conversion for NZ$1,229.95. These are Kiwi Web Design’s own account results for 3 September to 11 October 2026, not a New Zealand industry average or a count of confirmed new customers.
This is a comparison of advertising in ChatGPT with Google Search Ads. It is not a test of ChatGPT’s answers against Google AI Overviews or AI Mode. Kiwi Web Design runs both channels for its Auckland web design service, so we can publish the aggregate numbers we observed in our own NZ-dollar accounts. Our conclusion is clear for this test: ChatGPT Ads was the better source of affordable, tracked lead opportunities. Whether those opportunities become qualified clients still needs to be checked against our enquiry records.
The real New Zealand campaign numbers
| Metric, 3 Sep–11 Oct 2026 | ChatGPT Ads | Google Search Ads |
|---|---|---|
| Spend | NZ$975.79 | NZ$1,229.95 |
| Impressions | 34,062 | 757 |
| Clicks | 269 | 48 |
| Click-through rate | 0.79% | 6.34% |
| Average cost per click | NZ$3.63 | NZ$25.62 |
| Platform-tracked lead goal events | 7 | 1 primary form conversion |
| Spend per tracked lead goal | NZ$139.40 | NZ$1,229.95 |
What stands out: ChatGPT Ads produced about 5.6 times as many clicks while spending NZ$254.16 less. Its average click cost was about one-seventh of Google’s. The 7-to-1 lead-goal comparison is striking, but the two platforms have different tracking and attribution systems. The spend-per-lead figures are platform-reported costs per tracked goal, not costs per qualified sales enquiry.
The Google campaign’s higher click-through rate should not be read as proof of better commercial performance. Search ads and conversational ad placements have different exposure and intent patterns, so their impression and click-through rates are not like-for-like measures of lead quality.
What we ran, and how we collected the data
Kiwi Web Design is an Auckland-based website design business advertising its own service. The Google side is one Search campaign in our KWD Google Ads account. Its campaign was enabled when we checked the account, used Maximize Conversions and had a NZ$30 daily budget. We pulled the figures read-only through the Google Ads API, for 3 September through 11 October in the account’s confirmed Pacific/Auckland timezone.
The ChatGPT side is the combined result of two campaigns in our ChatGPT Ads Manager account. The earlier clicks campaign spent NZ$170.85, generated 52 clicks and recorded no goal conversions. The newer Auckland Website Leads campaign spent NZ$804.94, generated 217 clicks and recorded all seven lead goals. We read the campaign totals and conversion-event setting directly from Ads Manager on 12 October 2026. The newer campaign was active at the time of that check with a NZ$25 daily spend limit.
Both accounts report in NZ dollars, and both date filters cover 3 September to 11 October as platform reporting dates. Ads Manager did not return an account timezone, so the exact start and end hours may differ from Google Ads. Campaign settings, creative, delivery systems and landing pages were also not identical. This is a first-party case study, not a controlled experiment or a claim that all NZ advertisers will get these rates.
Why ChatGPT Ads looked better for our small business
More visits for less money
The most robust comparison is spend and traffic: 269 ChatGPT clicks for NZ$975.79, against 48 Google Search clicks for NZ$1,229.95. A lower click price gives a small business more chances to test its offer and landing page within a limited budget. It does not make every click equally valuable, but it materially changed the amount of traffic we could buy in this period.
More attributed lead events in the platform reports
ChatGPT Ads reported seven click-attributed lead_created events, using a 30-day click window with view-through attribution disabled. The configured goal is Kiwi Web Design Website Lead. OpenAI’s conversion measurement guidance explains that an event must match a configured campaign goal and be attributable within the selected window to appear as a conversion.
Google Ads reported one primary conversion for KWD LP Lead Submit - Ads tag. That is the lead-form action counted in Google’s main Conversions column. Google also showed six “all conversions” in total: the primary form event, one secondary lp_lead_submit event, one secondary general form_submit event and three secondary phone_click events. A phone-link click is an expression of interest, not a confirmed phone conversation. The form events may refer to overlapping activity, so six all-conversions must not be presented as six separate enquiries. Google’s primary and secondary action guidance and All conversions explanation describe why these columns differ.
A useful fit for considered purchases
Buying a website often involves comparing needs, timelines, budgets and providers. ChatGPT Ads can present a sponsored option in a relevant conversational context. OpenAI says context hints describe suitable customer situations; they are not exact-match keywords or guaranteed delivery rules. That makes the channel worth testing for a considered service like web design. Our account numbers show an outcome, not proof that conversational context caused it.
Where Google Search still has a role
Google Search can reach someone who explicitly searches for a web designer in Auckland. We would not dismiss that intent or advise every NZ business to replace Search with ChatGPT. Our own Search campaign still produced a primary form event, and its 6.34% click-through rate shows people engaged with the ad when it appeared.
The cost needs attention. In the same 39-day period, Google Search averaged NZ$25.62 per click. The reported search terms “website design auckland” and “web design auckland” spent NZ$379.18 combined with no primary conversion attributed to those terms in this window. That is a reason to review queries, bids, landing-page fit and lead quality before increasing the budget. It is not enough evidence to declare those searches worthless.
For our current campaign, the limited primary conversion signal also makes the Maximize Conversions bidding result worth examining. We would first reconcile the actual enquiry and inspect search-term intent, rather than assume that a larger daily budget will fix the cost per lead.
What this comparison cannot tell you
- Seven ChatGPT events are not seven confirmed customers. The platform reports attributed website events. We have not published a verified count of qualified enquiries, proposals or sales from them.
- The conversion definitions are not identical. Google’s one primary Ads-tag form event is the closest headline comparison; its five additional secondary actions are not five additional confirmed leads.
- This is one NZ business and 39 reporting days. Results will vary by offer, audience, creative, competition, destination and tracking. The figures are not a national benchmark.
- The reporting windows are not hour-for-hour identical. Google Ads confirmed Pacific/Auckland; Ads Manager did not provide its account timezone. Recent spend and attributed conversions can also be revised as platforms finish processing.
- No return on investment is established. A tracked lead goal has commercial value only if the enquiry is genuine, suitable and eventually becomes business.
Our ChatGPT Ads conversion-tracking walkthrough explains why a successful form response, a dispatched event and an attributed conversion are separate records. We will update this case study if our reconciled lead outcomes materially change the conclusion.
Should a New Zealand small business try ChatGPT Ads before Google Ads?
For a business with a considered purchase and a working lead form, ChatGPT Ads now deserves a measured test. Our own test generated substantially more affordable traffic and more tracked lead-goal events than Google Search over the same named dates. If those seven events reconcile to suitable enquiries, we would favour continuing that channel at its current controlled budget while improving the Google Search campaign’s cost and targeting.
For urgent, exact-intent searches, Google Search can still be valuable. The decision should follow qualified enquiries and sales, not an impression count or a platform’s own conversion total alone. Agree on one lead-quality definition, tag the destinations, and keep a simple record of enquiries by source before moving a meaningful share of the budget.
Is ChatGPT Ads cheaper than Google Ads in New Zealand?
In Kiwi Web Design’s 3 September–11 October 2026 test, yes on average cost per click: NZ$3.63 for ChatGPT Ads versus NZ$25.62 for Google Search. That is one advertiser’s result, not a published market price or a promise for another business.
Did ChatGPT Ads bring Kiwi Web Design seven clients?
No such conclusion is supported by this report. Ads Manager attributed seven lead_created events to ad clicks. Qualification, contact, proposals and sales need separate evidence from our enquiry records.
Are ChatGPT Ads the same as being recommended by ChatGPT?
No. Ads are sponsored placements, separate from ChatGPT’s answer. OpenAI says advertisers do not influence ChatGPT answers. Paying for an ad does not buy an organic recommendation or an AI citation.
Data source and next step
This article reports Kiwi Web Design’s own aggregate advertising account data, extracted read-only on 12 October 2026 from ChatGPT Ads Manager and the Google Ads API. The date range, currency, conversion definitions and campaign composition are stated above so the numbers can be cited with their limits. Individual customer enquiries and private account identifiers are not published.
Read our broader ChatGPT Ads versus Google Search decision guide for use cases beyond this one test. If you want to assess the channel for your business, see our ChatGPT Ads service and Google Ads service, or ask us about a measured NZ campaign.