Level 7 · Review the momentum chart

What a cumulative chart shows

A cumulative chart plots a running total: January, then January plus February, and so on. Each point is everything so far. As long as a month brings in any revenue, the line goes up. A month with half the usual revenue still pushes it up, just less steeply.

monthly = revenue.groupby(revenue.index.to_period("M")).sum()
monthly.cumsum()     # the line on Jordan's slide
monthly              # what each month actually brought in

The information is in the slope, and people are poor judges of changing slopes. “It went up every month” is true of almost every business with a running total, which is why the chart looks good on every slide.

What the board actually asks

“Momentum into 2026” is a claim about the end of the year compared with what came before. The honest comparison for monthly retail data is year over year: each month against the same month last year, because seasons repeat. November against October always looks great in retail.

On Harbor & Pine’s data, 2025 brought in $2.30M against $1.99M in 2024, 15% more. But month by month, September, November and December 2025 were below 2024, November by 8%. The holiday quarter, the one that matters most for a retailer, was weaker than the year before. The cumulative line hides all three months.

Reviewing any chart

Before a chart goes into a deck, ask four questions:

1. What does each point represent? A running total, a monthly value, a 7-day mean?
2. Compared with what? Last year, a target, another group? A number with no comparison can’t show “good” or “bad”.
3. Where does the axis start, and is there only one? Bars from zero, one y axis.
4. Does the title claim more than the data shows? “Momentum” is a claim about the trend at the end of the period.

Why not two y axes

A second y axis on the right lets you put two measures on one chart, and the scales can be chosen so the lines cross, run together or diverge, whatever the author likes. Readers can’t tell. Plot comparable measures on one axis, or use two charts side by side.

Common mistakes

• Running totals as performance. Use them for progress toward a goal, like “revenue to date against the annual target”.
• Month-over-month comparisons in seasonal data. December beats September every year. Compare with the same month last year.
• Fixing the look instead of the content. A different colour or axis range doesn’t change what the line measures.