The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
The 80-10-10 rule is a workflow heuristic: spend the first 10% defining what a good result means, let AI help with a suitable middle portion, then use the final 10% to check and refine its work. It is not a research-validated ratio or a guarantee of time saved. The practical principle is to set a clear target, delegate only work you can review, and keep responsibility for the final result.
What the 80-10-10 rule means
Thanh Pham, founder of Asian Efficiency, describes the rule as three parts of a task: the first 10% is deciding what good looks like, the middle 80% is work AI may be able to help with, and the final 10% is human review and refinement. The percentages are a way to frame a workflow, not measured divisions that every task should follow. Pham’s article was last updated September 4, 2026: The 80-10-10 Rule.
The sequence matters more than the numbers. If you ask AI to work before deciding what the result should accomplish, you may get fluent output that does not solve the right problem. If you skip the final check, an apparently finished draft can still contain errors, omit context, or use the wrong tone.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Start by defining what done looks like
Before delegating, write down the outcome, who it is for, and the criteria the finished work must meet. Keep the definition specific enough that you can tell whether the AI’s result is usable. For a client meeting brief, for example, you might specify the client and meeting purpose, the background and open questions to include, and ask the AI to flag decisions that need human judgment.
#1 Best Overall
If you cannot yet state the requirements, ask the AI what context it needs before asking it to draft. Then supply or verify that context yourself. This first stage is not busywork: it gives both you and the tool a clear target.
Choose work AI can help with and you can check
The middle portion is a better candidate for AI when the work is repetitive or time-consuming, has a structure you can explain, and can be reviewed without rebuilding it from scratch. The test is not whether AI can produce something; it is whether its contribution moves the task forward and is practical to verify.
- Turn meeting notes into proposed follow-up tasks for you to confirm.
- Find recurring themes in customer comments, then check that the themes reflect the original feedback.
- Draft an agenda from a stated meeting goal and list of topics.
- Organize research into sections, leaving factual verification to you.
Delegation may add effort when the task is quicker to do manually, depends heavily on personal preference or judgment, or needs to sound unmistakably like your own words. In those cases, AI can still be useful for a narrow supporting step, but handing over the whole task may create more checking and correction than it saves.
Recommended Free Tools
Review the result before relying on it
Human review is not just proofreading. Check whether the output is true, whether it missed context, whether it has the right voice, and whether it is useful for the intended outcome rather than merely polished. You remain responsible for deciding whether it is ready to send, share, or act on.
- Accuracy: Verify claims, names, dates, figures, and proposed actions against reliable source material.
- Completeness: Look for omitted constraints, caveats, or important details from the input.
- Context and voice: Make sure the answer fits the audience, situation, and tone you intended.
- Usefulness: Ask whether the result advances the task, not merely whether it looks finished.
Try the rule on one recurring task
- Pick a repeatable task. Choose one such as preparing for a meeting, writing a project update, reviewing customer feedback, or planning a week.
- Write a one-sentence target. State the intended outcome and a concrete definition of done.
- Delegate a reviewable part. Give the AI the relevant context and ask it to perform a bounded, structured portion of the work.
- Check and refine. Verify the result using the accuracy, completeness, context, voice, and usefulness checks above before using it.
- Assess the outcome. Consider whether the task moved forward, how much correction it needed, and whether you trusted the result enough to use it. Judge the result, not the number of prompts or AI-generated steps.
Thanh Pham’s related episode, “The 80-10-10 Rule: How to Actually Delegate to AI,” is dated April 29, 2026: The Productivity Show. The framework is presented as practical guidance; the article and episode do not establish an independent productivity effect or a fixed amount of time saved.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
