An arc is a shape for the sequence in which a reader meets evidence. For analysis, the version that works has three movements, and each one opens on a specific number. None of them opens on a transition.
What are the three movements of a story arc?
What is establishes the reader's current picture and then breaks it. You open by agreeing. State the thing the room already believes, in their own terms, with the number they'd quote. For Turant that's the ₹10.82 loss per order, because it's what the board said in the last meeting and what the finance pack repeats. Then comes the fracture. The same data, cut differently, doesn't support what everyone believes. Cohort by cohort, the estate holds a business earning ₹71.72 an order and a business losing ₹6.90 an order, and the average describes neither.
What could be establishes the alternative, and most analytical stories collapse into wishing as they reach it. You have to build the alternative from something observed, because an assumption can't carry it. For Turant, "the new stores will mature" won't do, because it's an assumption wearing a forecast's clothes. A recorded fact will. At six months of age, the FY25 cohort was running 402 orders a day and earning ₹4,910 a day per store. At the same age, the FY26 cohort is running 310 and losing ₹2,139.02, a same-age gap of ₹7,049.02 per store per day. So the alternative is that the newest stores reach the density the previous cohort reached. The stake is that they're starting 22.9% behind, so the alternative won't arrive on its own.
The call to act converts the gap into an instruction with a cost, an owner and a date. Pause the forty planned openings, which defers ₹13.60 crore of capital spending. Spend ₹5.70 crore on a density programme across the 72 stores. If they reach 420 orders a day, the estate gains ₹13.11 crore a year and the programme pays back in 5.22 months. And 32 of the 72 stores getting there is enough to cover it inside a year.
Advertisement
Common mistakes with a story arc
The three movements answer three different questions. When a reader is lost, it's almost always because two of them have been merged. One common merge puts the recommendation inside the diagnosis, so the reader is asked to accept a conclusion before the fracture has landed. Another puts the alternative in the appendix. Then the story diagnoses a problem and stops, and you'll know the shape, because it's every deck that gets thanked and ignored.
Duarte's Resonate alternates between what is and what could be repeatedly across a long presentation. That's right for a keynote and wrong for a twenty-minute readout with a decision at the end. For a decision you run the alternation once and land it. That back-and-forth is a device for holding attention across forty minutes of a talk that has no decision in it, and a board meeting has the opposite problem.
How many pages should each movement get?
Duration matters to the shape. In a nine-page readout, what is takes about four pages, what could be takes three, and the call takes two with its costing and its risks inside them. Analysts habitually stretch the first block to six pages, because that's the part they did the work on. They squeeze the call into one, because that's the part they're least sure of. Your reader's interest runs the other way.