DA1: Thinking in workflows

Why each option is right or wrong

• “Redesigning the signboard, since design software can be scripted” (Incorrect): It happens twice a year and needs creative judgement, so automation wouldn't pay back.
• “Both equally, because any task done by hand can be automated” (Incorrect): Rare or judgement-heavy tasks usually cost more to automate than they save.
• “Neither, because a small shop has too little work to automate” (Incorrect): Forty minutes a day is over 4 hours a week; small businesses often gain the most.
• “Copying the website orders into the spreadsheet” is correct.

  Copying orders happens every day, follows the same steps each time, and the information already arrives digitally. Those three signs together make it a strong candidate. The signboard is rare and creative, so automating it would cost more than it saves.

  • Test: frequent, rule-based, digital input.
  • Bonus sign: the same details are typed into more than one place.

📖 Revise: Lesson 1: Spotting repetitive work

Why each option is right or wrong

• “Leave it out, since the brief only asked for reminders” (Incorrect): Dropping a step people rely on would quietly change how the clinic works.
• “Ask the receptionist to stop adding notes, to keep things simple” (Incorrect): Changing the clinic's practice is their decision, and the notes may matter for revenue.
• “Automate the reminders now and add the note in a later version” (Incorrect): Launching without it would change what patients receive; agree on it first.
• “Map it as a decision and confirm the rule with the clinic” is correct.

  Mapping the process as it really runs is meant to uncover hidden steps like this one. Automating only the "official" version would drop the payment note. Put it on the map as a decision, check the exact rule with the person in charge, and build it in or agree to drop it on purpose.

  • Lesson: map real work, including workarounds.
  • Confirm: rules come from the business, not from your guess.

📖 Revise: Lesson 2: Mapping a process step by step

Why each option is right or wrong

• “Check the email every minute instead of every five minutes” (Incorrect): Checking more often doesn't help if the text it reads keeps changing shape.
• “Ask staff to forward each order email in a fixed format” (Incorrect): That brings back manual work, and forwarding can introduce new errors.
• “Use a larger AI model to read the email more cleverly” (Incorrect): An AI model may cope better, but structured data from the source is more dependable and cheaper.
• “Get the order from the shop as a webhook with named fields” is correct.

  Reading details out of email text depends on the exact wording and layout, so any template change breaks it. A webhook from the website sends the order as structured data with named fields, such as phone and amount, which stay stable when the email design changes.

  • Check: many shop platforms offer order webhooks.
  • Fallback: if not, an export or API is still sturdier than email text.

📖 Revise: Lesson 3: Triggers, actions and the data between them

Why each option is right or wrong

• “Send penalty notices automatically on day 1 to keep things fair” (Incorrect): Automatic penalties without review risk upsetting residents with genuine reasons.
• “Don't automate anything, as money matters must stay manual” (Incorrect): Friendly reminders are low-risk and save time; only the penalty needs judgement.
• “Let an AI decide which residents deserve a penalty” (Incorrect): Penalties affect people's money and relationships, so a person should decide.
• “Automate reminders, and draft penalties for approval” is correct.

  Reminders are routine and low-risk, so they can run automatically. Penalties affect money and relationships, and some cases deserve an exception, so the workflow should prepare a draft and let the secretary decide. This keeps the time saving without removing judgement.

  • Pattern: assistive automation, where the workflow drafts and a person approves.
  • Add: a list of approved exceptions the workflow can check.

📖 Revise: Lesson 4: When not to automate

Why each option is right or wrong

• “About 6 months, since automation projects are slow to pay back” (Incorrect): The numbers here show a much faster payback.
• “Immediately, because automation has no ongoing cost” (Incorrect): Building took 6 hours and upkeep continues, so there is a real cost to recover.
• “It doesn't pay back, because upkeep cancels the saving” (Incorrect): Upkeep of about 15 minutes a week is far smaller than the 90 minutes saved.
• “About 4 to 5 weeks, once the monthly upkeep is allowed for” is correct.

  Ninety minutes a week is 1.5 hours, and 6 ÷ 1.5 gives 4 weeks. Upkeep of about an hour a month takes away roughly 15 minutes a week, so the real saving is nearer 1.25 hours, which gives about 5 weeks. Either way, it pays back within about a month or so.

  • Formula: build hours ÷ weekly hours saved.
  • Honest version: subtract upkeep from the weekly saving first.

📖 Revise: Lesson 5: Measuring time saved

Why each option is right or wrong

• “Ask every customer to type the venue's phone number on the form” (Incorrect): Customers may not know it and would introduce errors.
• “Send the message to the events company's own number instead” (Incorrect): That doesn't reach the venue, which was the point.
• “Skip the message whenever the phone number is missing” (Incorrect): The form doesn't collect it, so no message would be sent.
• “Look up the number in a venues list” is correct.

  The workflow has the venue name but needs a phone number. A small "venues" sheet or table, kept by the company, lets a lookup step find the right number by name. Listing each step's fields before building is what reveals gaps like this one.

  • Keep it tidy: one venues list, kept up to date by one owner.
  • Handle misses: alert a person if a venue isn't found.

📖 Revise: Lesson 3: Triggers, actions and the data between them

Why each option is right or wrong

• “Agree and switch the automation off” (Incorrect): The log shows real savings; switching it off would bring the 20 hours back.
• “Build more automations until staff are no longer busy” (Incorrect): The issue is how the value was shown, not the number of workflows.
• “Tell the manager that automations can't be measured” (Incorrect): They can: run logs and task diaries give clear numbers.
• “Show the hours saved and what the team now spends that time on” is correct.

  Busy staff don't prove an automation failed. Comparing the run log with the old task diary shows 20 hours of invoicing moved off people's desks, and the team is spending that time on other work, such as customer calls. Presenting both facts lets the manager judge fairly.

  • Measure: before and after, in hours, errors and delays.
  • Keep records: the task diary is your baseline.

📖 Revise: Lesson 5: Measuring time saved

Why each option is right or wrong

• “With test marks, because paper sheets cause the most errors” (Incorrect): Paper input has to be digitised before it can be automated.
• “With both at once, to show the biggest saving quickly” (Incorrect): Starting with the task that's ready gives a quicker, safer win.
• “With neither, until all records are moved to a new app” (Incorrect): Reminders can be automated now, without waiting for a big change.
• “With fee reminders, and move marks onto a digital form first” is correct.

  Fee reminders are frequent, rule-based and already digital, so they can be automated straight away. The test marks start on paper, so the first step is collecting them through a simple form. After that, marks entry can be automated too.

  • Order: automate what's ready, prepare what isn't.
  • Quick win: builds trust for the next project.

📖 Revise: Lesson 1: Spotting repetitive work

Why each option is right or wrong

• “A training problem, so staff should type more carefully” (Incorrect): Care helps, but the repeated typing itself is the waste and the source of errors.
• “A sign that the courier should be changed” (Incorrect): The courier may be fine; the problem is moving data between systems by hand.
• “Nothing important, since each step only takes a minute” (Incorrect): A minute per order adds up quickly, and errors cause returns.
• “A pain point: the same data copied between systems” is correct.

  Typing the same details into two systems wastes time and creates typos, such as wrong pin codes. Marking it as a pain point on the map highlights it as one of the best places to automate, for example by sending the address from the order sheet to the courier automatically.

  • Look for: re-typing, waiting and lost work.
  • Check: whether the courier offers an API or an upload option.

📖 Revise: Lesson 2: Mapping a process step by step

Why each option is right or wrong

• “Automate the replies fully, so complaints are handled faster” (Incorrect): Complaints need empathy and the authority to offer fixes; automatic replies can make things worse.
• “Refuse the project, since complaints can't involve automation” (Incorrect): Parts of the work, such as routing and logging, can still be automated.
• “Send every complaint a discount code automatically” (Incorrect): That rewards complaints blindly and doesn't solve the customer's problem.
• “Automate routing and logging; people write replies” is correct.

  Complaints need judgement, empathy and sometimes a refund or a fix, so a person should reply. The workflow can still help: route each complaint to the right person, log it, and remind the team if it hasn't been answered within a set time.

  • Middle path: automate the routine parts around a human decision.
  • Measure: time to first reply, before and after.

📖 Revise: Lesson 4: When not to automate