Final Case Analysis

From the dashboard to a final decision

Take out your Week 16 management dashboard. Each line on it names something the company watches, its owner, the threshold that calls for action and the action itself. Some lines came from the Week 15 risk register, such as written photo approval and Mina's working hours. Others came from the Week 13 cash calendar, such as the lowest cash point. This week, those lines help answer the largest question in the course.

In the practice scenario, Judy's company has finished its 12-week test of Option A · Expansion (expand cheerleading and sports-entertainment talent management). Now Judy's company must decide what to commit to next, and all three options are open again. Option A could grow. Option B · Boutique (a small cross-border talent and advisory service for a few clients) would narrow the work. Option C · Content-centered return (return to content, advertising, food service and personal brand) would bring the company back to the work it did before the public attention.

In Week 17 you choose one option and defend it in a final decision memo. This is the Week 3 decision memo, grown into a full recommendation. In Week 3 you compared the options before there were any test results. The final memo uses everything the course has built since then, from the evidence audit to the dashboard.

As in earlier weeks, this chapter follows the practice scenario, which is invented for learning. This week's question is: how can you recommend one option in a way that another manager can check, challenge and change when the conditions change?

Where this week fits

Week 16 turned risks and cash dates into a dashboard with owners, thresholds and actions. Week 17 compares all three options in a final decision memo and uses the dashboard's early-warning lines as stop signals. Week 18 turns that memo into a company blueprint that another manager could run.

A recommendation is a choice with its conditions attached

A recommendation is a choice offered to the person who decides, together with its reasons, the assumptions behind it and the signals that would change it. The last part matters most, because conditions change. A team that pays late, a coordinator who falls ill or a new platform rule can turn a good choice into a poor one. A recommendation that names its conditions tells the next manager what to watch.

Compare two sentences about the practice scenario.

• An unsupported claim: "Judy's company cannot manage larger events." The case has no record of the company's capacity, so nobody can check it.
• A conditional recommendation: "The company should accept no more appearances until payment timing, Mina's hours and written consent have been checked against the dashboard thresholds." It tells the manager what to check and when to act.

The second sentence also keeps apart three kinds of statement you have used since Week 1. Evidence is what a source shows, such as the Taipei East page that describes the Tokki Cutie collaboration. An inference is our interpretation, such as "public attention may bring more requests". An unknown is something nobody has checked yet, such as whether those requests turn into paying customers.

In Week 9 your evidence audit sorted the claims in your decision memo into four kinds: these three, plus design, a choice the company makes in order to test something. Bring that audit into the final memo, so that the reader can see which reasons rest on evidence, which rest on inference, which are still unknown and which are the company's own design choices.

Write the decision statement first

"Which option is best?" is too vague to answer. Best for which period, for which people, and with how much risk? A decision statement names the choice, the period and the limits the choice must respect. It does not answer the question. It tells you what the comparison must include.

Here is a decision statement for the practice scenario: "Over the next 12 months, which option should Judy's company commit to, while keeping cash above NT$100,000, publishing no image without the person's written approval, and keeping Mina's work within an agreed number of hours?" The cash floor and the photo rule are red lines from the Week 5 goal tree. The limit on Mina's hours comes from the Week 11 capacity check and the Week 15 risk register.

The limits are written as thresholds, not as goals to balance. As Week 5 explained, a red line must be met before an option can be chosen at all, so an option that would push cash below the floor needs redesign before it can be compared.

The five criteria

The final decision memo compares the options on five criteria. Four are the management questions you have used since Week 1: Target, Value, Capability and Revenue Model. The fifth, People and risk, gathers what Weeks 14 and 15 added, such as fair treatment, consent and the risks in your register. Use these five criteria, with these names, everywhere in the memo.

You may still score the options with a weighted matrix, as in Week 3. Remember that the weights are value choices. If a small change in the weights changes the winner, say so in the memo, because the recommendation then depends on values as much as on evidence.

Compare the three options as whole systems

An option is more than a slogan such as "grow" or "focus". Each option is a business system: a set of activities, people and payments that must fit together. Zott, Amit and Massa (2011) reviewed research that describes a business model in this way, as connected activities that create value and capture part of it. Amit and Zott (2012) turned the idea into practical questions about which activities are done, how they link and who does them.

The table asks the same five questions of each option. Each cell states what must be true for the option to work; it is a starting point, not a verdict.

Each option moves the difficulty to a different place: A to coordination and cash timing, B to a few relationships and Judy's time, C to running several kinds of work at once. The memo judges which difficulty the company can best carry.

Turn early-warning lines into stop signals

A disconfirmation signal is a piece of evidence, named in advance, that would weaken the recommendation enough to make the company pause, redesign or stop. In this course we call it a stop signal. Writing it down in advance stops the company from defending a choice only because it has already been announced.

Your Week 16 dashboard is the best source of stop signals. A leading indicator shows a problem before the final result arrives, such as a payment that is already late or a coordinator whose hours keep rising. A lagging indicator, such as the year's profit, arrives too late to prevent the harm. So the final memo takes its stop signals from the leading lines. The table adapts the Week 16 lines: the thresholds come from the dashboard, plus design choices for the memo, such as holding new agreements with a late-paying team until its fee is paid.

A stop signal is useful only if someone can use it, so each row names who acts. People inside the work often see a stop signal first. A performer who reports an unsafe schedule is giving the company information it needs, and should be thanked for it.

The parts of a final decision memo

A final decision memo has seven parts, each built from work you have already done.

Worked example: one dashboard, two final decision memos

This example shows how two student groups can read the same dashboard and reach different options. Both memos are reasonable, because each states its assumptions and its reasons follow from them.

The dashboard at the end of the 12-week test

Both groups also use the same planning figures. Each appearance costs NT$35,000: NT$25,000 is paid in the week of the game, and NT$10,000 on the payday after it. Under the Week 14 rule that nobody waits for pay, all NT$35,000 goes out before the fee arrives. The test ended with NT$230,000 in cash after the game 2 fee arrived (Week 13). The company then paid for next season's preparation and other commitments, so it plans the next period from NT$150,000.

Two groups, two sets of assumptions

Group 1 makes three assumptions. The late first payment was a one-time problem, and the home team has now agreed in writing to pay on day 30. A second part-time coordinator who reads both languages can be trained in eight weeks. The company values building a repeatable service more than anything else in this period.

Group 2 makes three different assumptions. New teams will usually ask for 45-day terms, as the second team did. No second coordinator can be found this year. The company values protecting its cash buffer and Judy's and Mina's time more than anything else in this period.

The assumptions change the cash arithmetic. Group 2 expects bookings from new teams on 45-day terms with no deposit. Suppose two of them book appearances close together. Before either fee arrives, the company pays 2 × NT$35,000 = NT$70,000. Cash falls from NT$150,000 to NT$80,000, which is NT$20,000 below the floor.

With only one appearance at a time, cash falls to NT$115,000. That stays above the floor, although it is below the NT$130,000 action threshold.

Group 1 expects different bookings: one from the home team, which pays 30 days after the game with no deposit (its late game 1 fee has since been paid, so the late-fee stop line does not block a new agreement), and one from a new team, which pays a deposit of NT$30,000, half the fee, before the game. If both are booked close together, cash falls to NT$150,000 − NT$70,000 + NT$30,000 = NT$110,000. That is above the floor but below the action threshold, so Judy decides on the second booking before it is accepted. Group 2 does not believe new teams will accept a deposit, so it cannot use this solution.

The two memos side by side

What made the memos differ

The two groups used the same evidence. Their memos differ because their assumptions differ, and each memo says so openly. Group 1 would change its recommendation if new teams refused deposits; Group 2 would change its recommendation if they accepted them.

A third group could reach Option B · Boutique, for example by assuming that a sponsor would pay in advance for advice on working with Korean performers. This chapter does not tell you which option is right. Your memo is judged by whether its assumptions are stated, its reasons follow from them and its stop signals can be used.

A sensitivity check: change one assumption at a time

A sensitivity check tests a memo by changing one assumption at a time and asking whether the recommendation still holds. The table lists common changes for the practice scenario.

For Group 1, the first row would change the recommendation first: if payments slip and deposits are refused, cash falls below the floor. So that is the assumption its first review must check.

Commit in stages: four gates and four phases

Whichever option a memo recommends, the company should commit in steps, so that evidence from each step can shape the next one. Gavetti and Levinthal (2000) used computer models to show how learning by thinking ahead and learning from experience can support each other. Week 16 introduced March's (1991) distinction between improving what already works and exploring what is new. A staged commitment keeps room for exploring without betting the whole company on it.

A useful staging tool is a set of four gates. The company passes a gate only when the evidence is in.

1. Customer gate. The customer understands the offer, and it is clear who uses the service and who pays.
2. Capacity gate. The company can deliver without hidden, unpaid hours from Judy or Mina and without unsafe schedules for performers.
3. Cash and quality gate. Payment timing keeps cash above the floor, changes to the scope are agreed in writing, and there is a remedy if the promise is missed.
4. Learning gate. The work has produced routines that someone else could repeat, and people have been able to raise concerns.

The gates sit inside four phases of implementation.

In the review phase, ask whether a problem came from a bad assumption (the plan rested on something untrue) or poor execution (the plan was sound but the work went wrong). Each leads to a different next step.

Sitkin and colleagues (2011) argue that very ambitious goals are most dangerous when time, money and feedback are scarce. After the test, Judy's company is in that position: cash dipped below the action threshold, and Mina used all her reserved hours before game 1. So whichever option a memo chooses, it should grow in stages.

Explain the options you did not choose

A final memo explains why the other options come later, and it leads with the reason. "Option A waits because the company cannot keep cash above the floor on 45-day terms" is useful. It tells the reader which evidence would bring Option A back. "Option A is too risky" gives the reader nothing to check.

Write one reason for each option you did not choose, and link it to one of the five criteria. For example, a Capability reason might be that the work depends on one person. Then write what would change your mind.

Every memo also ends with a dissent box. It states the strongest argument for an option you did not choose and names the evidence that would make that argument stronger. It also suggests a small test that could settle the disagreement. A dissent box makes the memo more honest and gives the next manager a question to check.

Can the company change when conditions change?

In Week 7 you met dynamic capabilities: a company's ability to renew what it can do as conditions change (Teece et al., 1997). Teece (2007) divides this ability into three activities. Sensing means noticing a change, such as a team asking for longer payment terms. Seizing means committing resources to respond, such as asking every new team for a deposit. Transforming means changing routines, roles or structure, such as training a second coordinator so that the work no longer depends on Mina alone.

In the final memo, these three words become one question for each option: how much would the company have to transform to make it work, and who will lead the change?

Check who can read and use the memo

Week 4 showed how to compare settings by mechanisms, such as who approves and which language version controls, rather than by national character. In the final memo, that method raises one practical question: can every person affected by a stop signal read it and use it? If the memo exists only in Chinese or English, the Korean-speaking performer may not know that she can pause a schedule that breaks the 12-hour rest rule. The performer liaison already sends her Korean versions of schedule messages, so the stop signals that concern her should travel the same way. State which version controls if the two differ.

Where AI can help, and this week's risk

An AI tool can sort your evidence record. The risk this week is specific: asked to "choose the best option", a tool will usually give a confident answer that hides the assumptions and the unknowns. If a tool proposes an option, ask what evidence it used and what it left out. The NIST Generative AI Profile (Autio et al., 2024) recommends naming the person accountable for AI-assisted work; for your memo, that person is you.

Class practice: write and defend a final decision memo

This practice is done without AI tools. Your teacher will tell you which worksheet to use and how to submit it. Bring your Week 16 management dashboard and your Week 3 decision memo.

1. State the decision. Write your decision statement, the option you recommend, and the two assumptions your recommendation depends on most.
2. Give your reasons on the five criteria. Write one sentence each for Target, Value, Capability, Revenue Model, and People and risk. Say why your option works on that criterion, or what must be true for it to work.
3. Choose two stop signals. Take two leading indicators from your dashboard. Turn each into a stop signal with a threshold and the person who acts.
4. Defend, then dissent. A classmate changes one of your assumptions, using a row of the sensitivity table. Write whether your recommendation continues, changes or pauses, and why. Then write your dissent box: the strongest argument for another option, and the evidence that would make it win.

There is no single correct option. A strong memo shows what you assumed and what would change your mind. Label any number you invent, and keep your memo for Week 18.

Key points and what you bring to Week 18

• A recommendation is a choice with its conditions attached: its reasons, its assumptions and the signals that would change it.
• Start with a decision statement that names the period and the red lines. An option that breaks a red line needs redesign before it can be compared.
• Compare all three options as whole systems, on the same five criteria: Target, Value, Capability, Revenue Model, and People and risk.
• Take stop signals from the leading indicators on your dashboard, and name who acts on each one.
• Two careful groups can reach different options from the same evidence. What makes a memo strong is stated assumptions, reasons that follow from them and a dissent box.

For Week 18. Bring your final decision memo. In Week 18 you will turn it into a company blueprint, a design that another manager could run tomorrow without calling the founder, and every earlier artifact will find its place in it.