ChatGPT and Claude in a marketing team
Short answer
ChatGPT and Claude are general AI assistants a marketing team can use to research customers, draft copy, analyse exported data and turn notes into briefs. Both write well and work fast, and both can state something false with complete confidence. They earn their keep when they are given your brand material and real numbers, and when a person checks everything before it is published or sent.
By Jakub Cambor, founder of AI for Marketing, updated
How do you use ChatGPT for marketing?
You use ChatGPT for marketing by giving it one clear job at a time along with the material the job needs: your brand guide, last month's ad export, a batch of customer reviews or the notes from a sales call. Then you read what comes back as a first draft, check it, and edit it into something you would put your name to.
- Customer research: paste in reviews, support tickets or call notes and ask for the words customers use, the objections they raise and the reasons they give for buying.
- Ad and email drafts: ask for a dozen headline options for a Google Ads responsive search ad, or three subject lines for a Klaviyo campaign, each tied to one of those customer reasons.
- Analysis: upload a CSV from Google Ads, Meta Ads Manager or GA4 and ask which campaigns spent the most for the least return, then check its sums in the spreadsheet.
- Briefs: turn a messy thread of notes into a one-page campaign brief with the audience, the offer, the message and the measure.
- Concept visuals: use its image generation for mock-ups and mood boards, and leave the finished creative to your designer or Canva.
The setup matters more than the prompt. On the paid plans, Projects and custom GPTs let a team keep the brand guide, product facts and examples of approved copy in one place, so nobody pastes them in again for every job. OpenAI lists Projects, scheduled tasks and custom GPTs on the Plus plan, and a shared workspace with admin controls on Business.[1][2]
- 1
Pick the right plan
Use a business plan if any customer data will go near it, so your content is not used for training by default.
- 2
Make one project per brand
Load the brand guide, product facts, pricing rules and a few pieces of copy you are proud of.
- 3
Write standing instructions
Voice, spelling, words to avoid and claims the brand may never make, in plain sentences.
- 4
Keep a prompt library
Save the prompts for jobs you repeat, such as the monthly ad refresh or the weekly summary, with an example of a good answer.
- 5
Check before it ships
One named person checks facts, figures and tone against the source before anything reaches a customer.
Search behaviour differs between the two markets. In the US, far more people search for ChatGPT for marketing than for Claude; in the UK it runs the other way, with Claude searched more often.
| Search | UK | US |
|---|---|---|
| chatgpt for marketing | 30 | 1,000 |
| claude for marketing | 90 | 390 |
| All of these | 120 | 1,390 |
Monthly Google searches for each phrase, in each country.
Source: Google Ads monthly search volume, via DataForSEO (September 2026)
Is Claude good for marketing?
Yes. Claude is good for marketing work that depends on reading a lot and writing carefully: turning long documents, call transcripts and research into briefs, drafting copy that holds a consistent voice, and editing. Like ChatGPT, it can be confidently wrong, so its drafts need the same checking.
In practice the two are closer than their fans suggest, and most teams would do good work with either. The differences that matter show up in the plans and the extras around the model. Anthropic lists Projects, Artifacts, connectors to your apps, and Claude Design, Slides and Docs on its paid plans, while OpenAI puts image generation, deep research, scheduled tasks and custom GPTs on its paid plans. Pick one for the whole team so the brand material and the prompt library live in one place.[4][1]
| Plan | ChatGPT (OpenAI) | Claude (Anthropic) |
|---|---|---|
| Free | Yes, $0 a month | Yes, $0 a month |
| Main individual plan | Plus, $20 a month | Pro, $20 a month billed monthly, or $17 a month billed annually |
| For heavy daily use | Pro, from $100 a month | Max, from $100 a month |
| Team seat | Business Standard, $25 a seat a month billed monthly, or $20 billed annually | Team Standard, $25 a seat a month billed monthly, or $20 billed annually |
| Team size | 2 to 200 people | 2 to 150 people |
| Training on your content | Individual plans: you can opt out. Business: none by default | Individual plans: you can opt out. Team: none by default |
Sources: OpenAI, ChatGPT pricing, opened 29 September 2026; OpenAI, ChatGPT Business and Enterprise pricing, opened 29 September 2026; Anthropic, Claude plans and pricing, opened 29 September 2026
The prices are in US dollars as the vendors' pages showed them at the time of writing, September 2026. Anthropic's page says its prices exclude tax, and what you pay locally depends on your country's currency and VAT. The row to act on is the last one: on the individual plans your content can be used to improve the models unless someone switches that off, while on both team plans it is not used by default.[4][1][2]
What can ChatGPT and Claude do for a marketing team, and what can they not?
They can do most of the drafting, summarising and first-pass analysis in a marketing team, quickly and for the price of a subscription. They cannot know your customers, your numbers or your brand unless you give them that material, they cannot tell a true claim from a plausible one, and they take no responsibility for what gets published.
| Job | Good at | Weak at |
|---|---|---|
| Copy | Many variations fast, in a voice you have shown them | Knowing which variation will sell; that takes a live test |
| Research | Summarising reviews, transcripts and long reports | Anything they have not been given or cannot search for |
| Analysis | Spotting patterns in an export and explaining them plainly | Arithmetic over large files, which needs checking |
| Strategy | Laying out options and questioning a plan | Weighing your margins, capacity and history, which they do not have |
| Facts and claims | Finding where a claim needs a source | Inventing figures, quotes and product details that sound right |
| Customer data | Working on data with personal details removed | Being the place customer records are kept or processed |
Three habits keep a team safe. For brand voice, give the assistant examples of copy you approved and a short list of words and claims to avoid, rather than adjectives such as friendly or premium. For facts, check every figure, price, quote and product detail against its source, because a made-up statistic reads exactly like a real one; OWASP lists misinformation among the main risks of language model applications. For customer data, use a business plan where your content is not used for training by default, strip names and contact details before anything goes in, and follow your own data policy; UK teams can start from the ICO's guidance on AI and data protection.[5][6][2][4]
That rule is how the system behind our managed service works: AI drafts at volume, Jakub reviews, and the client approves before anything goes out. When HEAVENSAKE, a premium sake brand, wanted content in its own luxury voice, the answer was an agent built from the brand's own materials whose every output is a draft for the brand team. It was delivered in full, and the brand chose not to deploy it into its live workflow.[7]
Proof from our own work
$528K
Revenue tracked from Google Ads
$528K over 12 months (October 2023 to October 2024), with the highest search impression share among tracked competitors (20.51%, October 2024).
ToastPal, Google Ads across the US, UK, Australia and Canada. Delivered by Jakub as a freelance consultant before founding AfM.
Read the case study4
Brand-trained SEO content agent
An agent with 4 capabilities built for a premium sake brand, delivered to the client.
HEAVENSAKE, a premium sake brand. Delivered to the client, not yet deployed on their site.
Read the case study10,976
Organic search from AI-written content
10,976 organic clicks and 770,576 impressions in 152 days, with no paid promotion.
Harmonance, Jakub's own project, not a client result. It is where the AI content method was first tested; client work adds human review before anything is published.
Read the case study
What we do
We run your marketing for you and make it pay for itself.
- First comes a free written breakdown: we look at your site and any numbers you choose to share, with no logins, and show you where the money leaks and the first thing we would fix.
- Then we fix and run the connected commercial journey: the work that earns attention, the pages that turn it into a lead or sale, and the follow-up that creates the next valuable conversion.
- Everything is actively managed and improved, with a weekly report showing what changed and what it made, and you approve everything before it goes out.
- Rolling monthly. No lock-in, cancel any time.
What it is not
- Not a single-channel agency: we do whatever makes you money next, not ads forever or SEO forever.
- Not a junior team on a six-month lock-in: you work with Jakub directly, rolling monthly.
- Not AI left to run on its own: every piece is reviewed by a person and approved by you.
We also build custom automations, such as reporting digests, lead enrichment and routing, and support triage in tools like n8n, Make and Zapier. They are an add-on to the managed service, not a separate offer.
More questions
Should my team use ChatGPT or Claude?
Either will do the everyday work well. Choose on the plan and the tools around it: which one connects to the apps your team already uses, whether you need image generation built in, and which team plan fits your size. Then put the whole team on the same one, on a team plan, so the brand material and the saved prompts live in one place.
Can ChatGPT or Claude run my ads?
Not on their own. They can research keywords, draft the ads, suggest tests and explain an export, but the bidding and targeting happen inside Google Ads and Meta Ads Manager, and someone has to set the budget, approve the ads and read the results against what you sell. Treat them as a fast assistant to the person running the account.
Related answers
- AI marketing toolsThe best AI marketing tools are the ones that do a job you already have, inside the systems you already run.
- AI content marketingAI content marketing means using tools such as ChatGPT, Claude or Gemini to research, draft and adapt marketing content while people supply the ideas, the facts and the final judgement.
- AI in marketingAI marketing is the use of AI models to do marketing work: researching customers, writing and testing ads, building pages, following up leads and reporting results.
Sources
- 1.OpenAI, ChatGPT pricing, opened 29 September 2026chatgpt.com
- 2.OpenAI, ChatGPT Business and Enterprise pricing, opened 29 September 2026openai.com
- 3.Google Ads monthly search volume, via DataForSEO (September 2026)dataforseo.com
- 4.Anthropic, Claude plans and pricing, opened 29 September 2026claude.com
- 5.OWASP, Top 10 for Large Language Model Applications (2025)genai.owasp.org
- 6.Information Commissioner's Office, Guidance on AI and data protectionico.org.uk
- 7.Brand-trained SEO content agent: AI for Marketing case study