Blog

Even if you’ve been using ChatGPT for a while, you probably make mistakes. Lots of people to do. Falling into common prompt traps, generating content that sounds too ChatGPT, or ignoring the essential components of the perfect prompt; even the smartest entrepreneurs don’t get it right every time. Get back to basics with prompting with these five simple tips for maximum productivity in minimum time.

Research paper Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning by Large Language Models was created by a number of scholars from Singapore Management University and others. In addition to specific quick additions that will create super tips, the document shares five simple tips that no one should ignore.

Applying these tips will help large-scale language models (LLMs) like ChatGPT, Claude, and Gemini handle multi-step reasoning tasks more efficiently so that entrepreneurs can use them more effectively.

ChatGPT Tips for Better Prompt Results: 5 Science-Backed Methods

Customize prompts for different tasks

Just as you would instruct different members of your team in different ways depending on the task at hand, you should customize your prompts to suit different tasks by adjusting the level of detail and specific instructions. You wouldn’t sign someone up on Loom to discuss a critical and urgent business decision. You wouldn’t call a meeting for a simple accounting matter. Apply the same thinking to the ChatGPT prompt.

You don’t need a superprompt for every request. Adjust the prompt to fit the specific business task. Don’t make it complicated when you don’t have to. If you want ChatGPT to critique your landing page, for example, include best practices, explain your target audience, ask them to accept a role, and specify the feedback you want. On the other hand, for a simple “find a synonym” task, a single line instruction may be sufficient.

Include real-world examples

ChatGPT does not live in the real world. It processes over 300 billion words in its training data to make predictions, and sometimes it’s wrong. Give your enthusiastic AI colleague a hand. Use real-world business scenarios in your prompts to make answers more relevant and practical. Help him understand what you are trying to achieve.

Use real-world business examples to illustrate the problem so ChatGPT gets the vibe you’re going for, whatever you ask for. Include square brackets with the need-to-know context, such as “my business offers [explain your offer] for [explain your audience] and I face challenges including [describe them].” Get closer to the perfect answer the first time.

Break down the tasks

You wouldn’t give a human team member a bunch of rambling instructions for their week without any context or sense of priority. You wouldn’t want them to do it all at once. Apply the same thinking to ChatGPT. Instead of sending a huge prompt with many sections and complicated instructions, go step by step. Prompt for the first result, then take the result and ask for more.

Breaking down complex tasks into smaller stages, each with their own instructions and goals, helps ChatGPT manage them better. For example, when asking for a blog post, ask for title options first before choosing the best one and asking for an abstract. Enter your summary edits before asking for section titles, refining those before asking each section to be written. This is better than asking for a blog post all at once and means you have more control over the results.

Combine the efforts of people and ChatGPT

Contrary to what the AI ​​hype will tell you, ChatGPT cannot solve all your problems. At least not yet. Hallucinating, exaggerating. It comes out with some random stuff that is not customer ready and would hurt your business more than help. Don’t forget the human touch. Use ChatGPT for the initial steps and have people review and confirm the final results. This combines the best of both worlds.

The paper recommends generating an initial solution and having it reviewed by a human expert. Don’t go from prompt to publish without this step. If you publish ChatGPT-generated content without proper verification, it will harm your brand. Understand the telltale signs, go over each word carefully, and only include phrases you would actually say. Fact-checking, gut-checking, and maintaining your integrity.

Keep improving

The document acknowledges that ChatGPT is constantly evolving. Even between model numbers, results will be very different within a few weeks. Keep learning as the platform progresses. Read the studies, be hypercritical of the results, and continually refine your prompts based on feedback and effectiveness.

Get the best results from ChatGPT with the same approach you take to business. What once worked like a charm may now be way off the mark. Experiment, learn, iterate, improve. Even try the same prompts with alternate LLMs. Don’t settle for your tried and tested prompts; review them regularly.

Back to basics with ChatGPT: quick improvements for better results

By following these simple tips, entrepreneurs can get better results from ChatGPT, especially for complex tasks. Customize prompts depending on the desired outcome, include real-world examples, break down tasks, combine machines and humans, and keep improving. Simple steps that help ChatGPT understand and solve problems more accurately, for smarter solutions and better business results. Stop wasting time with suboptimal prompts. Develop this skill and keep improving. You will need it.