Introduction
If you have been watching ChatGPT Ads from the sidelines for the last several months — interested, but unwilling to burn a national-scale budget on national-scale impressions to find out whether the platform actually works for a Northeast Indiana small business — May 22, 2026 was the day the math changed. OpenAI announced a meaningful expansion of its Ads Manager Beta. The two additions that matter for a Fort Wayne or Auburn small business are simple: geo-targeting and daily-budget controls. With those two levers, you can finally test ChatGPT Ads with a budget that makes sense for a local service business — without paying to serve impressions to people 800 miles away who will never call you.
This post is the setup playbook. We will walk through what the new controls actually let you do, how to plan a first $500 test that gives you a real signal in a reasonable timeframe, three vertical scenarios for Northeast Indiana service businesses, and an honest discussion of the limitations you should expect. The headline up front: ChatGPT Ads is now usable for a Fort Wayne SMB, but it is not a replacement for Google Ads or Meta. Treat it as a complementary high-intent surface to test with a modest monthly budget while the platform matures.
Key Takeaways
- OpenAI's May 22, 2026 update added state-level, designated market area (DMA), and zip-code geo-targeting to ChatGPT Ads Manager, along with daily budget controls for new campaigns.
- The DMA and zip-code granularity is what makes ChatGPT Ads usable for a Fort Wayne small business for the first time.
- Early CTR data Search Engine Land reported looked strong on engagement compared to Display and Podcast benchmarks, but inventory remains limited and conversion data is unproven.
- A reasonable first test is $500 over 30 days targeted to Allen County and DeKalb County, with daily-budget pacing and conversion tracking installed before launch.
- The platform is best for high-intent service queries (HVAC emergency, dental booking, specialty retail) rather than awareness-stage advertising.
- Honest limitation: the Beta is still small, advertiser controls are limited, and ChatGPT-paid is not yet a replacement for Google or Meta paid search.
What Actually Changed on May 22, 2026?
The Search Engine Land coverage of the Ads Manager expansion lays out the specifics. The Beta now offers:
- Geo-targeting at three granularity levels across the U.S.: state, designated market area (DMA), and zip code. Settings can be configured during campaign setup or adjusted in campaign settings afterward.
- Daily and lifetime budget options, with the limitation that daily budgets are currently restricted to newly created campaigns.
- Aggregate reporting with impressions, clicks, and spend now displayed across campaign, ad group, and ad-level views.
- Dynamic call-to-action testing, where OpenAI automatically selects from CTAs like “Shop Now,” “Book Now,” “Sign Up,” and “Learn More” — advertiser control over CTA selection is described in the announcement as something OpenAI “could explore in the future.”
The Search Engine Land piece notes that the geo-targeting expansion brings ChatGPT's ad tools closer to the location controls available on Google and Meta. That is the practical translation: for the first time, you can run a ChatGPT Ads campaign that does not serve impressions across the entire country.
For a Fort Wayne small business, this matters in two distinct ways. The Fort Wayne–Auburn Nielsen designated market area covers most of Northeast Indiana including Allen, DeKalb, Whitley, Wells, Adams, and Huntington counties — about the right geographic radius for a regional service business. Zip-code targeting goes finer than that, useful for a contractor who only services within 20 miles of their shop or a specialty retailer focused on a specific neighborhood. State-level targeting is too broad for most local businesses but useful for a multi-location operator with locations across Indiana.
The daily-budget control is the second half of the unlock. The Search Engine Land self-serve buying expansion coverage earlier in May described the platform's pre-update model as primarily lifetime budgets, which made small-business pacing difficult — you would commit to a total spend and the platform would optimize as it saw fit. Daily budgets let you pace spend predictably, kill campaigns that are not working at the end of week one, and avoid the spike-and-burnout patterns that small-budget lifetime campaigns are prone to.
What Do the Early Numbers Actually Tell Us?
Before you plan a test, it helps to know what the platform looks like across other advertisers. The early signal is encouraging but limited.
Search Engine Land's analysis of early ChatGPT Ads CTRs found that early data showed ads “outperforming traditional benchmarks” on engagement metrics compared to Display and Podcast channels — without specific CTR percentages disclosed publicly. The piece also notes that Mother's Day–related prompts triggered ads about three times more often than average per SimilarWeb data, with brands like Etsy, Nordstrom, and flower retailers showing strong visibility. The format works well because ads appear directly inside conversational responses, making placements feel more contextual than typical display.
The piece is also honest about three significant limitations: inventory is limited because the platform is still in restricted testing, the novelty factor is real (users may engage simply because the format is new), and high CTRs do not guarantee conversion quality or cost efficiency. Or, as the piece summarizes it, ChatGPT ads “are off to a strong start on engagement — but until scale, cost and conversion data catch up, advertisers should treat this as a promising test channel, not a proven one.”
Earlier reporting in Search Engine Land's coverage of advertisers testing the Ads Manager noted that Best Buy and Expedia were spotted in early ad tests, though it is unclear whether they specifically used the new Ads Manager interface itself. Before the dashboard launched, advertisers were relying on “basic reporting like weekly CSV files.” The platform has clearly matured fast, but it is still pre-1.0 by most measures.
A deeper look at the ad inventory and roadmap is in Search Engine Land's “Inside ChatGPT Ads” piece, which traces what is shipping next and where the gaps are. The combination of those signals is what informs the test budget we recommend below. The platform is real, the engagement is real, but the scale and conversion data are not yet at the point where you should redirect significant Google Ads or Meta budget into it. The honest framing is that a $500 first test buys you a real answer about whether the platform works for your business in your geography — not a commitment to scale.
Planning a First $500 ChatGPT Ads Test for a Fort Wayne SMB
The first test should be designed to produce a clear yes or no answer in 30 days. Anything more elaborate is a budget for a different post. Here is the structure we use with clients.
Budget and pacing. $500 over 30 days at a daily budget of $16-17. The daily-budget control is what makes this work — it lets you pace evenly, catch underperformance by day 10, and kill the campaign with budget remaining if the signal is bad. Do not use lifetime budget for a first test; the early data is too volatile to commit a lump sum to.
Geographic targeting. For most Fort Wayne–area service businesses, start with the Fort Wayne–Auburn DMA. That covers Allen, DeKalb, and surrounding counties without forcing you to manually assemble a zip-code list. If you are a hyperlocal business — a specialty retailer in a specific Fort Wayne neighborhood, an Auburn contractor with a 20-mile service radius — start with a hand-picked list of 8-15 relevant zip codes instead. State-level Indiana targeting is too broad for most local SMBs.
Campaign structure. Two ad groups. One for high-intent service queries (e.g., “emergency HVAC repair,” “same-day dentist appointment,” “best [specialty retailer category] near me”) and one for branded plus competitor queries. The intent-stage signal is much stronger than awareness-stage on ChatGPT in early data; lead with intent.
Creative. Two to three ad variants per ad group. Keep copy concrete and benefit-led — what the customer gets and when. Avoid superlatives. The dynamic CTA testing the platform now runs (“Shop Now,” “Book Now,” “Sign Up,” “Learn More”) means you do not need to A/B test CTAs yourself; OpenAI is doing that automatically per the platform's Ads Manager documentation.
Tracking. Install the ChatGPT Ads conversion pixel before launch. Send a primary conversion event (phone call, form submission, completed booking) and a secondary engagement event (button click, page-2 scroll). If you do not have a clean conversion-tracking setup on your site at all, fix that first — running paid ads without conversion tracking is a Fort Wayne–specific budget mistake we covered in Fort Wayne Google Ads wasted spend, and the same logic applies here.
Measurement window. 30 days minimum. The first week will look noisy; week two starts to give you signal; week four lets you compare cost per conversion to your existing Google Ads or Meta baseline. The right yardstick is not CTR — CTR will probably look great because of the novelty factor — it is cost per qualified lead measured the same way you measure your other paid channels.
A note on attribution: ChatGPT Ads conversion tracking is still maturing. Expect some attribution gaps, particularly for users who see the ad in ChatGPT but convert later via direct search. Do not expect to credit every conversion cleanly. The broader attribution problem is something we addressed in our marketing attribution for small business guide, and the same multi-touch reality applies here. Do not kill a campaign because attribution looks soft; cross-check against overall lead volume during the test window.
Three Vertical Scenarios for Northeast Indiana
The framework is the same; the specifics vary by business type. Three scenarios from real campaigns or planned engagements:
Fort Wayne or Auburn HVAC contractor (winter emergency intent)
The strongest use case we have seen for ChatGPT Ads in Northeast Indiana is winter heating-failure intent. When the temperature drops in January and a furnace stops, the affected homeowner often goes to ChatGPT or Perplexity to ask “what to do when my furnace won't turn on” or “how do I find an emergency HVAC repair in Fort Wayne.” These queries are high-intent, time-sensitive, and dollar-rich.
The setup: target the Fort Wayne–Auburn DMA. Daily budget of $25-30 during cold snaps, reduced to $8-10 in mild weeks. Two ad groups — “emergency furnace repair” and “no-heat 24/7 service” — with copy emphasizing same-day dispatch and an after-hours phone line. Conversion event: phone call lasting more than 60 seconds. Expected outcome from a 30-day winter test: enough call volume to compute a cost-per-call and decide whether to expand. The companion to this work is our ChatGPT advertising for small business pillar, which covers the platform-wide strategy.
DeKalb County home services (pre-storm-season roof and gutter)
A second pattern: pre-storm-season intent for exterior home services. April through June and September through October bring spikes in queries about roof inspections, gutter cleaning, and storm damage assessment. These are research-stage queries, and the consultation-booking conversion has a longer lag than the emergency HVAC scenario.
The setup: target DeKalb County by zip code (Auburn, Garrett, Butler, Waterloo, Ashley). Daily budget of $15-20 for an 8-week window around the season. Two ad groups — “roof inspection” and “gutter cleaning consultation” — with copy emphasizing free in-person assessment. Conversion event: completed booking form. Expect a longer time-to-conversion than the HVAC scenario; budget patience accordingly.
Auburn specialty retailer (holiday gift intent)
The third pattern is harder to predict but worth testing for non-service retailers. November and December bring meaningful spikes in gift-related conversational queries — “best gifts for [specific category]” or “unique gifts under $50 in Northeast Indiana.” Specialty retailers with distinct product categories can show up well in these contexts if their inventory matches the query intent. The ChatGPT product-feed ads for small e-commerce piece covers the feed-driven side of this work in more depth.
The setup: target Auburn plus surrounding zip codes during the November-December window. Daily budget of $20-30. One ad group per top-3 gift product category. Conversion event: completed checkout. The honest caveat: e-commerce-style intent on ChatGPT in 2026 is still small relative to Google Shopping or Meta, so treat this as a complementary surface rather than a primary channel.
What ChatGPT Ads Is Not Yet Good For
A few categories where we would hold off, even with the geo-targeting unlock:
Pure awareness-stage advertising. The early data suggests ChatGPT Ads is most useful when the user is actively asking a question with intent. Broad-reach awareness budgets are better spent on Meta or YouTube right now.
Long-tail or zero-intent queries. The inventory is too thin and the targeting controls do not yet support detailed exclusions. If your business needs very specific keyword-level negative lists, the current Beta will frustrate you.
Replacement of existing high-performing paid search. If your Google Ads or Meta paid campaigns are producing predictable cost-per-lead at scale, do not redirect that budget. Test ChatGPT Ads with incremental dollars. Our AI visibility tools for small business piece covers how to monitor whether your brand is even being mentioned in AI search before you commit further.
Aggressive geo-targeting outside the U.S. The geo-targeting expansion described in Search Engine Land's coverage is U.S.-only. For Northeast Indiana businesses serving cross-border customers (rare but real for some niche manufacturers), the targeting controls do not yet cover that.
Fort Wayne Setup Checklist
For a Northeast Indiana SMB owner who wants to ship a first ChatGPT Ads test in the next two weeks, here is the sequence we use:
| Step | Action | Owner | Time |
|---|---|---|---|
| 1 | Get approved into ChatGPT Ads Manager Beta | Owner / agency | 0-7 days wait |
| 2 | Install conversion tracking pixel on the site | Developer | 1-2 hours |
| 3 | Verify the pixel fires on the right conversion events | Developer / marketer | 30 minutes |
| 4 | Define audience: DMA-level (regional) or zip-code list (hyperlocal) | Marketer | 30 minutes |
| 5 | Set daily budget for new campaign ($16-17 for $500/30-day test) | Marketer | 5 minutes |
| 6 | Build two ad groups: intent-stage and branded/competitor | Marketer | 1-2 hours |
| 7 | Draft 2-3 ad variants per ad group with concrete benefit-led copy | Marketer | 1-2 hours |
| 8 | Launch and pace daily for 30 days | Marketer | 15 min/day |
| 9 | Review at day 10 (kill switch) and day 30 (decision point) | Owner + marketer | 1 hour |
The total time investment for the first test is roughly 6-10 hours plus a small daily check-in. The hardest step in our experience is step 1 — getting approved into the Beta — because OpenAI is still gating access and the queue moves at its own pace. Submit the access request as soon as you decide to test; do the other steps in parallel.
A non-obvious point: if you already run Google Ads and Meta well, the muscle memory transfers. The Beta's interface, as described in Search Engine Land's overview of the early Ads Manager, is converging toward the conventions advertisers know from Google and Meta. The platform-specific quirks are smaller than the underlying patterns are familiar.
Want Help Getting Into the Beta and Setting Up a Test?
If you are not sure whether ChatGPT Ads is right for your Fort Wayne, Auburn, or Allen County business — or if you want help getting access, installing tracking, and structuring a first test — our Google Ads management practice now covers ChatGPT Ads alongside Google and Meta. We start with a 30-minute call to understand your business, your existing paid mix, and where ChatGPT Ads might fit. If the answer is “not yet,” we will tell you. We do not pitch tests for the sake of running tests.
For broader AI-readiness questions — not just paid, but how your business shows up across the whole AI search surface — our AI solutions advisory covers paid, organic, schema, and content together. ChatGPT Ads is one lever in a larger AI visibility strategy, and the right answer for most small businesses is to think about all the levers at once rather than chasing the newest one.
Ready to Ship Your First ChatGPT Ads Test?
A 30-minute call to see whether ChatGPT Ads makes sense for your Fort Wayne, Auburn, or Allen County business — and to scope a $500 starter test if it does.
Frequently Asked Questions
- How do I get into the ChatGPT Ads Manager Beta?
- Access is currently gated through OpenAI's ChatGPT Ads documentation and request flow. Approval times vary — we have seen waits from one day to several weeks. Submitting through an agency partner can occasionally accelerate access, but the underlying queue is OpenAI's. Plan the rest of your setup in parallel rather than waiting for approval before doing the work.
- What is the minimum budget that actually produces signal?
- In our experience, $500 over 30 days is the floor for a Northeast Indiana DMA-level test. Below that, the volume is too low to compute reliable cost-per-conversion. Above $1,500 in a first test is premature — the inventory and platform stability do not yet justify a larger initial commitment for a small business. The sweet spot for a first test is $500-1,200.
- Should I run ChatGPT Ads instead of Google Ads?
- No, not as a replacement. ChatGPT Ads is a complementary channel, not a substitute. The right framing is to treat your existing high-performing Google Ads or Meta budget as untouched and use new dollars to test ChatGPT. If the test produces a competitive cost-per-conversion, then scale gradually.
- Does ChatGPT Ads work for B2B service businesses?
- Early signs are encouraging for B2B intent queries, especially research-stage queries where a buyer is asking ChatGPT to help shortlist vendors. The geo-targeting expansion makes it newly viable for regionally focused B2B providers — a Northeast Indiana commercial contractor, for example, or a Fort Wayne-area accounting firm. The honest caveat is that B2B conversions take longer to attribute than consumer service calls, so the measurement window needs to be longer.
- Can I exclude specific queries or topics from triggering my ads?
- Negative-keyword controls and topic exclusions in the Beta are limited compared to Google Ads. OpenAI has not publicly committed to a feature-parity timeline. For now, plan around the fact that some impressions will be served against queries you would not have chosen — and budget around the noise rather than expecting to filter it out.
- What conversion event should I track?
- For service businesses: phone calls lasting more than 60 seconds, completed booking forms, or quote-request submissions. For e-commerce: completed checkouts. Avoid using upper-funnel events (button clicks, page views) as your primary conversion; they will inflate CTR while obscuring whether the channel actually drives revenue. We recommend tracking both a primary (revenue-linked) and secondary (engagement) event from the start.
- How is reporting in the Beta compared to Google Ads?
- The Search Engine Land update notes that aggregate totals — impressions, clicks, and spend — now display across campaign, ad group, and ad-level views. That is a meaningful improvement from the earlier "weekly CSV file" era. It is still less granular than Google Ads. Expect to do some of your own attribution stitching, especially in the first 60 days as both you and the platform are calibrating.
Sources & Further Reading
- Search Engine Land: OpenAI expands Ads Manager Beta with new budgeting and geo targeting controls — May 22, 2026 announcement coverage.
- Search Engine Land: Advertisers test ChatGPT Ads Manager — April 22, 2026 early-tester reporting.
- Search Engine Land: ChatGPT ads expand with self-serve buying — May 5, 2026 self-serve buying expansion.
- Search Engine Land: ChatGPT ads show strong early CTRs but scale is still the question — May 5, 2026 engagement analysis.
- Search Engine Land: Inside ChatGPT Ads: what the data tells us and what's coming next — May 1, 2026 roadmap and inventory analysis.
- OpenAI: ChatGPT Ads (OpenAI Help Center) — canonical documentation for the Ads Manager Beta.