Revenue Decline Cases 2026: Market First, Then Price

The first revenue decline question, and why it is not about the client

Revenue can fall for two structurally different reasons. The market shrank and the client held its position, or the market grew and the client lost position. Those two facts produce opposite recommendations from identical revenue lines.

If the market fell, the conversation is about cost structure, capacity and whether to consolidate. If the client lost share, the conversation is about the offer, the price and the channel. Spending the case on product assortment when the whole category contracted is a correct analysis of the wrong problem.

So ask for market size over the same period before asking anything else. If the client does not have it, ask for a proxy: competitor revenue, category volume from a trade body, or a public statistical series. The Census Bureau publishes one for American retail, reporting sales of "$773.9 billion" in August 2026 against "$764.5 billion" in revised July, checked 30 September 2026 at census.gov. That release also states that it "no longer contains the most up to date estimates", with revisions published on 28 September 2026, so use it with the caveat attached.

There is a second reason to start outside. A revenue line that is flat in dollars is falling in real terms whenever the category is growing, and clients rarely present it that way. A business reporting stable revenue across five years while its market compounded at 5 percent has lost nearly a quarter of its relative position, and nobody in the room has said the word decline.

Bain sets its published case up exactly this way. The brief is to "Advise a player in the fashion market on why it has experienced declining revenues for the past five years", checked 29 September 2026 at bain.com. Five years is long enough that the market question has to come first, and short enough that a single cause is unlikely.

Splitting a revenue decline into price, volume and the cross term

Once the market question is settled, the arithmetic can start. Here is a client to run it on.

An apparel retailer turned over $420 million in 2021 and $402 million in 2026. The fall is $18 million, which is 4.3 percent over five years. Unit volume went from 12.0 million to 10.0 million. Average selling price went from $35.00 to $40.20.

Confirm both years reconcile before going further. 12.0 million units at $35.00 is $420 million. 10.0 million units at $40.20 is $402 million. Both check.

Now the two movements. Price rose $5.20, which is 14.9 percent. Volume fell 2.0 million units, which is 16.7 percent. A candidate who reports those two percentages and concludes that volume dominates has skipped a step, because percentages of different bases cannot be compared directly.

The decomposition uses dollars. Price effect is the change in price across the old volume: $5.20 across 12.0 million units gives plus $62.4 million. Volume effect is the change in volume at the old price: minus 2.0 million units at $35.00 gives minus $70.0 million. The cross term is the change in price across the change in volume: $5.20 across minus 2.0 million units gives minus $10.4 million.

The three add to exactly minus $18.0 million, which is the fall in the revenue line. A decomposition that does not reconcile to the total is a decomposition with a term missing, and the missing term is nearly always the third row.

The cross term is why the revenue decline is net negative

Row three deserves its own discussion, because it is where the interesting sentence lives.

Price contributed plus $62.4 million and volume contributed minus $70.0 million. Those two alone net to minus $7.6 million, which is less than half the observed fall. The remaining minus $10.4 million comes from the interaction: the price rise was earned on 2.0 million fewer units than the base it was measured against.

Read what that means commercially. The retailer raised price and the volume it lost was the volume that would have carried the price rise. So the price action was not merely offset by the volume loss, it was partly cancelled by it. At a large enough volume fall, a price rise stops adding revenue altogether.

Here is the threshold. Revenue holds when the new price times the new volume equals $420 million, so at $40.20 the retailer needed $420 million divided by $40.20, which is 10.45 million units. It delivered 10.0 million, so it fell 0.45 million units short, which is 4.3 percent below the volume the price rise required.

That is the number to put in the recommendation. The price move was 4.3 percent of volume away from working, which is a much more useful statement than saying price rose and volume fell. It also tells the client where to look: a 4.3 percent volume recovery at the current price restores the 2021 revenue line without touching price at all.

Half the price rise was mix

The $5.20 price rise is itself a composite, and splitting it changes the diagnosis completely.

Suppose the retailer sells two lines. A core line at $30.00 and a premium line at $55.00, and neither list price has changed since 2021. In 2021 premium was 20 percent of units, so average price was 0.8 times $30.00 plus 0.2 times $55.00, which is $24.00 plus $11.00, so $35.00.

By 2026 premium is 30 percent of units. With both list prices unchanged, the average becomes 0.7 times $30.00 plus 0.3 times $55.00, which is $21.00 plus $16.50, so $37.50.

So $2.50 of the $5.20 rise came from mix alone, and that is 48.1 percent of the move. Only $2.70 came from actual pricing. Nearly half of what looked like a price increase was the composition of the basket changing.

This matters because mix and price fail differently. A list price rise can be reversed next week. A mix shift toward premium usually means the core line lost its cheapest buyers, which is a volume problem wearing a price disguise. The volume fall and the mix shift may be the same event seen twice.

The test is simple and worth naming. Ask for units and revenue by line for both years. If core units fell faster than premium units, the mix shift is the footprint of a lost customer group rather than a successful trade up. The Census Bureau's own definitions keep this discussion honest, noting that gross margin "represents total sales less cost of goods sold", checked 29 September 2026 at census.gov, so a mix gain that comes with a volume loss can raise margin percentage while shrinking margin dollars.

How much of the revenue decline the channel explains

External causes get asserted in these cases and rarely sized. The online channel is the usual suspect in retail, so size it.

The Census Bureau reports that "E-commerce sales in the second quarter of 2026 accounted for 17.1 percent of total sales", with online growing 12.2 percent year over year against 6.7 percent for total retail, checked 29 September 2026 at census.gov.

From those three figures the offline growth rate follows. Weighting online at 17.1 percent and offline at 82.9 percent, the two have to average to the total, so 0.171 times 1.122 plus 0.829 times one plus g equals 1.067. That gives 0.191862 plus 0.829 plus 0.829g equals 1.067, so 0.829g equals 0.046138 and g is 5.57 percent.

Round that to 5.6 percent. Offline retail grew 5.6 percent while total retail grew 6.7 percent, so the online channel costs the rest of retail about 1.1 points of annual growth at current weights. That is a real drag and a smaller one than the story usually implies. One caveat belongs with the derivation: the correct weight is the share at the start of the period and this uses the share at the end, which nudges the offline figure slightly high.

Now measure our client against it. Revenue fell from $420 million to $402 million over five years, a ratio of 0.9571, which is about 0.87 percent of decline a year. Against an offline market compounding at 5.6 percent, relative share moved by 0.9571 divided by 1.0557 to the fifth power. That denominator is 1.3113, so relative share is 0.7299 of where it started.

The client lost 27.0 percent of its relative position in five years. The channel accounts for 1.1 points of the annual gap and the client lost 6.1 points a year, so the channel is worth roughly a sixth of the problem. Everything else is the client. The honest limitation is that this uses a total retail series rather than an apparel one, because the release does not publish apparel separately, so it is a first pass to be replaced with category data.

What Bain publishes about a five year decline

The only published decline case with quantified options is worth reading closely, and worth reading for what it withholds.

Bain describes a client that "has experienced declining revenues each year for the past five years" and then offers two options, both with full assumption sets. Option A is an awards program: 25 percent participation in year one, 10 million customers, $100 of average annual spend, a one time $50 activation fee, and a permanent 20 percent discount. Option B is intermittent sales: the same 10 million customers at $100, a 20 percent discount for three months a year, with sales volume doubling during those three months. The firm concludes that it recommends Option B "as it is projected to generate $50M more than the awards program (Option A)", checked 29 September 2026 at bain.com.

Both totals are published as images, which we have not read, so we do not restate either as a fact. What we can do is run Option B from the stated assumptions and show the shape.

Baseline revenue is 10 million customers at $100 a year, which is $1.0 billion. Three months carry a quarter of that, so $250 million. Double the volume and discount 20 percent and that quarter becomes $250 million times two times 0.8, which is $400 million. The nine untouched months stay at $750 million, so the year totals $1.15 billion, a gain of $150 million and 15 percent.

Option A cannot be run the same way, because the published assumption set leaves one term ambiguous. A one time $50 activation fee could be revenue charged to the joiner or a cost borne by the retailer, and the case does not say. At 25 percent participation there are 2.5 million joiners, and 2.5 million participants at $50 is $125 million. That single ambiguity is two and a half times the $50 million gap the firm reports between the options, so it decides the comparison on its own.

That is the right way to use a published case with an unreadable exhibit. State what the firm states, run what the text supports, and name the term that would change the answer rather than guessing at it.

Where a decline answer loses its points

The weights below come out of our own grading code in src/lib/grading. They are ours, no firm publishes an equivalent, and the argument for each sits on how an answer is assessed.