Seventeen ChatGPT prompts in this AI Central guide share one design choice: not one of them asks the model what makes your product different. Every prompt is a bracket-slot template that makes you supply your own product facts and your competitors' facts first, then puts ChatGPT to work as an analysis engine over that input. That is the technique, and it is why the outputs read like positioning work rather than your own marketing copy handed back.
Reviewed August 7, 2026. The source is a first-party AI Central document in the library.
The problem these prompts are built around
Ask any chatbot what makes your offer unique and it will answer. That is the trap. The model holds no facts about your business beyond what is in the conversation, so it fills the gap with the most ordinary language available. You get adjectives, and nothing you can put in front of a buyer comparing three vendors on the same afternoon.
The document opens by stating that standing out from the competition is essential in business, and offers the seventeen as a set for differentiating your offer. Read the prompts themselves and a stricter idea shows up underneath that framing.
One pattern runs through all seventeen
Every prompt is a fill-in template, and the placeholder vocabulary repeats: your product description, your features, competitor features, a named major competitor, a specific industry, client feedback, a potential client role. Sixteen of the seventeen carry a slot for your own product. Eleven name a competitor inside the prompt itself.
Method note on those counts. I read each numbered prompt and checked for a bracketed placeholder referring to the reader's own offer, and separately for the word competitor anywhere in the prompt. The one without a your-product slot is the client feedback prompt. The six that never mention a competitor are numbered one, four, five, six, seven and sixteen.
That ratio is the argument the guide makes without saying it out loud. Differentiation is a comparison, and a comparison needs two sets of facts. These prompts cannot run on one.
The wording is the real instruction
Look at how the second prompt, labelled Research, is phrased. It does not ask for differences. It asks for two grades of difference.
Taking into account the nuances of [YOUR PRODUCT/SERVICE FEATURES] and those of [COMPETITOR PRODUCT/SERVICE FEATURES], delineate not only the overt differences but also the subtle distinctions that might influence specific market segments.
That is prompt two in the AI Central guide, verbatim. The phrase about overt differences and subtle distinctions rules out the obvious answer, which is the feature table both companies already publish. The first prompt does the same job pointed inward, asking ChatGPT to extrapolate potential secondary and tertiary benefits not directly mentioned.
This is the difference between a prompt and a wish. A wish asks for a good answer. These specify what would make the answer bad, then rule it out.
Prompt five is the one that changes the mode
Sixteen of the prompts ask ChatGPT to produce something: a matrix, a sales narrative, a positioning map, a list of market gaps. One asks it to attack instead.
Your role is a skeptical [POTENTIAL_CLIENT_ROLE, e.g., 'CTO of a major firm'] who's seen numerous pitches and is looking for deep insights.
That is the second half of prompt five, the role-play, as AI Central wrote it. It is the only point in the set where the model is not on your side, and it is the cheapest quality control in the document. Run it against the sales narrative from prompt four and you learn which claims collapse under one question. The ones that survive are worth putting on a pricing page.
It is a sequence, not a menu
The order in the document is not decorative. Read straight through, it is a workflow that starts inside your own business and moves outward in four movements.
- Prompts one to three build the raw material: your unstated benefits, the competitor comparison, then a matrix positioning every player against market needs and potential niches.
- Prompts four to six make it usable: a layered sales narrative, a skeptical stress test of it, and a pass over real client feedback.
- Prompts seven to thirteen produce the artifacts: a refresh after market shifts, a sharpened value proposition, market gaps, case study framing, and a positioning map.
- Prompts fourteen to seventeen handle defense: niche markets, a competitor's cited weaknesses, barriers against new entrants, and emerging technology as a differentiator.
Run the positioning map prompt alone and you get a map built on whatever the model assumes. Run the first three before it, in the same thread, and you get a map built on your inputs. Same prompt, different answer, and the difference is entirely upstream.
How to run the set properly
Five things decide whether this works, all of them about what you bring rather than what you paste.
- Work in one continuous ChatGPT thread, so context from the early prompts carries into the later ones.
- Paste a real product description into the first prompt, the full one, not the tagline.
- Pull competitor features from their own pricing pages and paste them in, rather than letting the model recall them.
- Keep the matrix from the third prompt. It is the artifact the rest of the set feeds off.
- Re-run the seventh prompt whenever your industry moves, it is the only one written for refresh rather than first setup.
The honest limits
Two things the guide does not do, worth saying plainly.
First, it offers no outcome data. No case study, no before and after, no named company, no conversion figure anywhere in the document. The word easy in the title is a claim about effort, not results. Judge it on whether the outputs are work you would act on.
Second, three prompts lean on knowledge the model may not have. The ninth positions you against a named major competitor, the thirteenth asks it to consider competitors' known attributes, and the fifteenth goes furthest.
Considering the commonly cited weaknesses or complaints about [A MAJOR COMPETITOR], how can [YOUR PRODUCT/SERVICE] position itself as a solution to those pain points?
That is prompt fifteen in the AI Central set, verbatim. Nothing in it supplies the complaints, so ChatGPT will supply them, and a model asked for commonly cited weaknesses will produce plausible ones whether or not they are real. Paste in actual review text or support tickets before running those three. That is the line between working and sounding like it worked.
The line the guide closes on
The last page is a short instruction rather than a pitch. It tells you to drop a prompt, polish the output, and not wait for inspiration. Then it separates the prompts from the thing that uses them.
This guide is your engine.
That is AI Central's own framing, and it is the right one. Seventeen prompts pasted into a blank chat give you seventeen answers. The same seventeen run in order, with real product detail in the brackets and a skeptical buyer on the other side of prompt five, give you a position you can defend.
Do I have to fill in every bracket?
Yes, and it is the step people skip. The brackets are the only place your facts enter the conversation. A prompt run with the placeholder left in, or with a tagline standing in for a real description, produces an answer about a generic product in a generic market.
How long does the full set take?
The document gives no time estimate and I will not invent one. The slow part is not the prompting, it is gathering the competitor features and the client feedback the prompts ask for. Collect those first and the seventeen runs go quickly in one sitting.
What if I do not have my competitors' feature lists yet?
Six of the seventeen never mention a competitor, so start there: the first, fourth, fifth, sixth, seventh and sixteenth. All six run on inputs you already own. Hold the comparison prompts until you have real competitor detail to paste in.
Do these only work in ChatGPT?
The guide is written for ChatGPT and names no other tool. The prompts are plain instructions with fill-in slots and no product-specific syntax, so structurally nothing ties them to one model. That is an observation about how they are written, not a claim the document makes.
If I only run one prompt, which one?
The third, the differentiation chart. It is the only prompt that asks for a matrix positioning you and two named competitors against market needs and potential niches, so it forces the comparison and leaves you an artifact you can keep.