What Changed in Digital Marketing This Week: Google Lead Forms, AI Overviews Images, and the ChatGPT Ads Reality Check

Key Takeaways

Google quietly removed the $50,000 lifetime spend requirement for Lead Form assets on 20 July, opening the format to any verified advertiser on Search and Performance Max. Zapier integration and email notifications for lead delivery were added at the same time.
Google AI Overviews can now generate images from text prompts, using Google’s Nano Banana model. The rollout is live in English across regions that already support AI Mode image creation, and it coincides with the 25th anniversary of Google Image Search.
Google’s SVP claimed AI search features send “billions of clicks to websites every week,” but independent studies conflict. Zero-click searches are running at 68%, and AI Overviews reduce click-through rates by 42% on queries where they appear. Google has not released granular data to support its claim.
Emarketer estimates ChatGPT Ads will miss OpenAI’s $100 billion 2030 revenue target by around 90%. The entire US chatbot ad market across ChatGPT, Copilot, Google AI Mode, and Alexa is projected at $5.41 billion by 2030.
Google brought business data feeds to Demand Gen campaigns on 17 July, letting travel, real estate, and automotive advertisers dynamically tailor creative from structured data without needing Google Merchant Center.
Google confirmed Local Inventory Ads will be enabled by default on all Shopping campaigns from 31 August. Advertisers who want to keep local and online inventory separate need to set up listing scopes before that date.

Five distinct updates landed between 14 and 20 July, and the pattern running through them is consistent with what Google has been building toward all year. On the advertising side, Google is lowering entry barriers and pushing AI deeper into campaign execution. On the search side, it is embedding AI further into the results page while publicly insisting that traffic to publishers is growing. Meanwhile, the AI competitor that was supposed to challenge Google’s advertising dominance is tracking well below its own revenue forecasts.

Here is what actually changed, and what it means if you run paid search or manage SEO for a New Zealand service business.

Google Finally Dropped the $50,000 Lead Form Barrier

This one deserves more attention than it got. On 20 July, Search Engine Land reported that Google removed the $50,000 lifetime spend threshold that had blocked smaller advertisers from accessing Lead Form assets. Lead Forms let users submit their contact details directly from a Search ad, without ever visiting your website. The $50,000 requirement kept the format in the hands of established accounts with long spend histories, and locked out smaller local businesses, new advertisers, and any account that had not yet pushed enough through the platform.

That requirement is now gone. The only eligibility criteria are advertiser reputation and verification status.

What changed alongside it matters. Lead Forms now support Search campaigns and Performance Max. Google added Zapier integration for lead delivery, joining the existing options of CSV download, webhooks, and the Ads API. Email notifications for new leads are also available. On data retention: manual CSV downloads are available for 30 days, but Google holds lead data for 60 days. Plan your export schedule around the second number, not the first. Geographic eligibility expanded to more than 25 new countries, including Croatia, Estonia, Jordan, Morocco, Qatar, Serbia, Slovenia, and Tunisia.

For a New Zealand service business, the question is whether the format suits your model. Lead Forms work well when your offer is clear and the conversion is a consultation request, a callback, or a quote. They work poorly when the prospect needs to absorb detailed information before deciding, because there is no landing page to do that work. If you have been running Google Ads for lead generation and did not try Lead Forms because your account was too new or your spend history too short, that barrier is now gone.

What to do: Set up a Lead Form on your best-performing Search campaign this week. Connect Zapier or configure webhooks so leads reach your CRM immediately, not when you remember to download the CSV. Test with a clearly defined offer. Lead Forms on Performance Max are also worth exploring, since the format now supports that campaign type natively. If you are already on PMax for lead gen, this is a straightforward addition.

Google AI Overviews Can Now Generate Images

On 14 July, Search Engine Land reported that Google has added image generation inside AI Overviews in Search. A user types a text prompt and Google’s Nano Banana model creates a custom visual from scratch, without leaving the results page. Google framed the announcement around the 25th anniversary of Google Image Search and a simultaneous redesign of the image search interface. The timing is deliberate.

The rollout is in English across regions that already support image creation in AI Mode. No specific date has been confirmed for any individual market, but the direction is clear.

The SEO implication is worth thinking through carefully. If someone searches “how does a heat pump work” and Google delivers an AI Overview that explains it, plus a generated diagram, that session produces no referral traffic. The same applies to any informational query where an image could satisfy the intent: explainer graphics, comparison charts, product illustrations, process diagrams, visual how-to content. If the image answers the question, there is no reason to click.

The traffic risk is concentrated at the top of the funnel, specifically in informational content that was already under pressure from AI Overviews expanding their footprint in search results. The format under less immediate pressure is content that requires depth, trust signals, and domain-specific knowledge to be credible. A “what is a heat pump” article faces more risk than a “heat pump installation costs Auckland 2026” page, because the second type needs local specificity, actual price data, and real experience to be useful.

We covered how AI Overviews are affecting click-through rates earlier this year in the AI Overviews CTR drop piece. Image generation inside the same interface compounds that pressure on shallow informational content.

What to do: Pull your Search Console data and identify which pages rank for query types that image generation could satisfy, specifically how-to, what-is, and visual explainer queries. If those pages are driving significant organic traffic, the risk is structural and not likely to reverse. The response is to go deeper: original research, local data, specific cost figures, or professional expertise that a generated image and a paragraph cannot replicate.

Google Says AI Search Sends Billions of Clicks. The Studies Say Otherwise.

On 17 July, Nick Fox, Google’s Senior Vice President of Search, published a statement that Google “is now sending billions of clicks to websites every week through AI features in Search alone.” The framing was a direct response to growing concern that AI Overviews are reducing organic traffic to publishers and that the shift to AI-generated answers is a net negative for the web.

Search Engine Land covered the claim and added the context Fox left out. Studies from the same period show zero-click searches running at 68% of all queries. A separate study found that AI Overviews reduce click-through rates by 42% on the queries where they appear. Google has declined to release granular traffic data to validate its claim, despite having launched AI performance reporting features in Search Console.

The problem with “billions of clicks” as a metric is that it does not answer the question anyone is actually asking. If AI Overviews are appearing on significantly more queries than last year, and reducing the click rate on each of those queries by 42%, the net effect on publisher traffic could still be negative even if the raw click count is growing, because total search volume is also rising. The claim does not resolve that question.

Google has a structural incentive to frame AI search as traffic-positive for publishers. If publishers pull content in response to traffic losses and AI training concerns, Google’s answer quality degrades. The incentive to reassure is real. That does not mean the claim is false, but it is reason to verify it against your own data rather than accepting it.

For NZ businesses: the useful action is to check your own Google Search Console organic click data over the past six months and filter for queries where AI Overviews appear. That is the actual data point for your site. If your clicks on AI Overview queries are flat or rising, the picture is better than the studies suggest. If they are falling, the macro claim is irrelevant to your specific situation. We have covered how to use the AI filter in Search Console for exactly this kind of analysis.

ChatGPT Ads Is Not Going to Be $100 Billion

On 14 July, Search Engine Land published Emarketer’s analysis of OpenAI’s advertising revenue projections. The gap between the two is striking.

OpenAI projected $2.5 billion in ad revenue for 2026 and $100 billion by 2030. Emarketer estimates the entire US standalone chatbot advertising market, covering ChatGPT, Microsoft Copilot, Google AI Mode, and Amazon Alexa for Shopping combined, will reach less than $1 billion in 2026 and $5.41 billion by 2030. If Emarketer’s read is correct, OpenAI’s 2030 target is off by roughly 90%.

The gap comes down to a single assumption. OpenAI’s projections assume ChatGPT Ads will capture a meaningful share of search ad budgets at scale, in a market that is already mature and dominated by Google. Emarketer’s view is that chatbot advertising is a distinct format, reaches a smaller and more fragmented audience than search, and will take considerably longer to develop the advertiser infrastructure and measurement frameworks that search has spent twenty years building.

None of this means ChatGPT Ads is irrelevant. The platform added audience list targeting, a new overview dashboard, suggested ad drafts pulled from existing website metadata, and new static ad formats inside a two-week window in early July, as we covered in last week’s roundup. The platform is real and building fast. The revenue projections just appear to be pricing in a future that has not arrived yet, and possibly will not arrive on OpenAI’s timeline.

For NZ businesses, the framing for ChatGPT Ads remains planning rather than action. The platform is not yet open to advertisers here. The Emarketer analysis is a useful calibration for how seriously to weight it against your current Google and Meta spend. A challenger platform tracking well below its own projections is still establishing itself, not displacing what already works.

Two Changes With Near-Term Deadlines

Google Demand Gen gets business data feeds. On 17 July, Google announced that advertisers in travel, real estate, and automotive can now connect business data feeds directly to Demand Gen campaigns without needing Google Merchant Center. The feeds let campaigns automatically tailor creative based on audience interests and available inventory, reducing the need to update ad assets manually when pricing or availability changes. The current limitation is that business data feeds are only supported on the Google Display Network within Demand Gen, not across the full campaign. If your business is in one of those three verticals and you are running Demand Gen, this is worth testing now.

Local Inventory Ads go default on 31 August. On 20 July, Google confirmed that all Shopping campaigns will have Local Inventory Ads automatically enabled from 31 August. Previously this required manually setting the enable_local field in the API. After that date, attempting to disable it via the Ads API on version 25.1 and later will return an error. For most businesses that sell both locally and online, the default setting is probably fine. But if you have a specific reason to keep local and online inventory displaying separately in Google Shopping, you need to set up listing scopes using CampaignCriterionService with product_channel set to ONLINE, or use campaign-level inventory filtering to exclude local products. Do it before 31 August.

What This Week Tells You About Where Things Are Heading

Strip the product names and the week’s announcements make a single argument. Google is consolidating more of what happens in both search and advertising, lowering friction to enter its ad platform while deepening AI’s role in what users see in search results.

On the advertising side, removing the Lead Form spend threshold, expanding Demand Gen to non-retail data feeds, and auto-enabling Local Inventory Ads for Shopping all point in one direction. Google wants more businesses spending through its platform, and reducing entry barriers is how it gets there. For smaller NZ advertisers who have been locked out of certain formats by spend requirements, this week’s changes are genuinely useful.

On the search side, the direction is consistent. AI Overviews generating images, the traffic claims from Nick Fox, the ongoing expansion of AI Mode in major markets. All of it points toward a version of search where more queries are resolved inside Google, with less need to click away. Whether that produces more or fewer referral clicks to external websites depends on the query type, and the honest answer is that Google’s claims and the independent research point in opposite directions.

For NZ service businesses, two things remain true at the same time. Google Ads is getting easier to access and harder to ignore. And organic search, particularly thin informational content, is under structural pressure that is not reversing. The businesses that navigate this cleanly will be the ones with strong conversion infrastructure on the paid side, and content that requires real expertise to be credible on the organic side.

The priority list from this week is short. Set up Lead Forms if you have not. Audit your Shopping campaign settings before 31 August. Check your Search Console click data on AI Overview queries. That is where the real signal is, not the platform-level claims.

Frequently Asked Questions

Do I need to spend $50,000 on Google Ads before I can use Lead Form assets? No. Google removed that threshold on 20 July 2026. Lead Form assets are now available to any verified advertiser, with no minimum spend requirement. The format supports Search and Performance Max campaigns, and Google has added Zapier integration and email notifications for lead delivery alongside the change.

What does it mean that Google AI Overviews can generate images? Google’s Nano Banana model can now create custom images from text prompts directly inside AI Overviews in Search, without the user clicking through to an external site. The practical effect is that informational content with a strong visual component, such as how-to guides, explainer diagrams, and comparison charts, faces greater risk of satisfying user intent without producing a referral click. The rollout is in English across regions already supporting AI Mode image creation.

Did Google confirm that AI Search is sending more traffic to websites? Google’s SVP stated that AI features send “billions of clicks to websites every week” but has not released the underlying data. Independent research from the same period shows zero-click searches at 68% and AI Overviews reducing click-through rates by 42% on queries where they appear. The most reliable way to assess this is to filter your own Google Search Console data by queries where AI Overviews appear on your site, not to take a platform-level claim at face value.

Should NZ businesses be concerned about ChatGPT Ads competing with Google? Not on the current trajectory. Emarketer projects the entire US chatbot ad market, covering ChatGPT, Copilot, Google AI Mode, and Alexa combined, at $5.41 billion by 2030. OpenAI’s own 2030 target is $100 billion. The platform is building fast, but it is not accessible to NZ advertisers yet, and the gap between its projections and independent forecasts suggests competitive displacement of Google’s advertising dominance is further away than OpenAI’s timeline implied.

What do I need to do about Local Inventory Ads before 31 August? If you run Google Shopping campaigns and want to keep your local and online inventory appearing separately, act before 31 August. After that date, Local Inventory Ads will be enabled by default on all Shopping campaigns, and disabling it via the API will return an error on version 25.1 and later. Set up listing scopes with product_channel set to ONLINE, or use campaign-level inventory filtering to exclude local products. If you have no specific reason to separate the two inventory streams, the default setting is fine to leave in place.

If any of the changes this week affect how your campaigns are set up and you would like a second opinion, we audit Google Ads and SEO accounts for NZ service businesses and can usually identify the highest-impact changes in a single session. Book a free strategy call and bring your account.