Read on: My website
Read time: 2 Minutes
Everyone wants the smartest model.
And a model that speaks to them.
Very few people…want to admit that the biggest problem is the sentence they just typed.
And you spend significant time, energy, effort, and burn tokens just to try to get the answer you are looking for.
Was this newsletter forwarded to you?
You can spend hours comparing benchmarks, context windows, and reasoning scores. You can upgrade from one release to the next and feel like you are “keeping up.”
If the way you ask for help is vague, lazy, or half‑baked, your results will not get meaningfully better. They will just look more confident on the surface.
Better models make better guesses. They do not read your mind.
So…what can you do about it? We are going to show you how to fix it today.
Let’s dance.
Your competitor just replied. You’re still typing.
A lead comes in on Instagram. Another on Messenger. Three more on SMS.
Your team switches tabs, repeats answers, and loses context while hot leads wait hours for replies. At 2am, nobody responds at all.
That’s not a people problem. It’s a process problem.
Wati brings Instagram DM, Facebook Messenger, TikTok, WhatsApp, SMS, RCS, and web chat into one AI-powered inbox. Automations instantly respond, qualify leads, and route conversations to the right person, 24/7.
Your team stops firefighting. Your leads stop waiting. Your pipeline starts moving.
Variables
Most AI conversations sound the same.
“Which model is best right now.”
“Did you see the new release.”
“Look at this side‑by‑side screenshot, this one clearly wins.”
If the output looks off, the reflex answer is simple: blame the model.
Assume you need a more advanced version. Subscribe to a new tier (spend more money). Wait for the next big announcement.
You are cruising…I mean burning…through your tokens faster than ever.
But if you actually use these tools every day, you start to notice something different.
The same model that produced a generic, disappointing answer in one conversation can generate a sharp, useful result in another. Nothing changed in the code. The only thing that changed was the way you framed the task.
Researchers are starting to quantify that. Early studies show that the quality of user prompts can influence generative AI performance as much as model improvements themselves.
In some experiments, roughly half of the gains came from upgraded models, and half came from better prompts layered on top.
So yes, model choice matters. But if you ignore the input side of the equation, you leave most of the upside on the table.
If you want to check out some of my other recent articles, check them out here:
You Do Not Have One AI Problem, You Have Two
When Loyalty Becomes a Liability: The Hidden Risk in Long-Term Clients
Real Cost
A bad prompt does not just waste a response, burn tokens, and frustrate you. It can distort your work.
If you are using AI to draft a blog post or do research, you might just get something bland. Annoying, but not catastrophic. If you are using AI to:
-
Summarize research for a strategic memo
-
Draft talking points for a board update
-
Outline risks in a policy decision
-
Help you compare vendors or scenarios
the stakes go up fast. Real fast.
A vague prompt that produces a confident but wrong summary does not just waste five minutes. It pushes you toward a bad decision faster. It creates false clarity.
That is the real risk for leaders, analysts, and knowledge workers who are leaning into AI: not that the model “fails,” but that a sloppy prompt produces a polished answer you trust too much.
When that happens, the problem is not the AI. It is the way you briefed it.
Bad
Most people equate “bad prompt” with “short prompt.” That’s not completely true, but close.
A bad prompt is any prompt that forces the AI to guess about the things that matter most.
Guess is the most critical word of the last sentence.
The patterns show up over and over:
-
You do not state the goal. You ask it to “write something,” “analyze this,” or “help with strategy” without saying what decision or outcome this will inform.
-
You do not name the audience. You forget to mention who will read it, what they know, or what they care about.
-
You skip the constraints. Length, tone, depth, examples, level of formality, all left to chance.
-
You never define the format. You do not say whether you want a brief, a list of options, a table, or a draft.
On the model’s side, the language is fine. It understands the words. What it does not understand is your intent. So it does what any system would do: it fills in the blanks with generic assumptions.
You ask, “Write about AI and infrastructure,” and expect something tailored to senior leaders in your industry. The model has no way to know that. So it gives you something aimed at everyone, which means it resonates with no one.
Prompt design guides say this in more formal language: clarity, specificity, and context vastly improve output quality. Short or long is less important than whether the prompt tells the model enough about the job it is actually doing.
The problem usually is not that AI cannot follow instructions. It is that it never received real instructions in the first place.
Raise the Bar
Do not believe“As models get smarter, prompting will matter less.”
Reality is moving in the opposite direction.
As models improve, people give them more important work:
-
Drafting contracts and proposals
-
Summarizing complex reports
-
Producing investor updates
-
Outlining strategic options
The outputs look cleaner. The language is polished. The answers sound confident. All of that makes it easier to trust what you see on the screen.
That is exactly why prompting becomes more important, not less.
When you give a mediocre model a bad prompt, you usually see the cracks. The answer feels off, so you discount it. When you give a strong model a bad prompt, it can produce something that looks right enough to lull you into using it without thinking.
If you care about the decisions built on top of AI output, you cannot shrug and say, “The model should just know.” You have to treat the way you ask as part of your professional responsibility.
Some organizations already see this. Business schools and enterprise guides are starting to treat prompt design as a core skill, emphasizing that user instructions directly impact accuracy, relevance, and clarity. Prompting is not a party trick. It is the new version of writing a good brief.
How
You do not need a 50‑page prompt library to get better results. You need a simple habit.
Before you hit submit on an important AI request, quickly check three things:
-
Goal
-
What decision, document, or outcome will this answer inform.
-
Are you asking for exploration, recommendation, or production.
-
-
Context
-
Who is the audience.
-
What do they already know.
-
What constraints matter (industry, geography, risk level).
-
-
Format
-
Do you want an outline, a comparison table, a list of options, or a full draft.
-
How long is useful, not just “as long as possible.”
-
Here is a simple before/after to make it less abstract.
Weak:
“Write a summary about the data center market in Virginia and North Carolina”
Stronger:
You are helping a VP of Sales for a Mid-Atlantic design-build firm develop a go-to-market strategy for entering and expanding in the data center market across Virginia and North Carolina.
Write a one-page strategic brief that outlines a 3-year implementation plan for breaking into and growing within this market.
Assume the reader is experienced, commercially minded, and deeply familiar with construction, business development, and regional relationship-building, but wants a clear strategic framework rather than generic market commentary.
The brief should focus on:
How to approach the Virginia and North Carolina data center markets differently based on regional dynamics.
How to build relationships and meeting cadences with key interested parties across state, local, federal, and commercial sectors.
Which stakeholders matter most early, and how engagement should evolve over a 3-year timeline.
Risks, barriers to entry, and political or regulatory friction points that could slow market penetration.
How to sequence business development, partnerships, policy engagement, and account targeting in a practical way.
Use clear, direct language and short paragraphs.
Avoid generic statements about “market opportunity” or “AI benefits.” Focus on actionable go-to-market strategy, stakeholder mapping, and execution priorities for a design-build firm trying to establish real presence and credibility in the data center sector.
If helpful, structure the response in these sections:
Market Entry Objective
Virginia vs. North Carolina Strategy
Priority Stakeholders
3-Year Implementation Plan, Mapping Plan Points every 6 Months
Key Risks and Watchouts
Competitors in our Space
Same model. Completely different guidance. The second version reduces guesswork. It gives the AI a role, an audience, a format, and a focus.
If the task is complex, break it into stages instead of throwing everything into a single monster prompt:
-
Ask for an outline first.
-
Then have it fill in one section.
-
Then ask it to critique or tighten that section.
-
Then move on.
You are not “bothering” the model. You are giving it a workflow.
Quick
If you want something you can tape next to your screen, make it this:
Before you send an important prompt, ask:
-
What am I trying to accomplish.
-
Who is this ultimately for.
-
What does “good” look like in terms of length, tone, and format.
If you cannot answer those questions clearly, that is your work to do before you ask the AI to do its work.
The way your organization prompts AI is now part of how it thinks.
Bad prompting is the new bad briefing. It leads to shallow thinking, sloppy decisions, and work that looks polished but is built on sand. Good prompting forces clarity: on goals, on audience, on constraints, on what matters.
You will keep seeing new model announcements. You will keep feeling a little FOMO when a new benchmark drops. That is fine. Upgrade when it makes sense.
But if you want to see real improvement in the work your teams are doing with AI, do not start by switching tools. Start by changing the way you ask.
The most powerful upgrade most people could make this week is not a new model.
It is their next sentence.
See you next week.
Whenever You’re Ready, Here are 4 Ways I Can Help You:
-
Unlocking Hidden Potential – Reconnecting with Past Clients for Explosive Growth – Check out my free eBook on how you can find hidden gems in your past clients and help you crush your sales goals.
-
AI for Business Development – Download our free eBook on how you can effectively leverage AI prompts to your advantage. From properly setting up your preferred AI tool, to how to shape your prompts, save time, and get the outputs you are looking for.
-
Sales Resources at Your Fingertips – From tools, tips, demos, and how-tos, check out our Pages and content that can provide you with additional support, whether it be social selling, account management, or something else.
-
Cribworks Advisor Program – Want more than just resources? Reach out to me and see if our Advisor Program can help you scale your business.
