The (Totally Free) AI Prompt Hack That Gives You Better B2B Content Ideas

If you ask most AI large language models (LLMs) for a list of blog ideas suitable for your business, the output is usually what I would call "fine". Not great, not terrible, but serviceable. Useful in parts, but rarely worth getting excited about.

The good news is that there is a simple way to use AI models like ChatGPT to generate more timely B2B content ideas, and it's really easy to do. All you need to do is think a little differently about how you're prompting your AI and you'll be presented with more timely, audience-specific B2B content ideas. It's the kind of output you'd expect from using buyer or data intent platforms, which are incredibly useful (but come with a hefty monthly subscription fee - more on that later).

You can try it in ChatGPT in under five minutes, and repeat it whenever you need a fresh editorial calendar.

It works because of how you frame the question. Instead of asking for content ideas in general, you tell the AI exactly who you want to reach and when. That one change can turn a generic list into a set of topics that feel like they have been pulled from a current market insight report.

The experiment

To explain how it works, let me take you through an experiment I did recently, where I ran two prompts back-to-back in ChatGPT.

Prompt 1: Give me some blog topic ideas for B2B content marketing.

The output was a list of what I call "evergreen topics". They're useful for filling your content archive, but there's nothing particularly groundbreaking about them. Ideas like “How to Repurpose a Webinar” or “Metrics That Prove ROI” are always relevant, but they are not tied to any specific moment or pressing pain point that your target audience is grappling with right now.

Prompt 2: What are senior Australian B2B marketers asking you about right now?

The output was completely different. It produced topics reflecting current conversations in the Australian market, and they're the things that are really keeping my own target audience up at night. This prompt surfaced a list of questions about omnichannel strategy, generative AI adoption and how marketers are using AI for account-based campaigns.

The only difference between the two prompts was the framing. One was general, while the other defined the audience and set it in time.

Why this works

LLMs work by predicting the next most likely words based on the patterns in their training data and the context you give them. If your prompt is broad (e.g. "give me some blog ideas about digital transformation in the finance industry"), the model draws from equally broad patterns. The results tend to be generic because they are designed to fit almost anyone.

When you define the audience and set the time frame (e.g. "what are Australian CFOs asking you about technology related themes right now"), you give the model a much narrower field to work within. It will pull from more specific and recent patterns, including news coverage, industry commentary and topical discussions.

Using the finance prompts above, this is what I got:

Generic prompt ("give me some blog ideas about digital transformation in the finance industry"):

• From Core Systems to Cloud: Why Finance’s Transformation is About More Than Technology

• Beyond Compliance: How Digital Transformation Can Unlock Competitive Advantage

• The New Operating Model for Finance: Agile, Data-Driven, and Customer-Centric

Timely, persona-specific prompt ("what are Australian CFOs asking you about technology related themes right now"):

• AI and Generative AI – Driving Productivity (and Seeking Value)

• Automation and Financial Transformation – Moving Beyond Spreadsheets

• ROI Clarity – How to Measure Tech Investments

The first set is generic. It reads like the kind of ideas a B2B marketing team might brainstorm on a whiteboard before taking them to sales for validation. They're fine, but probably too high-level to make your average CFO stop and take notice right now.

The second set is timely. It is anchored in real Australian CFO priorities and brings in the kind of language and pain points you would expect from a current market report. That makes it far easier to shape into thought leadership that feels relevant and credible today.

If LLMs are evolving to think more like people — and yes, that idea is a little uncomfortable to me too — then the best way to use them is to interact more like you would with a person. Think of it this way: if you were asking someone on your sales team for blog ideas, would you say "give me a list of blog ideas"? Or would you get a better outcome asking "What are we hearing from our customers and prospects right now”?

Generic vs timely content ideas

This is not to undermine the importance of evergreen content in your library. Both types of ideas have a place in your content plan. The real value comes from knowing when to use each.

• Generic prompts give you stable, repeatable (evergreen) topics you can use any time. They are ideal for building an always-on content library, creating SEO cornerstone pages, or filling gaps in your publishing schedule.
  

• Persona-specific, time-bound prompts give you ideas that match the conversations your audience is having right now. They are ideal for timely thought leadership, campaign-led content and event-linked assets where relevance and speed matter.

The risk is leaning too heavily on one or the other. A calendar built only on generic topics risks sounding detached from the market, while one built only on timely topics risks burning through its shelf life too quickly. The strongest approach blends both.

Do I still need expensive intent data tools?

The answer to this question is "it depends". Market sensing and intent data platforms — tools like 6sense, Demandbase and Bombora — are generally built for the big end of town. They are designed to help enterprise sales and marketing teams identify and engage accounts in large, complex deal cycles.

If you are working in that environment, the investment makes sense. These platforms give you account-level intent signals, competitor insights and buying-stage data that go well beyond what you can prompt from an LLM.

However for most small to mid-sized B2B teams, the ROI is usually harder to justify. You may not need account-by-account insights, and the subscription fees can outweigh the benefits. In those cases, more focus on better AI prompting can deliver a useful taste of what those tools offer: timely, audience-specific topics you can build into your campaigns, but without the cost.

Think of it as a lightweight market sensing approach. It will not replace a full platform, but it will improve your content planning and help you stay connected to the conversations happening in your market right now.

Try it yourself

You can run this in ChatGPT in under five minutes.

1. Define your audience clearly — include role, location and industry.

2. Ask: “What is [this audience] asking you about [topic] right now?”.

3. Use the ideas to shape your next round of blog posts, LinkedIn updates or campaign content.
   

If you've got time, try running the same exercise with a generic prompt for comparison. The difference will be obvious.

Of course, prompting is only part of the equation. There's also tone and style to consider, which is a whole other can of worms I covered in this article, Why Every B2B Marketer Needs a Great Style Prompt in their AI Toolkit. If that resonates, make sure you take a look at my Write Like a CCO Style Prompt Card — it's the style prompting card we use at Only Good Content to make sure AI outputs sound like a senior, experienced human is behind the wheel every time.