As ChatGPT Ads expand, more PPC teams are going to face a familiar question: Should we move some of our paid search budget there?
It makes sense that search marketers would be part of that conversation. Many of us already manage the budgets most likely to be considered for a ChatGPT Ads test. But I’m hesitant to treat ChatGPT as another search channel simply because the budget may come from paid search.
People aren’t using ChatGPT the same way they use Google Search. They’re having longer conversations, asking follow-up questions, researching options, and getting help with decisions. Advertisers also have a different set of targeting signals and far less history to rely on when measuring performance.
So when someone asks whether they should move paid search budget to ChatGPT Ads, there isn’t a simple answer based on cost per acquisition or return on ad spend.
You first need to know what you’re expecting ChatGPT Ads to contribute, which users you’re trying to reach, and what you’re willing to give up elsewhere to fund the test. Those answers should come before deciding how much budget to move.
1. What Job Are You Expecting ChatGPT Ads To Do?
Before deciding how much to spend on ChatGPT Ads, I’d want to know what we’re trying to get from that investment.
For some advertisers, ChatGPT could introduce a brand while someone is still researching a category or comparing options. Others may see an opportunity to reach people much closer to making a purchase or submitting a lead.
Those scenarios call for very different tests.
OpenAI can use the context and intent of a conversation to determine which ads may be relevant. That gives ChatGPT more information about what someone is discussing than we’d typically get from a single search query.
But having access to intent doesn’t mean every ChatGPT Ads campaign should be treated as another way to capture existing search demand.
If the goal is to reach people earlier in their research, I’d expect the test to look different from one focused on direct-response conversions. The budget source, campaign expectations, and measurement shouldn’t be identical for both.
This is where I think PPC teams need to be careful about carrying over too many assumptions from paid search. We have years of data telling us what certain searches are worth and how different types of intent tend to perform. We don’t have that same history with ChatGPT Ads yet.
Before deciding whether ChatGPT can fill a specific role in the media mix, we need a better understanding of the intent we’re paying to reach.
See also: ChatGPT Ads: New Acquisition Channel Or Just Another Brand Tax?
2. Do You Understand The Intent You’re Buying?
With paid search, we have a pretty good idea of what we’re buying.
If someone searches for “CRM software for small business,” we can see the search term, understand which keyword matched it, and use historical performance to determine how valuable similar searches have been.
ChatGPT Ads don’t give advertisers that same view into intent.
OpenAI can use the context and intent of the current conversation, along with signals from the ad, landing page, targeting selections, and advertiser-provided Context Hints to determine which ads may be relevant. Advertisers don’t receive users’ conversations or chat histories.
Context Hints are also different from the keywords PPC teams are used to working with. Advertisers can provide topics, keywords, or descriptions of conversations that may be relevant to their offering, but those inputs don’t work like exact-match keywords or guarantee an ad will appear within a specific conversation.
So OpenAI may have a lot of context about what someone is trying to accomplish without giving advertisers the same visibility we’ve historically had into the queries driving our ads.
For PPC teams, that requires some adjustment and mindset shifting.
We’re used to looking at search terms, identifying patterns in intent, adding negatives, restructuring campaigns, and changing bids or messaging based on what people are searching for. Some of those familiar levers either work differently in ChatGPT Ads or aren’t available in the same way.
That doesn’t mean the traffic is less valuable. We just have less history to tell us which types of conversations are likely to produce the outcomes we want.
And if we’re still learning what that intent is worth, I’d be cautious about taking a meaningful amount of budget away from channels where we already have years of performance data.
See also: ChatGPT Calls Lead On Quality, But Not Conversions
3. Where Is The Test Budget Coming From?
Most advertisers aren’t getting a new budget just to test ChatGPT Ads. In this economy, who’s getting new budget for anything?
The money will likely have to come from somewhere else.
Paid search may be the first place teams look, especially if PPC is responsible for running the test. But I wouldn’t pull budget from Search simply because ChatGPT Ads happens to sit with the same team.
If branded search is efficiently capturing existing demand, what would you gain by reducing that investment? The same goes for non-brand campaigns already producing qualified leads at an acceptable cost.
There may still be room to move budget.
A mature Google Ads account can have campaigns or segments that aren’t earning their share of the spend. ChatGPT Ads may be a better use of some of that money, but the comparison should make sense based on what you’re testing.
If the goal is to reach people earlier in their research, look at what you’re already spending to do that. If you’re expecting direct-response conversions, compare ChatGPT against other channels competing for the same outcome.
I also wouldn’t assume the entire test budget has to come from one channel. There may be a better case for pulling smaller amounts from several areas than cutting a proven Search campaign by 20%.
Whatever you decide, you should know what performance would justify keeping that budget in ChatGPT Ads. Otherwise, it’s difficult to know whether the test earned more investment or whether the money should go back to the channels that funded it.
See also: Google Ads Budget Misallocation Is More Common Than You Think – And Harder To Spot
4. How Will You Measure Whether Your ChatGPT Ads Test Worked?
Before moving budget, I’d decide what ChatGPT Ads needs to produce to keep that investment.
Advertisers have more measurement options than they did during the initial pilot. OpenAI now supports Pixel and Conversions API tracking, UTMs, and conversion optimization.
That gives PPC teams several ways to connect ad activity with actual business outcomes.
If you’re testing ChatGPT as a direct-response channel, the measurement plan may be fairly familiar. Track leads or purchases, pass qualified lead data back when possible, and look beyond the initial conversion.
The harder part is measuring a test designed to reach people earlier in their research.
If that’s the role you’ve assigned ChatGPT Ads, last-click conversions won’t tell you everything. But I also wouldn’t use that as an excuse to fall back on softer metrics like clicks or CTR.
You still need to decide what evidence would make you comfortable investing more.
That could include assisted conversions, changes in branded search behavior, CRM data, or other signals you already use to evaluate upper-funnel media. The right combination will depend on the business and what the test is designed to accomplish.
I’d make those decisions before the campaign launches. Otherwise, almost any result can look promising if you choose the success criteria after seeing the data.
There’s also another measurement question worth considering: What should paid ChatGPT Ads get credit for when your brand may already be appearing organically in ChatGPT?
5. How Do ChatGPT Ads Fit With Your Organic ChatGPT Visibility?
On Google, PPC and SEO have always been separate disciplines. But both can put a brand in front of someone on the same search results page.
ChatGPT works differently.
OpenAI keeps ads separate from its answers. Advertising doesn’t influence what ChatGPT says, and brands can’t pay to rank within an organic response. Product recommendations are also selected independently from ads.
A brand could advertise against relevant conversations while rarely appearing in ChatGPT’s answers. Another could already have strong organic visibility without spending anything on ChatGPT Ads.
I’d want to understand where your brand falls before deciding what paid ads could add.
There isn’t a native ChatGPT report showing every question where your brand appears organically. For a smaller analysis, marketers can build a list of questions their customers are likely to ask and test them manually. AI visibility tools can automate that process at scale and track brand mentions, competitors, and cited sources over time.
From there, pay attention to where competitors show up instead and which sources ChatGPT references around your category. That can help identify areas where paid placements could extend your visibility or reach users you’re currently missing.
It can also expose gaps that advertising won’t address.
If your brand rarely appears in relevant ChatGPT answers, increasing ad spend won’t change what ChatGPT says organically. That may require separate work across SEO, content, digital PR, or other areas influencing how your brand is represented online.
For marketers considering ChatGPT Ads, I’d look at paid and organic visibility within ChatGPT separately before deciding where more investment makes sense.
ChatGPT Ads Need Their Own Case For Budget
PPC teams will probably be responsible for many of the early ChatGPT Ads tests. That still doesn’t mean paid search should automatically be the budget source.
I’d make the case for ChatGPT Ads based on what you expect the channel to contribute. From there, you can decide which existing investments are competing for the same outcome and where there may be room to test.
For some advertisers, that could mean shifting money from paid search. For others, the better comparison may be paid social, display, video, or another part of the media budget.
We’re also still early enough that some of these decisions will change as advertisers get more data. OpenAI continues to add targeting, measurement, and optimization capabilities, while marketers are learning where ChatGPT fits within the buying process.
That’s a good reason to test. It’s also a good reason to be deliberate about what you’re giving up to fund it.
For marketers looking for a closer look at the channel, Search Engine Journal is hosting ChatGPT Ads & GEO: A Roundtable With Go Fish Digital & OpenAI on September 22 at 2 p.m. ET. The discussion will cover how ChatGPT Ads work today, how paid placements differ from search, and where advertising fits alongside organic visibility in ChatGPT.
More Resources:
- 6 Months Into ChatGPT Ads, Advertisers Still Don’t Know What ‘Good’ Looks Like
- How To Leverage AI Ad Placements For Stronger PPC Performance
- PPC Pulse: ChatGPT Ads CPMs, Ads Decoded Talks Analytics
Featured Image: PeopleImages/Shutterstock
