The summer peak is over, the data is in, and this is the most informative moment of the beverage year. It is also the moment most brands waste by looking at one number.
Total volume tells you almost nothing you did not already know. Here is what to actually pull apart.
Velocity per point of distribution
Total sales conflate two completely different things: how many places carry you, and how fast it moves in each.
A brand that doubled volume by doubling its store count has learned nothing about whether the product works. A brand that held store count and doubled velocity has learned that it works very well.
Calculate cases per store per week. That is your real number. It is the number a buyer will ask for, and it is the only one that predicts what happens when you expand.
The SKU distribution
Pull volume by flavor and rank it. Almost every multi-SKU beverage brand finds the same shape: one or two SKUs carry most of the volume, and the tail is thinner than anyone expected.
The useful questions:
- Is the top SKU top everywhere, or only in some accounts? Uniform means you have a lead product. Variable means you have a merchandising or placement problem.
- Is the bottom SKU earning its facing? A slow SKU occupies shelf space that your fast SKU could use. Cutting it is usually the highest-return decision available.
- Did the tail move at all? SKUs that did not sell during the peak will not sell in November.
Founders resist cutting flavors because they chose them personally. The shelf does not care.
Stockouts, which are invisible in sales data
A stockout looks like flat sales. It is actually lost sales plus a damaged relationship, and nothing in your revenue report distinguishes it.
Reconstruct it: compare delivery dates against expected depletion. Where the gap exceeds the replenishment cycle, you were empty. If you have a DSD partner, they know exactly where and when — ask.
Stockouts during a peak are the most expensive failure in this business, because the facing does not stay empty. It gets filled with something else, and getting it back costs more than keeping it did.
Account-level dispersion
Sort accounts by velocity, highest to lowest. Look at the shape.
Tight distribution — most accounts performing similarly — means the product's performance is a property of the product. Expansion should work.
Wide dispersion — a handful of accounts carrying everything — means performance is a property of those accounts. Understand what they have in common: demographic, placement, cold box, a manager who likes you. Then target expansion at accounts that share it rather than at anything with a door.
Most brands discover their winning accounts share something specific and unglamorous, like being within a mile of a particular kind of workplace.
Cold versus ambient
If you have both cold single-serve and ambient multipack placement, split the analysis. They behave differently enough that averaging them destroys the signal.
Cold single-serve spikes hard in summer and drops hard after. Ambient multipack is steadier. A brand reading a combined number in September will conclude demand is collapsing, when what actually happened is that the seasonal component ended on schedule.
What to actually do with it
Four decisions, in order:
- Cut the slowest SKU unless there is a strategic reason to keep it. Reallocate the facing.
- Produce Q4 against the mix you observed, not against last year's plan or an even split.
- Fix the stockout accounts before adding new ones. Reliability is worth more than reach.
- Define your expansion target by what your best accounts have in common.
The discipline part
The analysis is not hard. The hard part is acting on it when the answer contradicts something you decided emotionally — usually a flavor you love that nobody buys.
Every brand we work with has one. The ones that grow are the ones that cut it.