Source of Volume: What Really Drives Promotional Uplift

Source of Volume (SoV) decomposition answers the most important question in trade promotion: where did the incremental units actually come from?

Six productive and unproductive sources

A practical trade promotion playbook splits promotional volume into six sources, and each one is worth a different amount to you:

1. Category Expansion. New volume that grows the total category. Positive for both supplier and retailer.
2. Consumption Increase. Existing buyers genuinely using more at the lower price. Real extra demand for your own brand, and sales do not dip below normal once the promotion ends.
3. Competitive Switching. Volume stolen from competitor brands. Positive for supplier only.
4. Store Switching (some reports call it retail switching). Your own brand, bought in this store instead of another one. Positive for the host retailer only.
5. Internal Cannibalization. Volume stolen from your own products in the same store. Negative for supplier.
6. Stockpiling / Forward Buy. Volume pulled forward from future purchases, which shows up as the dip after the promotion. Its mirror image is the dip before it, from shoppers who waited for the deal. Negative for both supplier and retailer over a full cycle.

The hidden seventh source: Subsidized Base

A seventh source, Subsidized Base, captures volume that would have been purchased at full price regardless of the promotion: the baseline units sold on deal. As a share it is the % subsidized volume, baseline divided by promoted volume. The framework benchmark: roughly 50 percent of promoted sales would have happened anyway. On most promotional profit and loss (P&L) statements, this single line loses more money than any other.

Why the decomposition matters

A unit gained through category expansion is worth far more than a unit pulled forward from a future purchase. Most FMCG companies track total uplift but never decompose it, which is like measuring a patient's temperature without asking which organ is inflamed.

Net incremental volume, built up from the sources that stay in:

The same number, by taking the leaks out of the uplift:

Subsidized Base is not one of those deductions. It is baseline that sold on deal, so it never counted as uplift, and it costs you margin rather than volume.

The two shares that come out of it:

Whose gain each source is comes from two questions: does your brand sell more across the market, and does the retailer sell more in its store? The four answers:

• Shared gain (best): category expansion and consumption increase, and shoppers won from a competitor's brand at another store. Your brand sells more and the retailer sells more.
• Brand gain: competitive switching in the same store. Your brand sells more, and the store's category is the same size as before.
• Retailer gain: store switching, your own brand pulled from another retailer's shelf. The store gains a trip, and your brand across the market has not moved.
• No gain (worst): stockpiling, internal cannibalization and subsidized base. Your brand sells no more and pays the discount, and the retailer sells no more either. Volume moved in time or between your own packs.

Read it as three people reading the same event. Your brand across the market counts Shared gain plus Brand gain, which is net incremental volume. The retailer counts Shared gain plus Retailer gain, because it treats a shopper won from another store as new and a switch between brands on its own shelf as nothing. Your brand's sales in this one store count all three, and that is why one store's report reads better than the market does.

An illustrative scenario in biscuits. A brand ran a 25 percent off gondola‑end promotion and generated 12,000 incremental units. The headline looked strong; the decomposition told a different story.

What the headline hid

Decomposition broke the 12,000 units into:

• Category Expansion: 1,200 units (10 percent), genuinely new to category
• Competitive Switching: 3,600 units (30 percent), taken from rival brands
• Store Switching: 1,200 units (10 percent), your own brand shifted from other stores
• Consumption Increase: 1,800 units (15 percent), genuinely consumed more
• Stockpiling: 2,400 units (20 percent), pulled forward from future purchases, the dip after
• Cannibalization: 1,800 units (15 percent), stolen from own portfolio

Three readers of the same 12,000 units

• Net incremental volume, your brand across the market: 12,000 - 1,800 of cannibalization - 1,200 of store switching - 2,400 of stockpiling = 6,600 units, which is also 1,200 + 3,600 + 1,800. Net share of uplift = 6,600 / 12,000 = 55 percent, the number the sandbox tile shows for this mix. So 45 percent of the uplift either added nothing for you or cost you money.
• Your brand in this one store, the number on the buyer's report and on your key account manager's: 12,000 - 1,800 - 2,400 = 7,800 units. It reads 1,200 units higher because one store's sales data cannot see that those shoppers would have bought you somewhere else.
• The retailer's view: 1,200 + 1,800 + 1,200 = 4,200 units, 35 percent of the uplift. The retailer counts the 1,200 units won from other stores as new, and does not count the 3,600 units of competitive switching, which changed which brand sold without making its category any bigger. That 4,200 is the number the joint business plan conversation is about.

Post‑promotion tracking confirmed the diagnosis: sales dipped 18 percent below baseline for 2 weeks (stockpiling debt being repaid). Competitive switching was temporary: 75 percent of switched buyers returned to their original brand within 4 weeks. Only the 10 percent category expansion represented durable volume gain.

The redesign

The brand redesigned the next promotion: reduced depth to 15 percent, added sampling at the display, and cross‑merchandised with coffee. New SoV mix:

• Category Expansion 28 percent of the uplift
• Stockpiling 12 percent of the uplift
• Net share of uplift 72 percent

The decision rule

When the net share of uplift sits below 50 percent (this lesson's rule of thumb) and stockpiling is above 20 percent of the uplift, the promotion is selling next month's units this month. The fix is almost always shallower depth with better execution (display, sampling, cross‑category merchandising); a deeper price cut makes the stockpiling worse.

Breaking the volume down by source tells you which promotion type to run, when to run it, and how to set it up in store.

If your source is primarily competitive switching. You are winning the competitive battle but not growing the pie. Competitors will retaliate. Spend on getting shoppers to try the product and come back for a second purchase, using display support, sampling, and smaller trial packs, so the shoppers who switched for the deal turn into regular buyers.

If your source is primarily category expansion. This is the golden outcome. Retailers value category growth above all else. Document the expansion rigorously and use it as ammunition in JBP negotiations for more display space and promotional slots.

If your source is primarily stockpiling. You are training consumers to wait for deals and borrowing from future sales. Cut the discount, cap how many packs one shopper can buy on deal (limit 2), and move the money to offers that pay out on the next trip rather than on extra packs in this one: a collect‑and‑win, a loyalty‑card offer, or a coupon at the checkout for the next visit. A multi‑buy does the opposite, because it only rewards the shopper who takes several packs home at once, and the biscuit example on the Stockpiling / Forward Buy card shows a "Buy 2 for $5" doing exactly that.

If your source is primarily cannibalization. You are reshuffling your own P&L. Promote the SKUs that serve an occasion nothing else in your range serves, and keep the near‑identical ones off the plan. Analyze the cannibalization matrix to identify which products interact negatively and should never be co‑promoted.