侠Liangshen AIailiangshen.com
AI Implementation Workshop · Short topic lessons · Lesson 5

New tools are trending again. Should you jump on the bandwagon?

Use three questions to decide whether to switch

Freeabout 9 minNo experience neededWeb lesson + slides + video

This skills module focuses on a practical exercise: using three questions to decide whether to adopt a trending new tool, and designing a one-week pilot test. We will complete this in nine minutes.

By the end of this lesson, you can

  • Use three questions to determine whether to test a new tool
  • Design a one-week comparison between old and new methods for a specific, recurring task
  • Define success criteria and obtain feedback from a colleague who has tried it
  • Write a conclusion: adopt, reject, or use only in specific scenarios

Takeaway: a one-week comparison table, one success criterion, and one conclusion

01

Introduction

About 1 minutes

Let's look at a scenarioFictional example; numbers are illustrative

Zhou Kai, Sales Operations

His social media feed is filled with weekly posts claiming, 'This new tool is amazing.'

He has already spent two evenings testing it.

Yet, he cannot clearly articulate what problem it actually solved for him.

Chasing new tools does not mean you can effectively use them.
02

Methodology

About 3 minutes

Three questions, plus one red line

Question 1

Can it do the most time-consuming task for me? If you can't answer which task that is, don't try it yet.

Second question

Can my current tools already handle 80% of the job? The cost of switching is often not worth it.

Third question

Does company policy allow it? Where will the data go? If this point isn't cleared, there's no need to discuss anything else.

An additional question for tech companies

Will it have access to source code, API keys, unreleased products, or customer network data?

Tip

Use prompts to ask four questions

Suggested time
Four questions
I'm considering trying a new tool: [Tool Name, for example only].
The most time-consuming task I do is: [Task name and time per instance].
The tool I currently use is: [Current tool].
Please help me answer four questions:
1. Theoretically, can this tool perform this task? List the features that need verification; do not make a definitive judgment for me.
2. Where can my current tool already perform this task? What costs are incurred by switching?
3. If I want to try it, what data and usage policies do I need to confirm with the company first?
4. Data red lines: Could this tool potentially access source code, API keys, unreleased products, or customer network data? Who should I confirm this with?
It should list the features you need to verify on the official website, rather than making a definitive judgment for you.

One-week mini-experiment

Define standards firstFor example: Average time for the same task decreases by 30%, and error rates do not increase
→
Try both old and new methods three times eachFor the same fixed task, perform it three times using the old method and three times using the new tool
→
Record six linesRecord the time taken, number of errors, and whether you are willing to continue using it tomorrow
→
Ask a colleague to try it onceAsk them what the differences are, record their exact words, and do not answer for them
Only switch if all three metrics are better than the old method.
03

Practice

About 3 minutes
Tip

Example of a record (fictional, numbers are illustrative)

Suggested time
Six-line record table
| Attempt | Method | Time (minutes) | Number of errors | Willing to continue tomorrow? |
| 1 | Old method | 18 | 0 | Yes |
| 2 | Old method | 17 | 1 | Yes |
| 3 | Old method | 19 | 0 | Yes |
| 4 | New tool | 15 | 1 | Hesitant |
| 5 | New tool | 14 | 2 | Hesitant |
| 6 | New tool | 15 | 1 | No |
Illustrative conclusion: On average, about three minutes are saved, but errors increase by one on average, and verification takes about five extra minutes. Therefore, do not switch for now.

Common Biases and Remedies

5 Minutes: Preparing the Starting Point for an Experiment

Achieved: ______
04

Quiz and Wrap-up

About 2 minutes

Three quick questions to check yourself

Question 1: Why should an experiment use the old method and the new tool three times each, and why write the success criteria first?

Answer B.Comparing the same task ensures comparability; writing criteria first prevents your post-experiment fondness from skewing the results.
Show the full transcript (video narration)
New tools are trending again. Should you jump on the bandwagon?

This skills module focuses on a practical exercise: using three questions to decide whether to adopt a trending new tool, and designing a one-week pilot test. We will complete this in nine minutes.

By the end of this lesson, you will be able to:

By the end of this lesson, you will be able to do four things. First, use three questions to determine whether to test a new tool. Second, design a one-week comparison between old and new methods for a specific, recurring task. Third, define success criteria and obtain feedback from a colleague who has tried it. Fourth, write a conclusion: adopt, reject, or use only in specific scenarios.

Let's look at a scenario

Let's look at a fictional scenario; numbers are illustrative. Zhou Kai, in Sales Operations, sees weekly posts in his social media feed claiming 'This new tool is amazing.' He has already spent two evenings testing it, yet he cannot clearly articulate what problem it actually solved for him. Chasing new tools does not mean you can effectively use them.

Three questions, plus one red line

Focus on just three questions. First: Can it do the most time-consuming task for me? Second: Can my current tools already handle 80% of the job? Third: Does company policy allow it, and where will the data go? If this point isn't cleared, there's no need to discuss anything else. For tech companies, add a fourth question: Will it have access to source code, API keys, unreleased products, or customer network data?

Use prompts to ask four questions

This prompt ensures it lists the features you need to verify, rather than making a definitive judgment, because its knowledge may be outdated. It will also list the switching costs and tell you who to confirm regarding data red lines. Note: Specific features, pricing, and data policies are subject to the official website as you see it now. This course content was verified in November 2026; please double-check before use.

One-week mini-experiment

The one-week mini-experiment has four steps. Step 1: Define success criteria with numbers. Step 2: For the same fixed task, perform it three times using the old method and three times using the new tool. Step 3: Record the time taken, number of errors, and whether you are willing to continue using it tomorrow for each attempt. Step 4: Ask a colleague to try it once, ask them what the differences are, and record their exact words. Only switch if all three metrics are better than the old method.

Example of a record (fictional, numbers are illustrative)

This is a fictional six-row record, with numbers for illustration. The old method was used three times, and the new tool three times. Calculate the average time and average errors, then look at the last column. In this example, the new tool saves about three minutes on average, but causes one extra error on average, and verification takes about five minutes, so we will not switch for now.

Common Biases and Remedies

Several common biases. It directly praises the new tool, so ask it to list only the points you need to verify on the official website. The experiment did not record data, so set up the table first before starting. The new tool is faster but more error-prone, so calculate the error count and verification time together. Always wanting to try more, limit yourself to testing only one tool per week. A colleague says 'It's pretty good' but provides no follow-up, so ask what specifically is lacking and record a concrete quote.

5 Minutes: Preparing the Starting Point for an Experiment

Now, spend five minutes preparing the starting point for an experiment. Write down a new tool you want to try, the task it will perform for you, and the time it takes each time. Using the prompt from earlier, copy down the points that need verification. Write a success criterion that includes a number. Draw a six-row empty table, and invite a colleague to test it next week. Finally, clarify whether the company approves, who to ask, and when to ask.

Quiz

Here is a quiz question. Why should an experiment use the old method and the new tool three times each, and why write the success criteria first? A, To make the experiment look more formal. B, Comparing the same task ensures comparable time and error counts; writing criteria first prevents bias from your fondness for the tool. C, The new tool can only be tested three times, and cannot be used afterward. The answer is B.

Don't Chase, Don't Fall Behind

Final reminder: Not following every trend doesn’t mean you aren’t learning. Set aside one hour each month to review new developments, note promising tools in a ‘to-try’ list, and test them in small experiments when your schedule allows. Tools’ features, pricing, and data policies change frequently. This course content was verified in November 2026; always check the official website and your current version before use. Thank you for watching.