Budget allocator template for Ecommerce

The budget allocator with three columns a store needs: which slice of the catalogue a share belongs to, the margin on that slice, and how much of the total is being held back for a promotion rather than committed now.

Everything else is the general allocator, unchanged. One row per destination, a basis for every share, a floor and a ceiling, and a reallocation rule written before it is needed.

Use it when the money has to be divided across parts of a range whose economics differ, which is the normal condition of a store. A split argued channel by channel will quietly give the largest share to whichever segment is easiest to sell, and that is rarely the one that contributes the most.

Use it again before every promotion, because a peak that was not reserved for has to be funded by taking money from lines that are working.

The reason a store needs its own version of this document is that money behaves differently depending on what it is pointed at, and the difference is not visible in the amount.

A share aimed at a thin-margin segment is buying orders that contribute less per order. A share aimed at clearance is buying a sell-through date rather than a profit. A share aimed at new arrivals is buying information about a range nobody has data on yet. Three identical amounts, three different things bought, and a split argued only by channel cannot tell them apart.

So the catalogue segment and its margin come into the row, and the argument in the basis column is made with both in view. What you are deciding is not how much Meta gets; it is how much of the quarter's money is pointed at the part of the range that pays, the part that has to move, and the part you are still learning about.

The margin column is what makes a share arguable rather than merely agreed.

Fill it after discount and after expected returns, and after shipping where the store carries it. Those adjustments are where the surprises are: a segment that discounts hard and a segment with a high return rate can both look healthy on revenue and be the two thinnest rows in the allocator on contribution.

Then read the share against it. A generous share pointed at a segment whose margin cannot support the cost of winning an order is not an investment, it is a subsidy, and it is usually one nobody intended. The row is where that becomes visible before the quarter rather than after it.

Where the margin is genuinely unknown, leave it blank and say so in the notes. A blank margin with a name against it is a question somebody answers this week. A guessed one is a share defended for a quarter by people who never saw where the number came from.

The promotion reserve column exists to prevent the most predictable failure in retail budgeting.

A store knows its peak is coming. It allocates the quarter across its working lines, and then the peak arrives and the money for it is taken from those lines, in the fortnight when they are performing best and when demand is highest. The peak is funded by dismantling the thing that was working, and the post-peak weeks are then underfunded as well because the borrowed budget never comes back.

Reserving fixes it with one column. Hold a named portion of the total for the window, record the window in the cell, and treat the reserve as unavailable for ordinary reallocation. It costs a smaller working budget in the quiet weeks, which is the correct price.

Clearance segments belong in the allocator and they do not belong in the same row as full-price selling.

Their basis is usually not measured performance at all. It is a commitment: stock has to move by a date, and the share is whatever it takes. Their metric is sell-through rather than a cost per order, and their margin is by definition compressed. A clearance row judged against a full-price target will always look like the worst row in the sheet, and turning it off does not make the stock disappear.

Name the basis as a commitment, say what the date is, and judge it on whether the stock moved. That is honest, and it stops a clearance line dragging down the average that the rest of the split is argued from.

Every allocator should carry ceilings. In a store they earn their place in exactly one week a year, which is the week the peak arrives and money starts moving quickly.

Audiences saturate fast at a peak. A row that could absorb a steady weekly amount all quarter cannot absorb four times that in a fortnight, and money poured into it beyond its ceiling buys frequency rather than reach. The ceiling column is what tells the person moving the money where the headroom actually is, at the moment when there is no time to work it out.

Where a ceiling is not known, say so rather than leaving the cell empty in a way that reads as unlimited. Unknown and unlimited are very different instructions.

The reallocation rule in a store should move money slowly and on a condition that is not a single week.

Retail weeks are noisy. A promotion, a sold-out hero product, a delivery cut-off or a competitor's sale can move a cost per order in either direction for reasons that say nothing about the segment. A rule that reacts to one week will chase that noise around the catalogue all quarter.

Two consecutive weeks, a capped proportion, and a named destination with headroom. That is enough to be responsive and slow enough to avoid buying the last seven days of weather.

Every row names a catalogue segment and its margin, or says plainly that the margin is not yet known. Shares sum to one whole per currency. Every row has a floor and none is funded below it. The promotion reserve is set aside before the peak rather than taken from working lines during it. Clearance sits in its own rows with its basis named as a commitment. And every row carries a reallocation rule written while nothing was going wrong.