In-Answer Ads, the First Month: What Steps Can a Business Take Today?
The first month of in-answer ads is a month of preparation and a small test. It is not the month to decide whether the channel stays; it is the month to learn how the surface works and whether it reaches your customers at all. Businesses that rush this month usually end it with numbers they cannot interpret. 🗓️
This page is the first step of the in-answer advertising horizon series. It covers what is realistic in a month, which surface to start with, a four-week task list, typical first-month mistakes and what is rising on this horizon. Technical details were checked on 12 November 2026; forecasts are labelled as forecasts and reviewed in the update log.
What is realistic in the first month?
BU BÖLÜMÜN ÖZETİ
- Preparation comes first
- A small test
- Observation, not decision
A month is enough to learn a new ad surface and run a small test. It is not enough to decide whether the channel is permanent; that decision comes in the third month at the earliest.

Preparation comes first
The first two weeks are preparation weeks, not ad weeks: the conversion is defined, measurement is set up and the landing page is fitted to the offer. A test that starts without preparation cannot answer the question “did it work?” at the end. Preparation is the part of the month that keeps its value whatever the test shows, because measurement and a better page also help every other channel the business runs. 🧰
A small test
In the second half of the month, a test runs with one offer, in one region, on a limited budget. The aim is not profit but to see whether the surface reaches the business’s customers. On ChatGPT, the minimum daily budget depends on the billing currency, so the budget is set slightly above that floor and at a level that does not strain the business. A small, clean test teaches more than a large, messy one. How to open the account is covered in how to open a ChatGPT ads account.
Observation, not decision
At the end of the first month you write an observation note rather than a decision: how many impressions, clicks and conversions, and what happened on the landing page. Deciding firmly on small numbers is misleading, because one customer can swing the result. The first month’s note becomes the input to the second month’s plan. A written note is the cheapest tool in the whole test. 📝
Which surface should you start with?
The two surfaces are in different states, so the starting points differ too. The choice depends on where your customers ask their questions.
ChatGPT: a direct test
ChatGPT ads are set up in a separate ads manager and are available to legal entities in Türkiye. For businesses whose customers ask long questions and compare options before deciding, the first test can happen here. The account belongs to the business, and agencies are added as users rather than opening accounts on a client’s behalf. The setup takes little time; the preparation behind it is what decides whether the test can be read. Start with one offer that you can explain in a single sentence. 💬
Google: through existing campaigns
Google’s ads inside AI answers are not bought through a separate campaign; eligibility comes from existing Performance Max, Shopping and broad match or AI Max campaigns. So on the Google side, the first month is about fixing assets and measurement in campaigns you already run, not building new ones. The account-side checklist is in preparing an account for AI Mode ads. The formats were announced as tests on English-language searches in the United States, so for Turkish, preparation is the right first step.
A four-week task list
BU BÖLÜMÜN ÖZETİ
- Week 1: measurement
- Week 2: the landing page
- Weeks 3 and 4: a small test
The sequence below is enough for a small business to get through its first month in order. Each week builds on the one before.

Week 1: measurement
The conversion is defined: a form, a call, a sale or a booking. Measurement is set up and confirmed with a trial conversion. No ad goes live before measurement works, because the campaign objective cannot easily be changed later and a campaign started with the wrong objective has to be rebuilt. If you are unsure what to count, start with the action closest to revenue that is easy to verify. How to read results properly is explained in Google Ads performance measurement. 📏
Week 2: the landing page
The page the ad will lead to is reviewed: is the offer on the first screen, is the price visible, does contact take one step? Every claim in the ad must be backed up on the page. OpenAI says it may crawl landing pages for review, and a mismatch between ad and page can lead to rejection or wasted clicks. Read the page as a stranger would: if the offer is not clear within five seconds, fix the page before writing a single ad. The page is the other half of the test.
Weeks 3 and 4: a small test
The test runs with one offer, one region, two ad texts and a limited budget. You look at it twice a week, not every day; frequent changes ruin a small test because the system never settles. At the end of the fourth week, the observation note is written. The note is the second month’s roadmap: what to keep, what to change and which single change to test next. 🧪
Which mistakes happen in the first month?
BU BÖLÜMÜN ÖZETİ
- Writing with keyword logic
- Starting without measurement
- Deciding on one ad
Most first tests break in the same three places. None of them is technical; all of them come from carrying old habits onto a new surface.
Writing with keyword logic
Advertisers with search experience tend to write context hints like keyword lists. These hints describe the topic of a conversation and do not guarantee exact matches. Instead of a word list, write a plain description of the customer’s situation: who they are, what problem they have and what stage they are at. A hint that reads like something a customer might actually say usually matches better than a string of terms. Describing the person works better than listing the words. ✍️
Starting without measurement
A campaign launched without measurement tells you nothing beyond the number of clicks. Clicks are not the result the business is looking for. A test whose conversions are not measured is wasted budget, however cheap it is. Even a simple setup, with one conversion confirmed by a trial, turns the same spend into something you can learn from. If measurement cannot be set up this month, postpone the test rather than launching blind. A delayed test costs nothing; a blind one costs money.
Deciding on one ad
A test with a single ad text tests that text, not the surface. A poor result may come from the copy or from the surface. Testing with at least two different texts makes it possible to tell where a result comes from. Keep the difference between the two clear, for example one leading with the offer and one leading with the customer’s situation, so that the comparison actually teaches you something about the language that works here. 🔁
What is rising on this horizon?
BU BÖLÜMÜN ÖZETİ
- Today: the ads manager is changing fast
- Forecast: early testers learn
- Forecast and log
Most of what will happen on a one-month horizon has already begun. The first item is a present-day observation; the next two are forecasts.

Today: the ads manager is changing fast
The ChatGPT ads manager is in beta and its help pages are updated often; the country list was expanded several times during the year. That pace means a detail that was right a month ago may have changed today. Check the current help page before any setup, and note the date you checked it. A dated note saves you from arguing later about what the rules were at launch. 📆
Forecast: early testers learn
Forecast: in the coming months, businesses that learn this surface through a small test will build more accurate campaigns with less budget when competition increases. The value of entering early lies not in profit but in the lesson learned: how your customers phrase their needs, which offers they respond to and what your landing page is missing. Those lessons transfer to other channels too, which is why even a test that ends with a decision to stop is rarely wasted. 🌱
Forecast and log
Forecast: Google’s new formats will spread first on English-language searches, and giving a date for Turkish would be wrong. Update log. 12 November 2026: first publication. The forecasts on this page are re-read when they fall due; a forecast that turns out wrong is not deleted but corrected here. The next step in this series is the in-answer ads: three months page, which covers reading the first results and moving to a regular campaign.
How does the first month work with Adapte Digital?
BU BÖLÜMÜN ÖZETİ
- A single delivery package
- What is delivered
- Where to start
We do the first month’s preparation as delivered work: measurement, the landing page and the test campaign are set up and handed over to you.
A single delivery package
Adapte Digital does not charge a monthly management fee. Work starts with a single delivery package whose scope is written down in advance: 36,000 TRY + VAT. If the work grows, the new need is defined as a separate add-on package, priced by what will be delivered rather than by the hours spent. 📦
What is delivered
The output of first-month work is working measurement, a reviewed landing page and the structure of the test campaign. We look at the state of your existing ad account first, because a new surface added to a broken account inherits its problems.
Where to start
Which packages are open, and at what price, is shown on our campaign pages, item by item. 🧭
Quick Summary
- The first month is for preparation and a small test.
- ChatGPT can be tested directly; on Google, existing campaigns are fixed.
- Order: measurement, landing page, small test.
- Keyword logic, no measurement and a single ad are typical mistakes.
- Forecasts are reviewed in the update log.
Short Glossary
- Observation note
- A short written record of a test’s numbers and what happened, used to plan the next step.
- Trial conversion
- A test action performed to confirm that conversion tracking records correctly.
- Billing currency
- The currency an ad account is billed in; on ChatGPT it also sets the minimum daily budget.
Frequently Asked Questions
Next Step
This week, write down which single action counts as a conversion for your business and check that it is tracked. Our campaign pages list the packages that cover measurement and landing page work.
Sources: OpenAI Help Center — Ads Manager Availability · Search Engine Journal — How To Advertise In Google AI Mode For Ecommerce
Updated: November 2026
Sık Sorulan Sorular
For a small business, testing both surfaces in the first month makes the results hard to read. One surface is tested directly while the other is prepared. One variable at a time is the most reliable rule of the first month; the second surface can be added in the third month, once the first test has been read. Sequence beats speed at this stage. ⚖️
We do not give a figure, but the budget should sit slightly above the surface’s daily minimum and at a level that does not strain the business. The aim is learning, not profit.
No separate campaign is built. The assets and measurement of existing Performance Max, Shopping and AI Max campaigns are fixed.
At the earliest in the third month, after two test rounds have been compared. The first month produces an observation, not a decision.
