Working AI-first. So how do you make that practical for your team?
B2B marketing is in a state of flux. AI tools are changing not only how we write, but also how we structure, optimize and publish. Customers no longer just orient themselves via Google, but via ChatGPT, Perplexity and Claude. And as a marketing team you want to stay ahead .
At Blinqx, we’ve embraced AI as a marketing team. Everyone works with ChatGPT Pro, which is where we build our own Custom GPTs – smart digital assistants that make work more fun, faster and most importantly, more consistent. Every month we share the latest ‘finds’ as well as our fixes (because no, not every GPT is immediately brilliant).
I wrote earlier: AI-first is primarily a culture of experimentation and sharing what you learn. And why not share that more widely? Therefore: the five Custom GPTs we don’t want to live without anymore, with immediately applicable examples. Not to do everything faster, but above all: better.
1. Structure Spark GPT – thought leadership content
We all know it: that blank screen remains tricky. That’s why we created a Custom GPT to get from rough notes to clear structures for thought leadership content. The GPT here is not an editor-in-chief, but a smart colleague who stimulates your creativity and asks questions.
What does it provide? Within a minute you have a blog post, opinion piece or white paper with a logical flow and the right headings. Thanks to the built-in control questions, you automatically get a clear picture of what you do and do not know.
Applied in: first draft for blogs (this one), opinion articles and white papers, columns.
Tip: Give your GPT lots of examples in the knowledge section: blogs you’re proud of, podcasts, quotes, and a clear tone-of-voice. The richer your input, the better the output.
2. Feature-to-Value GPT – from product innovation to customer value
New feature? Nice. But what’s in it for the customer?
Our Custom GPT translates technical innovations directly into recognizable scenarios for the right target audience. Result: concrete situations where customers see the benefit immediately – ideal for product pages, sales decks and campaigns.
Then the GPT asks, “What do you want to create with this?” For example, you choose, “Webinar concept + invite + landing page.
What does it provide? Within seconds there is an outline saving some 70% of the time previously spent in this process.
Applied in: product marketing, launch campaigns, sales enablement.
3. Case Builder GPT – write customer cases faster.
Writing a good customer case is a recurring activity that takes quite some time. Perfect process to optimize. With a Custom GPT that builds a customer story based on an interview transcript. According to the right tone of voice, with the customer’s pain points as a clear starting point, in the template developed for that purpose.
What does it provide? The output is not a cut-paste final version (never do it anyway), but rather a solid first draft with headings, quotes and structure – allowing you to do in one hour what would otherwise take all afternoon.
Applied in: translation of customer interviews into publishable case.
4. Newsletter Optimizer GPT – better and faster setup and outline
Every product marketer regularly creates newsletters where multiple content pieces come together. The collective challenge: quickly move from existing content to core message with good intros and call-to-actions.
To do this, we use the Newsletter Optimizer: a GPT that creates a brief summary per URL (of blog, case or white paper) in newsletter style, with curiosity-inducing hook, benefit and CTA.
What does it provide? It does not produce indiscriminately, but creates A/B variants that you choose from, before taking the next step. Stay sharper yourself!
Applied in: monthly Blinqx newsletter and multi-content campaigns
5. Findability Checker GPT – search optimization tips
You simply enter a URL and get a report back within a minute: SEO score (technique still relevant for AI reviews), quality, Tone of voice, audience relevance, additional suggestions. The output includes suggestions like, “add a concrete scenario here” or “search term is missing from this sub-headline.”
What does it provide? No more manual checks, but an AI audit with clear improvement suggestions.
Applied in: review of all existing content and new content prior to publication.
What’s next?
6. LLM Discovery Optimizer – findable in the AI era
We are now testing the LLM Discovery Optimizer – a Custom GPT that analyzes existing content for factors important to AI tools.
The concept: paste the URL or copy you want to analyze into the GPT and get a content analysis + checklist of recommended action items (think snippetability, timeliness, E-E-A-T signals). This is done based on a comprehensive checklist that we have compiled ourselves in advance.
7. Tone of voice Checker – practical guide for non-marketers
Because AI makes content writing accessible to everyone, we as marketing don’t have the final say, we become more directive. Therefore, we create the tone of voice checker that every piece of content can go through. This setup can then be customized for each company in the Blinqx family to continue to recognize the nuances by label.
The concept: paste your text into the GPT and get tips and suggestions on how to use the right tone, what to better leave out and why X is better than Y for consistency.
Getting started with Custom GPTs yourself? What we’ve learned so far:
- First, inventory which core processes you want to optimize (what takes a lot of time, is repetitive, error-prone, etc)
- Build your own Custom GPT prompt library for your core processes (and share them with your colleagues)
- Ask for help! Don’t know how to write a good Custom GPT? Look around you for people who are further along. We asked Floor de Wit and Bouke Vlierhuis for their expertise.
- Evaluate the output of your GPTs with your colleagues and peers: what helps, what doesn’t? And keep doing this to continually optimize.
- Make it fun: experiment, share your finds and keep learning
Check out the latest Blinqx developments around AI here.
Frequently asked questions about GPTs in marketing
Where do I start if I want to build my own GPT for my marketing team?
Start with the processes that take a lot of time or often linger. Think of customer cases, blog structures or translations of product features to customer benefits. What you repeat often, you can usually smartly support with a Custom GPT. At Blinqx, we started small: one prompt, one use case, one experiment.
How do I prevent the output of a GPT from becoming generic or flat?
That’s in the preparation. Our GPTs all have a rich knowledge section: examples of content we’re proud of, an elaborate tone-of-voice, typical customer questions, and context about the product or market. The better the input, the sharper the suggestions. This remains people work.
How do you get your marketers to work with AI without splintering everything?
Work with templates, shareable prompt libraries and establish tone-of-voice guidelines together. You can customize each GPT by target audience or label, but the basic principles remain the same. And: have colleagues review each other’s output – AI makes it faster, but recognition comes from people.
What if a GPT doesn’t work?
Ask for help! In our team meetings, we make it a point to discuss non-working prompts or weird outcomes. Sometimes it’s a matter of bad input. Sometimes it’s faulty structure. And sometimes a task is just not suitable for automation. Then we delete it from the GPT library.
