It’s time for your business to get on board with AI prompting. Here’s how to begin.

• Generative AI tools are revolutionizing small business operations and output. 
• Leveraging generative AI tools requires users to craft effective AI prompts — which is both an art and a science. 
• With the right approach to AI prompting, businesses can drive operational efficiency.

Effective prompting — or writing questions and commands to generate responses from AI systems — is critical to getting better results from generative AI.

But like any skill, practice makes progress when it comes to AI prompting. Users who continuously experiment with and refine their prompts can expect higher quality outputs from generative AI. 

For small business owners in particular, this can lead to new levels of creativity and efficiency, whether it’s enhancing their ability to produce engaging visuals, compelling marketing campaigns, or data storytelling. 

Here’s a guide to help small businesses craft better AI prompts and maximize value.

AI prompting basics

Writing an AI prompt is different than asking a simple question or conducting an online search query. An effective prompt has several key elements: goal, context, expectations, and source. For example, if a user wanted to generate meeting talking points from an AI solution like Microsoft Copilot, it might break down like this: 

“When it comes to prompts, order matters,” Antony Cook, corporate vice president and deputy general counsel at Microsoft, said. “The order of your instructions can impact the response you get. Later parts of a prompt are likely to be emphasized more than earlier parts. Experiment with different orders to understand the impact.”

It’s also important to focus on telling the system what to do, rather than what not to do. Using “if-then” instructions within a specific prompt can lead to better output. For example, “Create a customer support email template. If a customer is upset about a delayed shipment, then include language that offers a 10% discount off their next purchase.” Users should also stick to a single request in each prompt and avoid entering overly general requests or using business jargon without additional clarification.

Refining AI prompts involves some experimentation. Testing different orders for requests or using different wording can lead to various outcomes. Users can then evaluate the results to determine what works best for their needs. Solutions like Copilot give small business owners the opportunity to experiment and learn with generative AI while benefiting their business. 

Prompts to power your business 

Business owners can apply generative AI tools in many different ways. The technology can automate routine tasks and streamline workflows, analyze data and deliver insights more quickly, and reduce potential data processing errors. Together, this boosts productivity, reduces risk, and improves operational performance. Some common scenarios where businesses might use AI prompts include:

Data management

Example prompt: “Copilot, create a PowerPoint presentation based on the sales data from Q1, including charts and key insights.”

Enhanced communications

Example prompt: “Copilot, design a marketing brochure using data from the attached report. Begin by pulling statistics from the chart on page 4. Then, use the product descriptions on page 6 for the ‘What We Offer’ section. Finally, incorporate the product images and icons from pages 6 and 7 into the brochure’s front cover design.”  

Customer engagement

Example prompt: “Copilot, create a series of social media posts announcing our upcoming holiday sale, including engaging captions and relevant hashtags.”

Content Creation

Example prompt: “Copilot, generate a compelling product description for a new line of organic skincare products.”

Financial management

Example prompt: “Copilot, categorize and summarize the business expenses from the last month for our financial review.”

Businesses can also use prompts to draft sales pitches, generate leads, conduct budget planning activities, craft email marketing campaigns, and more. However, for all the opportunities generative AI offers, people must use the technology carefully and responsibly. Businesses can put checks and balances in place and focus on generative AI solutions with built-in guardrails to avoid inadvertently infringing on copyrights and to ensure safe, effective use of these tools.

Navigating the boundaries of generative AI 

Every technology has its limitations, and users need to keep that in mind as they apply generative AI tools in their everyday work. For example, Copilot’s responses are generated based on patterns in data and language models, which means they can sometimes include inaccuracies or errors. 

“It is crucial to always review and verify AI-generated content before using it,” Cook said. “Always cross-check the AI-generated content with reliable sources to ensure accuracy. This can involve verifying data points, facts, and figures manually.”

Alternatively, Copilot is built on a foundation of trust. Microsoft Azure OpenAI services securely process prompts, so any organizational or company data a user enters is always protected, never accessible to other commercial customers, and is not used to improve Microsoft’s AI models, products, and services.   

For added protection, users should focus on proper data categorization and properly tag internal documents to ensure sensitive information isn’t inadvertently included in the output of AI tools. 

Copilot, like other products across the Microsoft 365 suite, also incorporates responsible AI principles. The design and deployment of Microsoft’s AI technologies anchor on six core principles, including fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability. Microsoft continuously updates and trains its AI models to enhance accuracy and relevance. 

Just like the printing press, desktop computers, and smartphones before it, generative AI will transform how people execute everyday tasks. In business, the technology can optimize everything from communications to strategic planning. However, generative AI is most effective when there’s a human in the loop, especially when it comes to AI prompting and validating system outputs. Businesses that want to get the most value out of generative AI should implement internal checks and balances at every stage of the AI content lifecycle — and use tools that do the same. 

Learn more about Microsoft’s Copyright Commitment here.

This post was created by Insider Studios with Microsoft.