SMEs' First AI Scenario: Writing Proposals, Weekly Reports, Customer Service Scripts

2026-06-03·8 min read·OKCodex Editorial

SMEs' First AI Scenario: Writing Proposals, Weekly Reports, Customer Service Scripts

AI Practical Handbook Part 3


Start with what produces results most easily

Many bosses say “we need to use AI,” but employees don't know where to start. This article gives three immediately effective scenarios that require no technical background and can be used today.


Scenario 1: Writing a promotion proposal

Traditional approach: The marketing department spends 2-3 days writing a draft proposal. AI approach: Get a first draft in 10 minutes; manual polishing takes 30 minutes.

Prompt template:

You are a senior marketing strategist, skilled in the [industry] field.

Background:

  • Company: [company introduction]
  • Product: [product name and core features]
  • Target customers: [customer profile]
  • Promotion budget: [budget range]
  • Promotion period: [time period]

Please create a promotion proposal, including:

  1. Goals and KPI setting
  2. 3 main promotion channels and execution steps
  3. Content creation direction
  4. Risk contingency plan

Output in Markdown format, total word count within 800 words.

Real case: An education institution used this template to generate a draft summer enrollment proposal in 5 minutes. Teachers only needed to fill in specific data, reducing total time from 2 days to 2 hours.


Scenario 2: Writing weekly/monthly reports

Traditional approach: Employees spend 30-60 minutes each week organizing weekly reports, with uneven quality. AI approach: Record work highlights → AI organizes them into a standardized format, done in 5 minutes.

Prompt template:

Please organize the following work records into a professional weekly report format:

[Paste your work log, for example:

  • Monday: Followed up with customer A, confirmed contract details
  • Tuesday: Attended product review meeting, made 3 suggestions
  • Wednesday: Handled customer B complaint, coordinated technical support
  • Sales completed this week: XX ten-thousand ]

Weekly report format requirements:

  • Completed items this week (sorted by importance)
  • Key data/results
  • Plan for next week
  • Items requiring support

Concise and professional language, no more than 400 words.

Suggestion: Have the team spend 3 minutes each day recording work highlights (can use phone notes), then generate the weekly report with one click on Friday.


Scenario 3: Writing customer service scripts

Traditional approach: Training customer service takes 1-2 weeks; scripts rely on oral transmission, and quality is unstable. AI approach: Quickly generate a standard script library and unify service quality.

Prompt template:

You are a professional customer service trainer, with a friendly and professional style.

Product information:

  • Product: [product name]
  • Core selling points: [1-3]
  • Price: [price range]
  • Common questions: [list the 5 most common questions]

Please write 2 reply versions for each question:

  1. Friendly colloquial version (suitable for 微信/小红书)
  2. Professional formal version (suitable for email/business scenarios)

Keep each version within 100 words.

Advanced use: Paste the customer's actual chat history to AI and have it analyze “which replies led to churn,” then continuously optimize the scripts.


Execution suggestions

  1. Pick one scenario to try first, don't launch all three at once, or it's easy to give up
  2. Create a “Prompt library” document, save effective prompts, and share them across the team
  3. AI produces the first draft, humans verify: AI can make mistakes, especially with specific numbers and facts, so manual review is required

好码未来 · AI Practical Handbook Part 3 | Next: Using AI to analyze Excel data

Share:Weibo

Subscribe to weekly briefing

Comments

Loading comments…

Leave a comment

0 / 500

Comments appear after review.

Find tools for your next task

Start with scanned documents, code review or research. Get a shortlist with matching evidence; no model is run.

Find tools for my task →
Analyze this topic with two models

DeepSeek and Qwen are preselected with the article topic. You choose when to run them.

Try it →

Keep reading

Explore

Explore published agents