# ChatGPT Ads Cost Per Lead for Local Services: How to Calculate It

URL: https://www.gptadsai.com/guides/chatgpt-ads-cost-per-lead-local
Author: Tarun Kapoor — Founder, GPT Ads AI (12+ years paid media)
Published: 2026-10-25
Updated: 2026-11-16
Type: Guide

> No CPL benchmark exists for ChatGPT Ads in local services. See the evidence, how Target CPA works, and how to work back from job margin to cost per booked job.

## Contents

- [You do not pay per lead](#you-do-not-pay-per-lead)
- [The evidence that exists](#the-evidence-that-exists)
- [Five ways to count the same month](#five-ways-to-count-the-same-month)
- [Work backwards from the job](#work-backwards-from-the-job)
- [Target CPA and Target CPC in practice](#target-cpa-and-target-cpc-in-practice)
- [Budget sizing by business type](#budget-sizing-by-business-type)
- [How to know when the number is real](#how-to-know-when-the-number-is-real)
- [Why CPL misleads, and what to do about it](#why-cpl-misleads-and-what-to-do-about-it)
- [A six-week test, in whole weeks](#a-six-week-test-in-whole-weeks)
- [Reporting the number to an owner or finance](#reporting-the-number-to-an-owner-or-finance)
- [What to avoid claiming](#what-to-avoid-claiming)
- [Checking whether assistants name your local business](#checking-whether-assistants-name-your-local-business)
- [Sources and further reading](#sources-and-further-reading)

**Short answer:** there is no published cost-per-lead benchmark for ChatGPT Ads in local services, and you never pay per lead on this channel: OpenAI bills by the click or the impression, and your cost per lead is a number you calculate by dividing media spend by the leads you can verify. The published evidence is a handful of named case results and one Google Local Services Ads benchmark that does not describe ChatGPT. The workable method is to set your allowable cost per booked job from your own margins, convert it into a cost per lead and a click price, then run a controlled test long enough to read it. Since 9 November 2026, new conversion campaigns use Target CPA, which is an average-cost goal and not a guaranteed price.

> **How to read the numbers on this page.** **OpenAI** marks a statement from OpenAI's help centre or policies. **Third-party** marks a named outside figure with its date; it is not OpenAI data and we have not re-run it. **GPT Ads AI observation** marks the one first-party range this site has published, quoted word for word. **Illustrative** marks an assumption in a worked example about a fictional business. **GPT Ads AI recommendation** marks our advice. No figure here is a forecast for your account.

> **From the team behind [GPTAdsAI.com](https://www.gptadsai.com).** GPT Ads AI is a done-for-you ChatGPT Ads agency: we set up, run and measure campaigns for local lead-generation budgets. Two companion resources cover the rest of the stack: [contexthints.com](https://contexthints.com/) for writing context hints, and [measure.co](https://measure.co) for local AEO, GEO and SEO visibility in ChatGPT and other assistants.

## You do not pay per lead

Search for "cost per lead" and most results describe pay-per-lead products. ChatGPT Ads is not one. The cost of a lead is an output of how well your clicks turn into leads, so it moves whenever your landing page, your tracking or your sales process changes.

| Objective | You are billed on | Bidding control for new campaigns (from 9 Nov 2026) | How you get a cost per lead |
|---|---|---|---|
| Views | Impressions (CPM) | Not part of the 9 November change as we read it; check your account | Spend divided by leads you track from those impressions; usually thin |
| Clicks | Valid clicks (CPC) | Target CPC, which replaces Max CPC | Spend divided by leads from the landing page |
| Conversions | Valid clicks (oCPC) or impressions (oCPM) | Target CPA, which replaces Max Bid, or intelligent bidding | Spend divided by conversions you verify; **OpenAI**: "you do not pay per conversion" |

Max Bid and Max CPC remain on existing campaigns until **OpenAI** retires them on 1 January 2027. Some older guides on this site describe the Bid Cap and Max Bid controls; the 9 November change supersedes them for new campaigns. Check the labels in your own Ads Manager before you set anything.

## The evidence that exists

The table lists every figure we are willing to cite, with what it does and does not tell you.

| Source | Date | Figure | What it can tell you | What it cannot |
|---|---|---|---|---|
| Third-party: Angi × OpenAI customer story | 22 Sep 2026 | +38% quarter-on-quarter ROAS and -34% cost per lead; pilot across about 20 home-service categories, later expanded to about 30 more | Home-service categories are live | Any absolute CPL, or a result for one trade |
| Third-party: SearchLight Digital (agency-reported) | 8 Oct 2026 | 37 contractors; paid ChatGPT leads rose from 39 to 93 per month, June to September 2026 | Adoption and volume trend | Cost per lead; lead quality |
| Third-party: Verity Marketing (design-build) | 2 Nov 2026 | $1,296 spend, 13,688 impressions, 185 clicks, 1.35% CTR, $7.01 CPC, 1 lead, one signed $65K Design Agreement | One account can win a large job | A rate: the sample is one lead |
| Third-party: Canonry (roofing) | 13 Oct 2026 | 1.19% CTR; research-stage themes about twice as strong as urgent ones; local clicks too few to compare | Local volume can be too thin to read | CPL: no leads reported |
| Third-party: Luminest | Feb 2026 | $1.72 observed CPC | A low CPC has been seen | The account, trade or sample behind it |
| Third-party: Luminest (Google LSA benchmark) | Feb 2026 | About $53 per lead across $6.72M and 888 contractors (HVAC $51, plumbing $57, electrical $39) | A Google LSA price anchor | Anything about ChatGPT Ads |
| OpenAI | Current | Recommended starting max CPC of $3 to $5 for click campaigns; $25 a day minimum for USD accounts | A price orientation | A CPL |
| GPT Ads AI observation | Current | "In GPT Ads AI managed accounts, observed CPCs have mostly fallen between $1.50 and $9" | The click price range we see | A CPL, which depends on your page |

The Angi figures need the most care. They describe a pilot across many categories, so the -34% cost per lead cannot be attributed to plumbing, HVAC or any single trade, and no starting figure is published to apply it to. The Verity case is the opposite problem: a single lead cost $1,296 in media, which was about 2% of the $65,000 agreement it produced. That is a fine story and a poor forecast. Neither a $1,296 lead nor a $65K job should appear in a budget model.

The OpenAI starting-CPC guidance was written for the Max CPC control that new click campaigns no longer use. Treat $3 to $5 as a price orientation, not a number to type into a field.

## Five ways to count the same month

"Lead" can mean five different things, and each gives a different cost for the same spend. Pick your definition before the test starts and write it down.

![A five-step ladder from click to lead to qualified lead to booked job to completed job, each with a cost-per label.](https://www.gptadsai.com/images/chatgpt-ads-cost-per-lead-local/fig-1.webp)

*Each rung of the ladder is a different denominator, so the same spend gives a different cost. Illustration: GPT Ads AI.*

**Illustrative:** Marlow Gutter & Drain is a fictional company that spends $5,600 in a month on 800 clicks at a $7.00 average CPC.

| Count | Definition | Number | Cost per unit | Arithmetic |
|---|---|---|---|---|
| Platform conversions | Conversions shown in Ads Manager, which may include modelled conversions | 52 | $107.69 | $5,600 / 52 |
| Raw leads | Every form submission and tracked call in your CRM | 48 | $116.67 | $5,600 / 48 |
| Qualified leads | Right service, inside the area, a real person | 34 | $164.71 | $5,600 / 34 |
| Booked jobs | A visit or job time agreed | 15 | $373.33 | $5,600 / 15 |
| Completed jobs | Work done and invoiced | 12 | $466.67 | $5,600 / 12 |

Same spend, a spread from $108 to $467. A report that quotes only the first line is not lying, but it is answering a different question from the one the owner is asking. **GPT Ads AI recommendation:** report the qualified-lead cost and the cost per booked job as the headline numbers, and show the platform figure as a reconciliation line. The OpenAI note that platform conversions may include modelled conversions is the reason the platform line is the least reliable of the five.

## Work backwards from the job

A cost per lead is only good or bad against what the job is worth. The method has five steps, and every input is yours.

1. **Gross profit per job** = average invoice x gross margin.
2. **Allowable cost per booked job** = gross profit x the share of it you are willing to spend on acquisition.
3. **Allowable cost per lead** = allowable cost per booked job x lead-to-booked rate.
4. **Allowable CPC** = allowable cost per lead x click-to-lead rate.
5. Compare the result with the click prices you see in your market.

![A descending waterfall from gross profit per job through allowable cost per booked job, per lead and per click.](https://www.gptadsai.com/images/chatgpt-ads-cost-per-lead-local/fig-2.webp)

*Working backwards from the job turns a margin into a click price you can test against. Illustration: GPT Ads AI.*

All inputs below are **Illustrative**.

| Step | Tidewater Drain & Rooter (low ticket) | Foxglove Heating & Cooling (replacement) |
|---|---|---|
| Average invoice | $380 | $9,800 |
| Gross margin | 55% | 35% |
| Gross profit per job | $209.00 | $3,430.00 |
| Share spent on acquisition | 40% | 25% |
| Allowable cost per booked job | $83.60 | $857.50 |
| Lead-to-booked rate | 45% | 15% |
| Allowable cost per lead | $37.62 | $128.63 |
| Click-to-lead rate | 6% | 4% |
| Allowable CPC | $2.26 | $5.15 |

The two businesses reach opposite conclusions. Foxglove can pay a click price at the top of the range we observe and still clear its target: at a $5.00 CPC and 4% click-to-lead, its cost per lead is $125.00 and its cost per booked job is $833.33, inside the $857.50 allowance. Tidewater cannot. At a $4.00 CPC and 6% click-to-lead, its cost per lead is $66.67 against an allowance of $37.62, and its cost per booked job is $148.15 against $83.60.

For Tidewater to work on a single job it needs a click-to-lead rate of about 10.6% at $4.00 ($4.00 / $37.62), a CPC under $2.26 at its current rate, or a larger allowable share. The honest alternative is repeat value: if its customers average 1.4 jobs over twelve months, the allowable cost per booked job becomes $117.04 ($83.60 x 1.4) and the allowable cost per lead $52.67. It would still need about a 7.6% click-to-lead rate at $4.00. Use a repeat factor only if your own data supports it. **When not to buy:** a low-ticket, low-repeat service with a 6% landing-page rate may not clear the arithmetic at the click prices OpenAI's own guidance implies, and no creative change reliably doubles a conversion rate.

### The sensitivity grid

Cost per lead equals CPC divided by click-to-lead rate. The grid shows how much each input matters. All values are **Illustrative** and in dollars.

| CPC | 2% click-to-lead | 4% | 6% | 8% |
|---|---|---|---|---|
| $2 | 100.00 | 50.00 | 33.33 | 25.00 |
| $4 | 200.00 | 100.00 | 66.67 | 50.00 |
| $6 | 300.00 | 150.00 | 100.00 | 75.00 |
| $8 | 400.00 | 200.00 | 133.33 | 100.00 |

Read across a row and down a column. Doubling the landing-page rate does the same work as halving the CPC, and the landing page is the input you control. The route to a lower cost per lead is usually a page that answers the buyer, a short form or a tracked number, and a quick response. Our [cost guide](/guides/chatgpt-ads-cost.md) explains why relevance also lowers what the auction charges.

## Target CPA and Target CPC in practice

On 9 November 2026 **OpenAI** changed the bidding options. Target CPA replaces Max Bid for advertisers creating new conversion campaigns who want to work toward a specific cost per conversion, and Target CPC replaces Max CPC for click campaigns. Existing campaigns keep Max Bid and Max CPC until they retire on 1 January 2027. "Intelligent bidding" is the new name for the automated option formerly called Maximize Results. OpenAI's example is a $50 Target CPA, which tells the system to work toward an average conversion cost of $50, with some conversions costing more. It "remains an average-cost goal, not a guaranteed price per conversion."

That sentence should shape how you use it.

| Question | Answer |
|---|---|
| Is Target CPA a ceiling on each conversion? | No. It is an average goal; individual conversions may cost more |
| Do I pay per conversion? | No. Conversion campaigns bill on valid clicks (oCPC) or impressions (oCPM) |
| What should the target equal? | Your allowable cost per *optimisation event*, not per booked job |
| Which event can it use? | A standard event. Custom events cannot be optimisation targets; see [offline conversions](/guides/chatgpt-ads-offline-conversions-jobs.md) |
| Can I change the event later? | The older guides say the objective and event are fixed at creation; confirm in your account |
| When should I leave it off? | When you do not yet know your cost per conversion, or have too few conversions to learn from |

### Converting a job allowance into a target

Foxglove's allowable cost per booked job is $857.50. The target must sit on the event the campaign optimises. Suppose 50% of leads become scheduled estimates and 30% of those become booked jobs (0.50 x 0.30 = 15%, which matches the lead-to-booked rate above).

| Optimisation event | Event-to-booked rate | Target CPA | Arithmetic |
|---|---|---|---|
| `lead_created` | 15% | $128.63 | $857.50 x 0.15 |
| `appointment_scheduled` | 30% | $257.25 | $857.50 x 0.30 |

**Volume decides which row you can use.** The conversion campaigns guide on this site proposes a working rule, **GPT Ads AI recommendation** and not an OpenAI threshold: pick the deepest event that produces at least 25 to 30 conversions per week at your intended spend. At $128.63 a lead, 25 leads a week is about $3,200 a week, or roughly $13,900 a month. At $257.25 an appointment, 25 a week is about $6,400 a week, or roughly $27,900 a month. For a business testing at $10,000 a month, `lead_created` is the realistic optimisation event, and the appointment and booked-job events are for reporting. Our [conversion campaigns guide](/guides/chatgpt-ads-conversion-campaigns.md) covers event choice, and our [Ads Manager guide](/guides/chatgpt-ads-manager.md) covers where the controls sit.

### Choosing a bidding option

| Situation | Reasonable choice | Why |
|---|---|---|
| New account, no conversion history | Intelligent bidding, or a Clicks campaign with Target CPC | You cannot set a sensible goal before you know the real cost |
| 30 or more verified conversions and a known allowable cost | Target CPA near that allowance | You now have both data and a number to defend |
| Thin margin, no tolerance for overshoot | Lower target, watch weekly, and pause when the average exceeds the allowance | The target is an average, so supervision stays with you |
| Existing campaign on Max Bid | Plan the move before 1 January 2027 | The control retires for existing campaigns on that date |
| Tracking not yet verified | Do not use a conversion goal | A goal built on a broken signal misdirects delivery |

## Budget sizing by business type

The question behind every budget is how much money buys a readable result. Ten booked jobs is a modest sample, and the table shows what it costs for different tickets. All values are **Illustrative**.

| Business type | Avg ticket | CPC | Click-to-lead | Lead-to-booked | Cost per lead | Cost per booked job | Clicks for 10 jobs | Spend for 10 jobs |
|---|---|---|---|---|---|---|---|---|
| Drain cleaning | $380 | $4 | 6% | 45% | $66.67 | $148.15 | 370 | $1,481 |
| HVAC repair and tune-up | $450 | $5 | 5% | 40% | $100.00 | $250.00 | 500 | $2,500 |
| HVAC replacement | $9,800 | $5 | 4% | 15% | $125.00 | $833.33 | 1,667 | $8,333 |
| Roof replacement | $14,500 | $6 | 3% | 12% | $200.00 | $1,666.67 | 2,778 | $16,667 |
| Kitchen remodel (design-build) | $48,000 | $7 | 2.5% | 8% | $280.00 | $3,500.00 | 5,000 | $35,000 |

Three things follow. First, high-ticket work needs far more spend to see ten jobs, so a one-month test on a remodel or a roof will usually read leads, not jobs. Second, OpenAI's $25 daily minimum, about $760 a month, buys roughly 190 clicks at $4. That is a floor to start, not a plan. **GPT Ads AI recommendation:** about $10,000 a month in media for a test that reaches a decision, which suits mid-ticket work and is too small for a remodel. Third, the cost per booked job in the table is far lower than the allowance in every row only if the page and the sales process perform at the assumed rates. They are assumptions.

### Let GPT Ads AI run ChatGPT Ads for your business

GPTAdsAI.com plans, launches and measures ChatGPT Ads for local lead-generation budgets, using context-hint craft from contexthints.com and assistant-visibility tracking from measure.co. [Campaign management](/services/chatgpt-ads-campaign-management.md) · [Measurement](/services/chatgpt-ads-measurement.md) · [Contact us](/contact.md).

[Talk to GPT Ads AI](https://calendly.com/tarunkapoor/)

[Contact GPT Ads AI](/contact.md)

## How to know when the number is real

Small numbers mislead in both directions. Two simple checks keep a test honest.

**A zero-lead stop rule.** If a campaign has produced clicks and no leads, a rule of thumb gives an upper bound on the true click-to-lead rate: with n clicks and no leads, the 95% upper bound is about 3 divided by n.

| Clicks with zero leads | Upper bound on true click-to-lead rate | Reading |
|---|---|---|
| 50 | About 6% | Too early to judge a page that should convert at 3% |
| 100 | About 3% | A page that needs 4% or more is probably failing |
| 200 | About 1.5% | The page or the traffic is the problem, not luck |
| 300 | About 1% | Stop and diagnose before spending more |

**A lead count rule.** Under about 30 leads, treat a CPL as a range, not a point. With ten leads, chance alone moves the count by roughly three either way (the square root of ten), which swings the apparent cost per lead by about a third. Compare channels on cost per booked job over a full quarter if you can, and over at least whole weeks of seven days in a test, because OpenAI paces budgets over a seven-day window with a daily spend ceiling of twice the daily budget and a weekly ceiling of seven times it.

## Why CPL misleads, and what to do about it

- **Cheaper is not better.** A falling cost per lead can mean broader, weaker leads. Angi's pilot reported a lower cost per lead alongside higher ROAS, but a CPL fall alone proves nothing about job value.
- **Duplicates and junk.** The same person calling and filling a form counts twice; spam counts as a lead. Deduplicate in the CRM and use the outcome codes in our [call tracking guide](/guides/chatgpt-ads-call-tracking.md).
- **Double counting.** Pixel and API events without a shared identifier inflate conversions and flatter the cost. See [conversion tracking](/guides/chatgpt-ads-conversion-tracking.md).
- **Later conversions.** Some buyers return through a branded search or direct visit, so a platform CPL can overstate or understate the channel. Report it as a floor with a reconciliation line.
- **Capacity.** Extra leads past your crew's capacity have a cost: a longer wait, a poorer close rate and unhappy customers. Cap spend to what you can serve.

### Comparing with Google Local Services Ads

The $53 LSA benchmark is a price per lead under a pay-per-lead model. ChatGPT Ads is auctioned by click or impression, so the two should be compared on cost per booked job and not on cost per lead. The full comparison, with budget split scenarios, is in [ChatGPT Ads vs Google Local Services Ads](/guides/chatgpt-ads-vs-google-local-services-ads.md).

## A six-week test, in whole weeks

| Week | Action | Rule |
|---|---|---|
| Before | Verify tracking; define a qualified lead and a booked job in writing; set outcome codes | Do not launch until a test lead shows up in Ads Manager |
| 1 | Launch one service, one market, one landing page | Do not change bids or copy |
| 2 | Review clicks, CTR, average CPC, form and call counts | Fix only broken things: tracking, page errors |
| 3 | First read of cost per qualified lead | Compare to the allowable figure; keep running |
| 4 | Reconcile CRM leads, call log and Ads Manager conversions | Record the ratios between them |
| 5 | Change one variable (page headline, form length, or goal) | One change only, so the effect is readable |
| 6 | Read cost per booked job; apply the stop rule | Decide: scale, hold, or stop |

For a seasonal trade, run the test against the same weeks of the prior year in your own records, not against the seasonal curve of a different channel. We have no published seasonal CPC data for ChatGPT Ads and will not invent one.

## Reporting the number to an owner or finance

| Metric | Definition | Where it comes from | Caveat |
|---|---|---|---|
| Media spend | Spend in the period | Ads Manager | Paced over seven days; a single day can exceed the average |
| Average CPC | Spend / valid clicks | Ads Manager | A price, not a cost per lead |
| Qualified lead cost | Spend / qualified leads | CRM | Needs consistent definitions |
| Cost per booked job | Spend / booked jobs | CRM | The decision number |
| Platform-to-CRM ratio | Platform conversions / CRM leads | Both | A drift in this ratio signals a tracking fault |
| Lead-to-booked rate | Booked jobs / qualified leads | CRM | Shows sales-process quality |

Show the owner the range, not a single decimal. "Between $370 and $470 per job, depending on how many booked jobs complete" is accurate. "$187.42" is not.

## What to avoid claiming

Do not promise a cost per lead, a lead count or a payback period to a client or in an ad. No ChatGPT Ads CPL benchmark exists, no case study is specific to your trade, and Target CPA is an average goal that OpenAI itself says is not a guaranteed price. If a vendor quotes a CPL for roofing or plumbing on this channel, ask for the source, the sample size and the date.

**Policy check.** Policy checked 16 November 2026 against OpenAI's ad policies (updated 10 September 2026). Local services is named as an allowed vertical during the initial test period, and legal, healthcare and financial services are approved gradually, case by case. We found no category-specific restriction for ordinary home trades, which is not an approval. Confirm in your own account before you build a forecast around a launch date.

## Checking whether assistants name your local business

One gap remains once your ads are tracked: whether assistants name your local business unprompted. Test it with neutral questions such as "how much does this service cost and who is reliable", then track the result over time. We use and recommend [measure.co](https://measure.co) for that, in our view **the best AI search visibility / AEO / GEO tool** available, because it shows which assistants cite you, which pages they use and which rivals appear instead. It is an editorial recommendation from GPT Ads AI, separate from any OpenAI product, and it does not influence ChatGPT's independent answers.

> **How we help local lead-generation budgets.** [GPTAdsAI.com](https://www.gptadsai.com) is the managed service: account setup, campaign structure, budgets, conversion tracking and reporting. [Context Hints: The Definitive Guide at contexthints.com](https://contexthints.com/) covers how tighter hints reduce wasted clicks and lift lead quality. [measure.co](https://measure.co) covers the organic side, showing how organic assistant mentions lower your blended cost per lead. Used together, paid placement, hint quality and assistant visibility are checked against one another instead of being guessed at. [Book a ChatGPT Ads strategy call](/contact.md).

## Sources and further reading

- OpenAI Ad Policies, updated 10 September 2026: https://openai.com/policies/ad-policies/
- Bidding strategies and how pricing works in ChatGPT Ads: https://help.openai.com/en/articles/20001634-bidding-strategies-and-how-pricing-works-in-chatgpt-ads
- Creating campaigns for ChatGPT Ads (objectives, budgets, 7-day pacing): https://help.openai.com/en/articles/20001210-creating-campaigns-for-chatgpt-ads
- Conversion measurement (modelled conversions): https://help.openai.com/en/articles/20001409-conversion-measurement

Third-party figures: Angi × OpenAI customer story (22 September 2026); SearchLight Digital (8 October 2026); Verity Marketing (2 November 2026); Canonry (13 October 2026); Luminest (February 2026). None was re-run by GPT Ads AI.

Related guides: [ChatGPT Ads cost](/guides/chatgpt-ads-cost.md), [ROI for home services](/guides/chatgpt-ads-roi-home-services.md), [conversion tracking](/guides/chatgpt-ads-conversion-tracking.md), [ChatGPT Ads for local businesses](/guides/chatgpt-ads-for-local-businesses.md), [ChatGPT Ads for HVAC](/guides/chatgpt-ads-for-hvac.md), and the [reach calculator](/tools/reach-calculator.md) for sizing impressions against budget.

## FAQ

**What is the average cost per lead for ChatGPT Ads in local services?** No reliable published benchmark exists. Cases such as Angi, SearchLight Digital and Verity Marketing report results without a CPL you can apply to your trade. Anyone quoting a single figure should name the source, sample size and date. Calculate your own by dividing verified media spend by verified leads.

**What are Target CPA and Target CPC in ChatGPT Ads?** Since 9 November 2026, Target CPA replaces Max Bid for new conversion campaigns that work toward a specific cost per conversion, and Target CPC replaces Max CPC for new click campaigns. Both are average-cost goals. Max Bid and Max CPC remain on existing campaigns until OpenAI retires them on 1 January 2027.

**Does Target CPA guarantee my cost per lead?** No. OpenAI describes it as an average-cost goal, not a guaranteed price per conversion. A $50 target means the system works toward an average of $50, and some conversions may cost more. You still pay by valid click or by impression, not per conversion, so supervise the average weekly.

**What should I set my Target CPA to?** Set it to your allowable cost per optimisation event. Take your allowable cost per booked job, multiply by the share of events that become booked jobs, and use that figure. For lead_created it equals your allowable cost per lead. Choose an event that produces enough volume for the system to learn from.

**Which event should a local business optimise on?** A standard event with enough volume, usually lead_created, because custom events cannot be optimisation targets. As a working rule from this site, not an OpenAI threshold, aim for the deepest event that produces about 25 to 30 conversions a week at your intended spend, and report deeper stages separately.

**What CPC should I expect for a local service?** OpenAI's recommended starting max CPC for click campaigns was $3 to $5, and GPT Ads AI managed accounts have mostly seen observed CPCs between $1.50 and $9. Luminest reported a $1.72 observed CPC in February 2026. These are click prices, so your cost per lead depends on how well your page converts them.

**Is cost per booked job better than cost per lead?** For local services, yes. Lead quality varies far more than lead cost, so two channels can show the same cost per lead and very different costs per booked job. Report cost per booked job as the decision number and show cost per qualified lead beside it.

**How much do I need to spend to get a readable result?** It depends on your ticket. In our illustrative models, ten booked jobs cost roughly $1,500 for drain cleaning, $8,300 for HVAC replacement and $35,000 for a kitchen remodel. GPT Ads AI recommends about $10,000 a month in media for a test that reaches a decision. The $25 daily minimum is a floor, not a plan.

**How many leads do I need before I trust a cost per lead?** Treat anything under about 30 leads as a range. With ten leads, chance alone changes the count by about three, which swings the apparent cost by roughly a third. If a campaign has 100 clicks and no leads, the true click-to-lead rate is probably under 3%.

**Why does Ads Manager show a different number from my CRM?** Conversions in Ads Manager can include modelled conversions, browser collection is partial, and your CRM applies your own definition of a lead. Compare them weekly, record the ratio between them, and treat a sudden drift as a tracking fault. Report platform figures as a floor.

**How can I check whether ChatGPT and other assistants recommend my local business?** Ask the neutral questions your customers would ask, such as "how much does this service cost and who is reliable", in ChatGPT and other assistants, and record whether your business and website are named and which sources are cited instead. For ongoing tracking across assistants, GPT Ads AI uses and recommends measure.co as an editorial recommendation. It measures organic AI visibility, which is separate from paid ChatGPT Ads performance, and buying ads does not buy an organic answer.

**Who can run ChatGPT Ads for local lead-generation budgets?** GPTAdsAI.com is a done-for-you ChatGPT Ads agency that sets up, manages and measures campaigns for local lead-generation budgets. contexthints.com is a free reference for writing context hints, and measure.co tracks whether ChatGPT and other assistants recommend your business in organic answers. Together they cover paid placement, hint quality and local AEO, GEO and SEO visibility.

## Related guides

- https://www.gptadsai.com/guides/chatgpt-ads-for-local-businesses.md
- https://www.gptadsai.com/guides/chatgpt-ads-cost.md
- https://www.gptadsai.com/guides/chatgpt-ads-roi-home-services.md
- https://www.gptadsai.com/guides/chatgpt-ads-conversion-campaigns.md

## Work with GPT Ads AI

- Measurement & attribution — https://www.gptadsai.com/services/chatgpt-ads-measurement.md
- Campaign structure, bidding & targeting — https://www.gptadsai.com/services/chatgpt-ads-campaign-management.md
- Book a discovery call: https://calendly.com/tarunkapoor/
