eduKate Yishun · Learning and the next step forward
Find a useful starting point
Choose the question closest to your situation, or read the full guide in order.
Browse all chapters · Explore recovery and transformation · Explore the learning library
Clear work instructions help someone complete a task without having to guess what the writer meant. A useful step-by-step procedure explains the starting conditions, the actions in order, the decisions along the way and the evidence that the result is acceptable. Writing work instructions therefore means describing a complete route through the work, including the points where the reader must pause.
A work instruction template can help organise that route, but a template cannot supply missing knowledge. If “use the latest version” leaves the reader choosing between two plausible documents, the instruction is still incomplete. If “check the finished pack” never explains what belongs in the pack, the reader can follow every sentence and produce the wrong result.
This guide explains how to write clear instructions for another person, from choosing a bounded task to trying the draft with an unfamiliar reader. It includes an original worked example, a reusable structure, a different practice task and a reasoned answer. The central test is practical: can the intended reader identify the right starting point, perform the task and recognise when the result should be accepted, corrected or held?
Your 50-second route: Start with the task contract if your instructions have become a long collection of reminders. Go to actions and names if readers keep asking what a sentence means. Use decisions and exceptions when ordinary steps work but variations cause trouble. Study the worked case and its revised procedure for a complete example. Finish with the reader tryout and independent practice before relying on your own draft.
Contents
Chapters 1–5 · Begin with a task contract
Chapters 6–10 · Make decisions and exceptions explicit
Chapters 11–15 · Repair the cause of confusion
A task contract is a short statement of what the instruction helps a reader produce, from which starting point, under which conditions. It need not be a formal document. “Prepare one exhibition envelope from an approved request and the matching caption cards” is already more useful than “Exhibition administration.” The first statement has a unit of work and an endpoint. The second could cover almost anything.
Name the intended reader. A person who understands the task but forgets occasional details needs a different aid from a person doing it for the first time. The unfamiliar reader may not know the local names for trays, who can clarify a request or what a finished example should contain. Stating that audience prevents the writer from silently relying on knowledge that the reader does not have.
Then name the result in observable terms. “Prepare it properly” describes an aspiration. “An envelope labelled with the exhibit identifier, containing the required current caption cards and one matching placement slip” describes something a reader can inspect. The description also makes it possible to discover disagreements before someone spends time producing the wrong thing.
Set the boundary at a useful point. Preparing an envelope may end when it is placed in the ready tray and its status is recorded. Delivering it, positioning the exhibit and opening the exhibition are separate tasks. Including every downstream activity can bury the procedure the reader actually needs. Omitting a necessary status record, however, can leave the next person unable to distinguish finished work from unfinished work.
Write one sentence identifying when these instructions do not apply. Perhaps they cover ordinary single-exhibit envelopes but not requests with unapproved wording or combined exhibits. That boundary gives a reader permission to stop using the procedure when its assumptions fail. It also prevents the writer from disguising an unresolved decision as a complicated exception.
This guide sits within Business Turnaround and Rebuilding, where reliable everyday work is one part of operational recovery. Begin with Recovery and Transformation when you need to decide which kind of problem deserves attention first.
Before writing polished sentences, collect one ordinary example from start to finish. Put the starting request beside the finished result. Ask which facts connect them. Which identifier makes the result belong to this request? Which information determines the required contents? Which reference resolves disagreements? Which action changes the status from prepared to ready?
Watch for decisions hidden inside expert movement. An experienced person may pick one card from a stack almost instantly. The choice could depend on language, revision, exhibit identifier and approval status. Writing “select the card” removes the very knowledge that makes the action correct. Describe the selection conditions, not just the visible movement of the person’s hand.
Separate what the task requires from what one experienced person happens to prefer. A particular pen colour might be personal habit. Recording the request identifier might be essential for tracing the result. If both are written as equally compulsory, the reader cannot tell which differences would invalidate the work. Ask what would go wrong if a particular detail changed, and use the answer to decide whether the detail belongs in the instruction.
Collect a plausible exception as well as the ordinary example. Missing material, conflicting references and partly completed work often expose gaps that a perfect demonstration hides. Do not manufacture a rare puzzle merely to make the document elaborate. Choose a variation that the intended reader could reasonably meet and that affects what they should do next.
If two knowledgeable people describe different acceptable methods, resolve the result requirements before choosing a method to teach. They may be using different assumptions, or they may have two genuinely acceptable routes. A beginner usually benefits from one supported route with a clear boundary. Additional routes can be documented separately when readers have a reason to choose between them.
Do not begin by shortening an old document. First discover which decisions it currently answers and which it leaves to memory. A shorter instruction that removes a necessary selection rule is less useful. A longer instruction that records every conversation is not automatically better. The aim is enough information to act correctly, placed where that information becomes necessary.
A prerequisite is something that must be true before the task can proceed as described. It may be a material, a reference, permission to perform the task, a skill or an agreed decision. If a reader discovers a missing prerequisite halfway through assembly, they may have to undo work or leave a confusing partial result for someone else.
List the actual inputs and how the reader identifies them. “Have the necessary documents” is too vague. “Use the request sheet marked Exhibition Batch M, revision 3, with the approval box completed” is testable within a fictional task that uses that reference. In a real setting, the document title and approval method must match what the reader can actually find.
Distinguish a prerequisite from an action that belongs inside the procedure. The existence of an approved request is a prerequisite if the reader has no authority to approve it. Matching the request identifier to an envelope is an action the reader performs. Mixing the two can accidentally tell someone to create an approval or make a decision outside the task’s scope.
Explain what to do if a prerequisite is absent. Give the location for held work, the information to record and the person or role that can resolve the missing input. “Ask someone” transfers the ambiguity to a conversation. “Record the request identifier and the missing approval, then ask the request coordinator before preparing that envelope” gives the reader a usable next move.
Do not assume that seeing a material means it is suitable. A stack may contain current and superseded cards. A folder may hold two similar request sheets. The prerequisite check should identify the reference against which suitability is judged. When references conflict, the reader needs a way to obtain clarification, not an instruction to choose whichever looks newer.
Keep preparation proportionate. A small paper task might need three inputs and two labelled trays. It does not need a ceremonial equipment inventory. The useful test is whether each prerequisite prevents a specific interruption, mismatch or unauthorised assumption. Remove entries that do no work, and add the missing condition when a reader repeatedly reaches the same avoidable dead end.
A clear step tells the reader what to do to which object. “Update it and move it over” contains two actions but leaves both objects and the destination uncertain. “Write the exhibit identifier on the envelope, then place the envelope in the prepared tray” makes those relationships visible. Whether those two actions deserve separate numbered steps depends on how the task is organised.
Use the same name for the same object throughout. If the document starts with “placement slip,” switching to “location card” halfway through can make a new reader look for a second item. If two objects are different, give them different names. A request sheet records what is wanted; a status sheet records what happened. Calling both “the form” destroys a useful distinction.
Avoid words that imply a standard without identifying it. “Correct,” “appropriate,” “complete” and “latest” can all be useful after the standard has been defined. Before that, they hide decisions. Replace “use the correct revision” with a rule that points to the revision recorded on the approved request. If that reference is missing, explain the hold action instead of encouraging inference.
Make the action visible enough to recognise. “Be aware of language requirements” does not produce an observable result. “Read the language field on the request and select the matching caption arrangement” does. The second sentence still needs the arrangement defined elsewhere, but it gives the reader a concrete operation and a reason to consult that definition.
Choose the amount of detail for the reader. Someone familiar with envelopes does not need instructions for lifting a flap. They may need an explanation of why the exhibit identifier must remain visible after packing. Useful detail addresses likely uncertainty. Excess detail about obvious movements makes the important selection rules harder to find.
Test pronouns by removing the previous sentence. If “it” or “this” now has several plausible meanings, repeating the noun may be worthwhile. Test location words by imagining that the page is printed or displayed differently. “The box on the right” may cease to identify the destination; “the box labelled Ready” continues to work if that label is present in the actual workspace.
Steps belong in the order that their dependencies require. A reader must identify the request before selecting materials for it. They must know the requested language before deciding how many caption cards belong in the envelope. They must inspect the assembled contents before marking the envelope ready. These are logical dependencies, not merely a writer’s preferred sequence.
Write the major state changes first: request identified, materials matched, envelope assembled, contents checked, status recorded. Then place the actions that produce each change beneath it. This makes omissions easier to see. If the outline jumps from materials matched to status recorded, assembly and acceptance may be missing. If the same selection happens twice, the document may be duplicating work.
Use a numbered sequence when order matters. Use a short reference table when the reader needs to choose between defined alternatives. Do not hide an ordered procedure inside a paragraph that demands several movements and a decision before the reader can look back at the page. Equally, do not turn every descriptive sentence into another numbered step.
Watch for circular instructions. “Use the completed status sheet to select the required contents, then complete the status sheet from the contents” gives the reader no independent starting reference. The request should determine the required contents; inspection should determine the recorded status. Each information source needs a clear role so that an error is not simply copied into the next record.
Consider interruption points. If a reader stops halfway through, what visible state tells them where to resume? A partly assembled envelope placed among ready envelopes invites a mistaken assumption. An identified work area or an explicit in-progress status can preserve the distinction. Choose a method that fits the task rather than adding records that nobody can maintain.
A procedure can be logically ordered yet awkward to use. After checking dependencies, try performing the task while reading it. Notice repeated trips between references or unnecessary handling of the same item. Improve the order only where prerequisites remain satisfied. Saving a movement is not helpful if it encourages the reader to record acceptance before the evidence exists.
A decision point needs a condition, an action for each relevant outcome and a way to rejoin or leave the procedure. “If necessary, add another card” supplies none of those clearly. The reader must decide what necessary means, which card to add and whether the task is complete afterwards. A short decision table can carry that logic more clearly.
For an exhibition task, an agreed rule might be: an English request requires one English caption card; a bilingual request requires one English card and one named second-language card. Both arrangements must match the request’s exhibit identifier and specified revisions. A missing or unrecognised language field places the request on hold. The writer has now defined the ordinary routes and the unresolved route.
Avoid writing branches that overlap. “If the pack is bilingual, add the translation” and “if the pack needs two cards, add the second card” may refer to the same action. A reader could add twice. Define the condition once and say exactly which set of items it requires. When two conditions can both apply, explain whether both actions are needed or one rule takes priority.
An exception instruction should preserve useful information. Instead of merely “stop,” state which unit of work stops, how it is identified, what remains unfinished and where the uncertainty is recorded. Often one unresolved request can be held while other independent requests continue. That is different from a missing master reference, which may make the whole batch impossible to interpret.
Do not invent an answer to make the procedure look complete. If the reader lacks authority to choose wording, substitute a material or resolve conflicting approvals, the correct branch is clarification. A complete instruction can end with a clearly described hold state. Completion of the instruction does not require pretending that every requested result can be produced immediately.
This is where work instructions meet repairing a broken handoff. A reader needs usable inputs and someone who can resolve gaps. The handoff guide explains that transfer. The instruction itself explains how to use the received inputs to perform the bounded task.
A reader can perform every listed movement and still produce an unacceptable result if the reference was wrong or a selection rule was misunderstood. Acceptance therefore concerns the resulting item, checked against the requirement. It should not be inferred merely because the steps were attempted, the allocated time passed or the envelope looks tidy.
Define the checks in observable terms. For an envelope, that might mean the identifier matches the request, the required language set is present, every caption revision matches the reference and the placement slip gives the requested location. “Looks good” cannot distinguish an attractive but obsolete caption from a current one. The criteria must concern the features that make the result usable.
Include the record and destination where they are part of completion. A correctly assembled envelope left in an unidentified pile may not be available to the next person. A status sheet marked ready while the envelope remains incomplete creates a different problem. Acceptance should connect the physical result, its recorded state and its agreed location without claiming that later delivery has already happened.
Describe repair and rechecking. If an incorrect card is replaced, recheck the relevant contents before changing the status to ready. Do not preserve the original pass mark after changing what was checked. If the repair introduces another dependency, such as a changed language arrangement, revisit all acceptance criteria affected by that change.
Be careful with totals. Eight requests do not automatically mean eight ready envelopes. A batch can end with six ready and two held, with all eight accounted for. The held items are not failed paperwork to hide; they are requests whose requirements have not yet been met. Distinguishing them keeps a neat total from concealing an unusable result.
Once a reader knows how to perform the task, a short aid may help them remember the final checks. Building a checklist that prevents repeated mistakes explains that separate job. The checklist refers to defined requirements. It does not replace the instruction that teaches how to select, assemble, repair and recognise exceptions in the first place.
Morrow Exhibition Room, its people, requests, quantities and outcomes are fictional. The original case concerns preparing paper materials for a small exhibition. It illustrates instruction writing; it is not evidence of a measured improvement in a real organisation.
A coordinator has approved eight requests, R01 to R08. Each request needs one labelled envelope, the specified caption card or cards, and one placement slip. An English arrangement requires one English caption card. A bilingual arrangement requires one English caption card and one second-language caption card. The approved request identifies the second language where required and specifies the revision of each card.
The old instruction says: “Get the latest captions. Put two sheets in each envelope. Check they are correct. Tick done and leave them for collection.” An experienced volunteer may mentally replace those sentences with the actual rules. An unfamiliar reader cannot know whether “two sheets” includes the placement slip, whether every envelope needs two caption cards or how “latest” is determined.
The fictional batch has these requirements and starting materials:
| Request | Caption arrangement | Material position before assembly |
|---|---|---|
| R01 | English | Required current card available |
| R02 | Bilingual | Both required current cards available |
| R03 | English | Required card missing |
| R04 | Bilingual | Both required current cards available |
| R05 | English | Required current card available |
| R06 | Bilingual | Current English card available; only a superseded second-language card available |
| R07 | English | Required current card available |
| R08 | English | Required current card available |
All eight approved requests, eight blank envelopes and eight correct placement slips are available. There are no opening envelopes already in progress and no further requests arrive during this task. The work is ordinary paper assembly; no printing, wording approval or translation is part of the instruction. The person preparing envelopes cannot substitute a revision or create missing text.
Five English requests require five caption cards. Three bilingual requests require six. The full batch therefore requires eleven current caption cards. Nine suitable current cards are available: eight for the six fully supplied requests, plus the English card for R06. One additional physical card is superseded and unsuitable. Two required current cards are missing, one for R03 and the second-language card for R06.
The old instruction provides no defensible route through this situation. “Latest” could tempt the reader to use the newest card they can find even when it does not match the approved reference. “Two sheets” could produce an incorrect arrangement. “Tick done” could conceal the two unresolved requests. The rewrite must clarify those decisions before it can usefully shorten anything.
The revised instruction begins with its scope: prepare individual Morrow exhibition envelopes from the approved Batch M request sheet, revision 3. It applies to the English and bilingual arrangements listed there. It ends with accepted envelopes in the Ready tray and unresolved requests recorded in the Hold tray. Collection and installation are separate tasks.
The reader needs the approved request sheet, the caption stock, matching placement slips, blank envelopes, the batch status sheet and trays labelled Ready and Hold. The request coordinator resolves missing or conflicting requirements. If the approved request sheet itself is absent or contradictory, the reader stops the batch and asks the coordinator to resolve the reference before selecting any cards.
The instruction defines a held request as an identified request whose required result cannot currently be accepted. It defines ready as assembled, checked, recorded and placed in the Ready tray. Those words describe actual states in this example. They do not mean that a collector has accepted delivery or that an exhibit has been installed.
The preparation sequence is:
- Select one unprocessed request from the approved sheet. Write its request identifier on a blank envelope and locate the placement slip with the same identifier.
- Read that request’s caption arrangement, required language or languages, card revisions and placement location. If any of those fields are missing or conflict, use the hold action in step 4.
- Find the required caption set. For English, select one English card. For bilingual, select one English card and one card in the named second language. Match each card’s request identifier, language and revision to the approved request.
- If a required card or matching placement slip is missing, unsuitable or uncertain, stop work on this request. Put its labelled envelope and any suitable matched materials in the Hold tray. Record Hold and the precise missing or conflicting item on the status sheet. Refer the question to the request coordinator. Return to step 1 for another request.
- When all required materials match, place the selected caption card or cards and the matching placement slip inside the labelled envelope. Keep one request’s materials together; do not combine different request identifiers.
- Check the assembled envelope against the approved request. Confirm its identifier, the complete language set, each caption revision and the placement location on the slip. Confirm that no extra caption card is included.
- If a check fails and the correct material is available, replace the incorrect material and repeat step 6. If it is unavailable or uncertain, use the hold action in step 4. Do not mark the envelope ready while correcting it.
- After every acceptance check passes, record Ready for that request and place the envelope in the Ready tray. Return to step 1 until every request has a recorded state.
- Compare the status sheet with the Ready and Hold trays by request identifier. Each of the eight requests must appear once in the account. Investigate an omitted identifier, duplicate entry or mismatch before reporting the batch status.
- Report the ready identifiers and held identifiers with their unresolved requirements. Leave the held materials identified. When the coordinator later supplies a resolved requirement, return that request to the material-matching work in step 3 and repeat the acceptance check before changing its state.
The return in step 4 applies to an individual request with a local exception. It does not overrule the opening instruction to stop the whole batch when its common reference is unreliable. That distinction matters because a request-specific missing card and an unusable master sheet affect different amounts of work.
The envelope remains a container, not proof of acceptance. Placing a current English card inside R06’s envelope is useful preparation, but the missing current second-language card keeps R06 on hold. Similarly, a superseded card is not transformed into an acceptable card by putting it in the correct envelope. The revision rule continues to govern the decision.
With the stated materials, R01, R02, R04, R05, R07 and R08 can become ready. R03 and R06 remain held. The final account is six ready plus two held, equalling eight requests. The six ready envelopes contain eight accepted caption cards and six placement slips. R06 holds its one suitable caption card and placement slip; R03 holds its placement slip. The unsuitable superseded card remains outside the ready and held envelopes, separate from the selected materials.
No assembly speed or productivity gain is claimed. The example establishes that the written logic gives the reader an answer for every stated condition. Whether the draft communicates that logic successfully still needs a tryout with the intended reader.
Choose a reader who reasonably resembles the intended audience and has the prerequisites the document assumes. If the instruction is meant for an unfamiliar volunteer, asking the expert who helped write it provides limited evidence. That expert may silently repair gaps. The aim is to discover whether the document carries the needed information without those private repairs.
Use representative materials and explain that the instruction is being tried, not that the reader is being examined. Ask the reader to use the draft and identify places where they cannot confidently proceed. Avoid immediately explaining every hesitation. A useful pause may reveal an unfamiliar term, a missing selection rule or a dependency placed too late.
Record where help was needed and what kind. “Needed assistance” is a weak observation. “At step 3, could identify the language but could not find the revision marker” points to a fix. It may require describing where the marker appears, changing the material label or correcting a reference. It does not automatically call for another paragraph of general encouragement.
Separate three questions: did the reader perform the described action, did the document describe the right action, and did the result meet the requirement? A reader who follows a faulty instruction precisely exposes a document defect. A reader who misunderstands a correct but ambiguous sentence exposes a communication problem. A defective input may require clarification outside the document itself.
In the Morrow case, a useful tryout would include an ordinary English request, a bilingual request and a request with an unsuitable revision. Those examples examine distinct branches. Watching one fully supplied English envelope succeed does not tell the writer whether the hold rule is usable. Keep the exercise small enough that the observer can reconstruct what happened.
If the observer gives a missing rule aloud, record that assistance. The resulting success cannot be counted as independent use of the original draft. Revise the relevant passage and try it again with a reader or example that does not merely repeat the explanation. For broader decisions about changing a live workflow, use testing a small operational change.
Do not respond to every difficulty by adding a warning at the end. If the reader chooses the wrong card because two items have the same name, repair the naming or the selection rule. If they discover a missing prerequisite after assembly, move the check before assembly. If they cannot interpret a decision table, inspect the conditions and outcomes before increasing its visual emphasis.
A useful revision note states the observed problem and the change intended to address it. “Clarified instructions” is difficult to evaluate later. “Reader treated newest available as approved; replaced latest with the revision recorded on the approved request” preserves the reasoning. Another reviewer can then inspect whether the changed wording actually resolves the ambiguity.
Keep explanations close to the decisions they support. A short reason can help a reader resist a tempting shortcut: for example, a revision must match the approved request because a newer-looking card may belong to another approved arrangement. A long history of how the rule developed is unlikely to help someone selecting a card at the worktable.
Check the rest of the document after a local edit. Changing the name of a tray in one step but leaving the old name in the exception section creates a new problem. Changing acceptance from “two cards” to “the required language set” must also update examples and any linked checklist. A small sentence change can alter several references.
Distinguish an instruction problem from a missing decision. If nobody has agreed which revision should be used, clearer wording cannot create that agreement. Record the unresolved requirement and obtain the necessary decision from the appropriate owner. Writing an authoritative sentence before that happens merely makes the uncertainty harder to see.
The useful movement is from a specific observation to a specific change and then to another check. Translating feedback into change develops that wider reasoning. Here, its practical expression is a draft whose revisions can be tied to reader difficulty rather than to a writer’s preference for more polished language.
A good instruction loses usefulness when the reader cannot tell whether it applies to the materials in front of them. Give it a descriptive task title, a version or revision identifier, its scope and an owner for questions. Those details need not dominate the page. Their job is to connect the document to the task and make a disagreement resolvable.
Keep one clearly identified current copy in the place readers actually use. If a printed sheet remains beside the worktable after the reference changes, the existence of a newer file elsewhere may not help. When replacing a working copy, check the points of use. Retained older copies should be visibly distinguishable from the version intended for current work.
Describe a substantive change in terms of what the reader must do differently. “Version updated” says little. “Bilingual requests now require separate English and named second-language cards; check both revision fields” communicates the changed action. If the change affects work already started, say how those items should be reviewed before they are accepted.
Review when something relevant changes: the requested result, starting materials, names, authority to resolve exceptions or acceptance criteria. A repeated question can also signal that the instruction no longer fits its audience. A calendar reminder may prompt inspection, but the passage of a date by itself does not tell the owner which parts need revision.
Keep the maintenance burden realistic. A short task does not need a large document register to be understandable. It does need someone to notice when the instruction and the actual task diverge. Assigning that responsibility prevents everyone from assuming that another person will update the wording.
The broader Rebuild guide concerns creating a system that can last. For work instructions, durability means that changes in the task reach the reader’s working copy and that evidence of confusion reaches someone able to repair it. An untouched document is not necessarily a stable process.
The fictional Orchard Media Shelf prepares resource folders for a community discussion group. This is a different task from exhibition envelopes. You are writing instructions for a new helper who must select existing materials, assemble folders and record their status. The helper cannot approve titles, change requested formats or create missing materials.
There are four approved requests, D01 to D04, no opening work and no arrivals during the task. Every folder requires a cover showing its request identifier and one matching discussion sheet. A print-only request also requires one printed reading. An audio-and-print request requires the printed reading plus one listening card showing the approved audio title and its version. A listening card is required even if the printed reading has the same title.
The approved request list names the required format and material versions. D01 is print-only, D02 is audio-and-print, D03 is audio-and-print and D04 is print-only. All covers, discussion sheets and required printed readings are available. D02’s listening card matches the approved title and version. D03 has a listening card with the same title but an older version. No current listening card for D03 is available.
The old instruction says: “Put the reading and discussion sheet in each folder. Add the audio if there is one. Mark complete when the folder has three things. Put finished folders on the shelf.” The workspace has a Ready shelf and a Hold tray. A request coordinator can resolve missing or conflicting materials. The status record has fields for request identifier, state and unresolved requirement.
Before reading the answer, identify four separate defects in that instruction. Then write a short selection rule for the two formats, an exception rule for unsuitable material and an acceptance rule. Finally, determine which requests can be ready with the stated materials, which must be held and what the final account should show.
Do not assume that “three things” counts objects consistently. The cover is a required part of the folder, but the old instruction does not say whether it is included in the count. More fundamentally, the two requested formats have different content requirements. Your revised rule must identify the required items, not rely on a number that accidentally fits one arrangement.
One further question makes the exercise harder. Suppose the helper notices the older version on D03 but has already written Complete in the record. What exactly must change before they stop work on that request? Consider the physical location, the recorded state and the information the coordinator needs. A useful answer preserves the partly prepared work without allowing it to masquerade as ready.
The old instruction does not tie material selection to the approved request. “Add the audio if there is one” makes availability determine the output even when audio is required but missing. “Three things” does not define which components count or accommodate the different formats. “Finished” has no acceptance test, and the instruction gives no route for an unsuitable version or an incorrect status entry.
A better selection rule is: read the requested format and material versions for the identified request. For print-only, assemble the matching cover, discussion sheet and printed reading. For audio-and-print, assemble those same components plus the listening card matching the approved audio title and version. Verify each item against the request; a title match alone does not establish a version match.
The exception rule should say that missing, conflicting or unsuitable required materials place that request on hold. Keep its suitable materials identified in the Hold tray, record the unresolved item and ask the coordinator to resolve it. Continue with other independently interpretable requests. If the approved request list itself is unreliable, stop the affected scope of work until the common reference is clarified.
Acceptance requires the right identifier, every component required by the format and the specified versions, with no unresolved mismatch. Record Ready only after those checks pass, then place the folder on the Ready shelf. A folder awaiting a listening card cannot pass merely because the remaining components are correct. Likewise, a numerical count does not prove that the components are the required ones.
D01, D02 and D04 can be ready. D03 must be held because its required listening card is unsuitable. The final account is three ready plus one held, equalling four requests. The ready folders contain three covers, three discussion sheets, three printed readings and one suitable listening card. D03’s suitable cover, discussion sheet and printed reading remain identified in Hold; its older listening card is kept separate from the selected materials.
If Complete was already recorded for D03, correct the current state to Hold and record that the required current listening card is missing. Move the identified partial folder to the Hold tray. Preserve the correction in whatever ordinary record method the group uses so that a reader does not continue relying on the earlier state. The request becomes ready only after a suitable card arrives and the acceptance check is repeated.
For your own task, use this compact structure: task and intended reader; start and finish; prerequisites and authoritative reference; named materials; ordered actions; decisions and exceptions; acceptance and repair; final state and destination; current version and question owner. Each part answers a different reader uncertainty. Leave out a field only when its function is already clear, not because it makes the template shorter.
Start with one task that repeatedly needs explanation. Write its required result, identify the starting reference and capture the choices an experienced person makes. Draft the ordinary route and one plausible exception. Then put the draft beside the real materials and let an appropriate reader use it. The first useful achievement is discovering where the instruction still depends on the writer being present.
Do not judge the result only by how professional the page looks. Ask whether a reader can choose the right input, make the required distinction, recognise a hold condition and verify the final result. A plain page that answers those questions is useful. Attractive formatting can support the same work, but it cannot resolve an undefined requirement or a contradictory branch.
Use the companion guides according to the remaining problem. If work reaches the reader without usable requirements, examine the handoff. If a trained reader understands the task but regularly omits a final check, consider a short checklist. If the task itself appears to create delay or rework, return to the operational recovery hub to investigate the wider process.
Two primary references inform the narrow writing conventions cited here. Microsoft’s style guidance recommends numbered procedures, separate instructions and clear action locations. Microsoft: Writing step-by-step instructions.
Google’s procedure guidance distinguishes optional steps and explains how to present actions with their results. Google developer documentation style guide: Procedures.
The fictional materials, selection rules, exercises and acceptance decisions in this guide are original teaching examples. The references do not establish that these examples improve productivity. Your instruction earns practical confidence when the intended reader can use it on appropriate tasks, with uncertainties made visible and the resulting work checked against its actual requirements.
Contents · Previous chapter · Return to recovery and transformation
Continue through this learning web: Branch hub · Build a Preventive Checklist · Diagnose a Process Bottleneck · Recovery & Transformation · Library