Lead Gen News: September 2026 - AI Max Auto-Migration Live, LSA Missed Call Billing, and Meta's Campaign Score
Five changes across Google and Meta this month affect how NZ service businesses should be running paid lead generation right now. Google’s AI Max auto-migration is live across accounts during September. Local Services Ads will bill for missed calls from October 1, which is tomorrow. Customer Match gained IP address matching on September 25. Google is testing a significantly larger LSA format. And Meta has launched a 0 to 100 campaign score that evaluates how well your setup aligns with their automated best practices.
Here is what changed, why it matters, and what to do before the month turns over.
Google’s AI Max Auto-Migration Is Running Right Now
If you have not opted into AI Max voluntarily, Google is not waiting any longer. Between September 1 and 30, 2026, Google is automatically migrating two groups of Search campaigns to AI Max: campaigns using Automatically Created Assets (ACA) and campaigns using campaign-level broad match settings.
This is not a notification asking for permission. It is an in-progress migration, running across accounts throughout the month.
Automatically Created Assets is the feature that generates headlines and descriptions on the fly based on your landing pages and existing ad copy. If ACA was turned on in your account, your campaigns have now crossed over into AI Max’s text customisation engine. If you had ACA switched off, the broad match migration still applies if you were using campaign-level broad match targeting.
The reported performance lift for the full AI Max feature suite is 7% more conversions or conversion value at a similar CPA or ROAS on average. That number will vary significantly based on how clean your conversion tracking is and whether your negative keyword lists are in reasonable shape before the migration runs.
What is not migrating until February 2027
This is an important correction to the July update, which reported that Dynamic Search Ads would be retired in September. Google pushed that deadline. The automatic migration of remaining DSA campaigns to AI Max now happens February 1 to 28, 2027.
What has changed in September is narrower: new DSA campaign creation through Google Ads, Google Ads Editor, and the API is no longer available. In-account notifications encouraging voluntary upgrades have been rolling out since July, and those notices are now more pointed. DSA campaigns already running will continue to serve through January 2027.
If you have DSA campaigns, the window to migrate on your own terms is open until January. That is a better outcome than a February forced migration with no preparation time, but it is not an excuse to let it sit until Q1.
What to check before month-end
Log into any account that was running ACA or campaign-level broad match and pull the search terms report. Look at what the AI is targeting under the expanded settings, specifically which categories are now receiving impressions that your keyword list never covered before.
The gap between what AI Max targets and what your exact match terms covered is where lead quality risk concentrates. A construction company whose campaigns previously captured only high-intent queries like “commercial builder Auckland” may now be entering auctions for earlier-stage research queries. Some of that traffic will convert. Some will produce time-wasting enquiries.
Add negatives for any search categories that would never produce a qualified lead for your specific business. The algorithm learns from every conversion signal it receives, so four weeks of poor-quality form submissions will point bidding in the wrong direction for months.
Our full guide to enabling AI Max safely on Search campaigns covers the negative keyword prep work and quality controls in detail.
Local Services Ads: Missed Calls Become Billable From October 1
This is the most time-sensitive item in this roundup.
Starting October 1, 2026, Google is changing how it charges for calls through Local Services Ads. Under the current policy, you are only charged for answered calls that meet minimum duration criteria. From tomorrow, a missed call during business hours can be charged if the caller stayed on the line for more than 20 seconds.
The 20-second clock starts when the call connects to your phone system. If your system requires callers to press a number to reach a department, the timer does not begin until they press that key. Callers who abandon before selecting an option will not be charged. But a caller who reaches your voicemail or sits on hold and hangs up after 20 seconds will now appear as a billable lead event.
A second change runs alongside this. Under the current policy, repeat calls from the same customer within a 15-day window count as a single lead. From October 1, if an initial call does not qualify as a chargeable lead, a subsequent call between the same customer and your business can become chargeable if the follow-up call meets valid-lead criteria on its own.
What this means for your call handling
The practical risk is straightforward: a missed call is now money spent with no corresponding opportunity to convert it.
Check your business hours settings in your LSA profile first. If calls route to an unmanned line outside hours, make sure your stated hours are accurate so Google does not count after-hours calls against your budget. Calls outside your nominated business hours are not charged under the new rules.
Review your call handling next. If your team regularly lets calls go to voicemail during business hours, that pattern now carries a direct cost. If a missed call costs the same as an answered one in LSA billing terms, the cost of not answering is no longer zero.
For businesses running LSAs as a primary lead source, an overflow answering service may now have a clear economic case that it did not have last month. The arithmetic is simple: if your average LSA cost per lead is $60 to $80, and a missed call costs the same, a $500 per month call answering service pays for itself in eight prevented missed charges.
The bigger LSA format on its way
Alongside the billing change, Search Engine Land reported on September 25 that Google is testing a significantly expanded LSA format in search results. The redesign gives each listed business more visual space, with a larger image and a highlighted information box beneath the main listing.
The test is running in the US market with no confirmed New Zealand rollout date.
If it arrives here, the effects compound what the billing change already implies. A larger format means more above-the-fold real estate controlled by LSAs, which pushes organic results and standard paid ads further down the page. Review volume and star rating become more visible at this size, making the difference between a business with 15 reviews and one with 90 immediately apparent to anyone scanning the results.
Getting your review count up and your call handling right are now both more urgent as the LSA format evolves.
For a full breakdown of when LSAs make sense versus standard Google Ads for NZ service businesses, the comparison between the two formats covers the qualification requirements and cost-per-lead economics in detail.
Customer Match Now Accepts IP Addresses and Timestamps
On September 25, Google announced that advertisers can now include a user’s IP address, and optionally the timestamp of their interaction, when uploading Customer Match data to Google Ads.
This is a new matching signal alongside the existing Customer Match identifiers: email address, phone number, full name, and postal address. Unlike those fields, which must be hashed before upload, IP addresses and timestamps are uploaded raw.
The mechanics are specific. An IP address uploaded with a timestamp is matched to the known user of that IP at that point in time. An IP uploaded without a timestamp is matched to the most recent known user of that address. A timestamp cannot be uploaded without an accompanying IP address.
Google expects match-rate improvements from IP signals to begin flowing through in October 2026, meaning the practical effect on audience size will start to appear from next month.
How to think about this practically
Customer Match has always done two useful things: re-engage existing customers who have lapsed, and suppress current clients from prospecting campaigns so you are not paying to advertise to people who already have a booking. The limitation has always been match rate. Email addresses go stale, phone numbers change, and postal records drift.
IP-based matching adds a signal that does not depend on a contact form submission or a phone call completing. If someone visited your website, did not convert, and your analytics platform captured their IP, that data can now be included in a Customer Match upload to widen your addressable audience beyond the contacts in your CRM.
For an Auckland mortgage broker who sees hundreds of sessions a month from people who browse rates pages and leave without enquiring, or a Wellington law firm with meaningful site traffic but a modest contact list, this is a mechanism to close the gap between traffic and retargetable audience, without requiring additional paid media spend to recapture those users through standard remarketing.
Match rates will vary depending on IP data quality and the ratio of returning versus new visitors in your analytics export. Treat the first upload as a baseline and measure audience size improvement in October against what your existing Customer Match lists were delivering.
The privacy point
IP and timestamp matching is not available for users in the EEA, UK, or Switzerland. For New Zealand, there is no current restriction. However, using IP addresses as customer data means you are handling personally identifiable information under the Privacy Act 2020.
Before uploading, confirm that your website privacy policy covers the collection and use of IP data for advertising purposes. If your policy only references cookies and analytics, it likely needs updating. Running an upload before that step is a compliance risk, not a technicality.
Meta Advantage+ Leads Campaigns Now Available Globally
Meta has been rolling out Advantage+ Leads Campaigns progressively through 2026, and they are now available globally, including for New Zealand advertisers.
The format is Meta’s fully AI-automated version of a leads campaign. You supply the creative assets, your website URL, your lead form or landing page, and your budget. Meta’s system handles audience targeting, placement selection, creative sequencing, and budget distribution. The controls available to the advertiser are creative inputs and the conversion goal, not the audience build or placement mix.
Meta’s early data shows an average 10% lower cost per qualified lead compared to manually configured leads campaigns. That figure applies to accounts with sufficient conversion history. Accounts running fewer than 50 lead form completions per month will see a longer learning phase before the system stabilises, and early results can look worse than the manual baseline during that window.
For NZ service businesses already running Meta leads campaigns, the practical question is whether to move to the Advantage+ format or stay with a manual structure. Below 50 leads per month, the longer learning phase means efficiency gains come slowly and the first few weeks can be disruptive. Above that threshold, testing is a reasonable call.
The new campaign score
Meta has also launched a campaign score, visible inside Ads Manager as a 0 to 100 indicator. The score evaluates your campaign across four dimensions: creative variety, signal quality through Pixel and Conversions API configuration, audience breadth, and how much of the Advantage+ automation stack you have enabled.
Research from third parties suggests campaigns scoring below 60 consistently underperform those scoring above 80 on Meta’s own benchmarks. That correlation is worth taking seriously as a diagnostic, but it is not the same as saying a score of 80 guarantees better leads for your specific business.
The score is a setup metric, not a performance metric. A well-performing campaign with a manual audience and limited automation features will score low. A poorly structured campaign that has every Advantage+ feature switched on will score high. Meta has a commercial interest in encouraging advertisers toward fuller automation, and the score reflects that interest.
Use it as a checklist. If your score flags that Pixel events are firing inconsistently, or that your campaign has only one creative variant, those are genuine problems worth fixing. If it flags that Advantage+ Audience is off and your manual audience is producing better-qualified leads than broad expansion would, leave it off. Apply the recommendations that make commercial sense, not the ones that move the number.
Predictive Budget Allocation
Meta is also rolling out Predictive Budget Allocation, which automatically shifts spend between ad sets in real time based on performance signals. Early tests report 8 to 15% improvement in ROAS compared to fixed allocation.
If you have been manually controlling budget splits between ad sets, check whether Predictive Budget Allocation is already active in your account. It changes how to read ad set level performance: a sudden budget reduction in one ad set is not a signal that the ad set is failing. It may be Meta responding to a stronger opportunity elsewhere in the campaign at that moment.
The practical shift is to evaluate performance at the campaign level rather than the ad set level when this feature is running. Ad set budgets become fluid; campaign totals and cost per qualified lead are the metrics that matter.
For a breakdown of how Advantage+ and Meta’s Andromeda system interact under the hood, the Advantage+ guide for NZ service businesses covers the mechanics in full.
What These Changes Have in Common
Four of the five updates this month push more decision-making further inside Google’s and Meta’s automated systems. The exception is the LSA billing change, which is a straightforward cost increase for businesses whose call handling was producing non-chargeable missed calls. That is an operations problem with a clear fix: answer the phone.
For the algorithmic changes, the pattern is consistent with everything that has happened across both platforms in 2026. Automation is expanding, and the quality of your inputs determines whether that expansion helps you or hurts you.
AI Max expanding reach on an account with thin negatives will produce mixed results. Advantage+ Leads optimising toward a conversion goal that does not reflect lead quality will maximise the wrong metric. IP-based Customer Match on a stale IP export will not move match rates in a meaningful direction.
The businesses that will benefit from these September changes are those already doing the work that these systems depend on: consistent conversion tracking, regular negative keyword reviews, current customer lists, and enough creative variation to give Meta’s algorithm something to test. None of that is technically complex. It is a maintenance discipline that platforms are now enforcing through performance rather than through explicit requirements.
Heading into Q4, the gap between accounts that are feeding these systems well and accounts that are not is widening. The platforms are not offering more control. They are offering better results for those who have put the right foundations in place.
If you want a clear-eyed view of where your Google or Meta lead campaigns stand heading into Q4, book a free strategy call with Lucid Leads.
Frequently Asked Questions
Do I need to do anything about the AI Max migration before October?
Yes. Pull your search terms report now and review what the expanded targeting is matching against. Add negatives for any query categories that would never convert to a qualified lead for your business. The algorithm is learning in real time, and poor signals from September will take time to reverse.
Does the LSA missed call billing change affect New Zealand businesses?
Google’s announcement did not specify a geographic limit. If you are running Local Services Ads, review your call handling before October 1. A missed call during your nominated business hours that exceeds 20 seconds on the line will be a billable lead event from tomorrow.
Is Customer Match IP address matching available in New Zealand?
Yes. The restriction applies to the EEA, UK, and Switzerland. New Zealand has no current restriction. Before using IP addresses in a Customer Match upload, verify that your privacy policy covers this use case under the Privacy Act 2020.
Should I follow every recommendation my Meta campaign score flags?
No. Treat it as a diagnostic checklist, not a target. Apply recommendations that address a genuine gap, such as inconsistent Pixel tracking or a single creative variant. Ignore recommendations that push you toward settings your campaign does not need. A high score on a poorly structured campaign means nothing. A low score on a campaign producing strong, qualified leads is not a problem worth fixing.
What is the actual DSA retirement deadline now?
The automatic migration of Dynamic Search Ads campaigns to AI Max runs February 1 to 28, 2027. New DSA campaign creation is no longer available as of September. Existing DSA campaigns continue to serve through January 2027. If you want to migrate on your own terms and monitor the transition, start in November or December rather than waiting for the February forced migration.
Related Articles
Continue learning about lead generation and paid advertising

Lead Gen News: Late October 2026 - Google's LSA Merger, Meta's Product Browsing Deadline, and New Channel Priority Controls
Google folds Local Services Ads into Performance Max. Meta enables product browsing in existing ads on 28 October. New channel controls cap Display budget spend.

Lead Gen News: Mid-October 2026 - Google Automated Promotions Live, Meta's AI Campaign Manager, and Free Incrementality Testing
Google automated promotions went live 12 October. Meta launched an AI campaign manager at Ad Week. Self-serve conversion lift now open for Search and PMax.
Written by
Founder & Lead Generation Specialist
Jason Poonia is the founder of Lucid Leads, helping service businesses across New Zealand generate qualified leads through paid advertising and conversion-focused funnels. With a background in Computer Science from the University of Auckland and over 5 years of experience running lead generation campaigns, Jason has helped businesses in construction, trades, real estate, and professional services generate thousands of qualified leads. His data-driven approach combines targeted ad strategies with rapid lead qualification to deliver prospects who are ready to buy.