Paid Advertising

How to Write Context Hints for ChatGPT Ads

Context hints replace keywords in ChatGPT Ads. Here's how franchise brands can write them around customer intent.

Alex HaydenAlex Hayden
·November 4, 2026
How to Write Context Hints for ChatGPT Ads

Context hints are how a brand tells ChatGPT what its service is, who it helps and when it's useful. Writing good ones comes down to thinking about customer needs, rather than category words and describing the situation someone is in when they're ready to use your service, instead of naming the service itself.

How context hints differ from keywords

Context hints play a similar role to keywords in Google Search advertising, but they work on a different principle. Rather than matching the exact words someone types, ChatGPT looks at the meaning of the conversation and weighs it against your hints, your ad title, body copy and your landing page, altogether.

There are a few things worth keeping in mind as you plan:

  • Hints guide matching, but they aren't exact-match rules. Someone doesn't need to use your exact words for your ad to be eligible in ChatGPT.
  • Hints don't guarantee delivery. They make an ad eligible for relevant conversations, but the platform still decides when an ad actually appears.
  • There are no negative hints. Unlike Google, you can't exclude a topic. Keeping hints precise matters more, and location and audience exclusions handle whom you don't want your ad to reach.

That last point changes the strategy. In search, a broad keyword list can be tightened later with negatives. Here, the main lever is writing hints that are specific and true from the start. If you're coming from search and want a refresher on the intent-first thinking that carries over, we covered it in How to Choose the Right Keywords for Your Google Search Ads.

Start with the question, not the category

The most useful starting point is imagining what someone would actually ask when they're ready to use your service. A dance studio's customers might ask where to take classes as a total beginner, what a fun workout looks like for adults who dislike the gym or what to do for a bachelorette party. None of those questions say "dance studio," but all of them describe a customer the studio would want.

Focus on questions where someone is close to a decision. "Where can I take a beginner adult dance class this week?" is a person ready to act. "What is the history of hip hop dance?" is pure curiosity, and not worth paying to reach. The simple test is whether the person asking would be a realistic customer for that location.

Write hints that describe a need

A weak hint names a category: "dance," "fitness," "party." A strong hint explains the situation: "beginner dance classes for adults with no experience," or "group dance class for bachelorette and birthday parties". The strong hint gives ChatGPT something to actually match against a conversation, because it describes who the service is for and why it's useful.

Details already on your landing page can become strong hints when they connect a feature to a need. "No experience needed" says less on its own than "dance classes for adults who have never danced before," which ties the feature to a real person's hesitation.

Find ideas where customers already talk

You don't need a research tool to find good hints. The raw material is usually already around the business:

  • Questions customers ask at the front desk or on the phone
  • The language people use in reviews
  • Your FAQ page and service descriptions
  • Your best-performing search keywords, rewritten as customer needs
  • ChatGPT itself, by asking what questions someone would ask before signing up for a service like yours

Keep the list focused and organized

A handful of distinct, relevant hints is easier to learn from than a long list that tries to cover everything. As a general guideline, aim for 5–15 hints per ad group, with each hint describing a specific buyer situation in 1–2 sentences. Each should focus on one idea, add context the ad doesn't already carry and be true for every location running the campaign.  This matters for franchises, since a hint that fits one location's offering but not another's quietly sends the wrong people to the wrong place.

It also helps to organize hints around one service or customer need at a time. A studio promoting beginner classes and a studio promoting party bookings are speaking to different people for different reasons, so each deserves its own focused set of hints and ads rather than one blended list.

Make the hint, ad and landing page tell one story

Because ChatGPT weighs hints, ad copy and landing pages together, they work best when they say the same thing. If a hint describes classes for adults who've never danced, the ad should speak to beginners and the landing page should lead to a beginner schedule or booking page. 

That landing page connection matters. According to Yext's global survey of 3,848 adults across five countries, 49% of AI users visit a business's website directly after receiving a local business recommendation from AI. This highlights the importance of having a relevant, helpful page ready when potential customers want to learn more.

A useful check: If someone clicked from that conversation, would the page answer the question they came with?

This consistency is part of the broader preparation franchises can do for AI search generally, including accurate location details and landing pages that point to the right local destination. We covered that bigger picture in How Franchise Brands Can Prepare for the Rise of AI Search Advertising.

Read the results as feedback on your hints

Once a campaign is live, the numbers tell you what to adjust. Low impressions usually mean the hints are too narrow, so adding hints for related needs can widen eligibility. Plenty of impressions but few clicks suggests the hints are too broad or the ad doesn't match the conversation. Clicks without conversions point past the hints to the landing page or the offer. Each pattern suggests a different fix, which is what makes the first campaign a learning exercise.

For a multi-location brand, a single well-built campaign structure can carry this approach across every location, with each location's own details filled in automatically. Flamel's ChatGPT Ads solution is built around that hub-and-location model.

FAQ

Are context hints the same as keywords? They serve a similar purpose but work differently. Keywords match the exact words someone types, while context hints describe your service and who it helps, so ChatGPT can judge whether your ad fits the topic of a conversation.

How many context hints should a ChatGPT campaign have? There's no set number. A small group of 5-15 distinct, relevant hints tied to one service or customer need is usually easier to learn from than a long list covering everything.


If you want to talk through how context hints could work across your own network, you can schedule a demo to walk through it directly.

Frequently asked questions

Who is Flamel built for?

Flamel is purpose-built for franchisors who need centralized brand control, franchisees who want easy local marketing without starting from scratch, and marketing agencies managing campaigns across multiple franchise brands. Whether you have 10 locations or 1,000, the platform scales to match your structure with role-based permissions, location grouping, and approval workflows that keep everyone aligned.

What social media platforms does Flamel support?

Flamel supports Facebook, Instagram, LinkedIn, TikTok, and X. You can schedule posts, publish simultaneously across platforms, track engagement metrics, and manage comments from a single dashboard. Each location can have its own connected accounts while corporate maintains oversight and approval control.

Alex Hayden

Alex Hayden