Syntha Lab RU

Post-promotion dip and pull-forward demand

From weeks before, during and after a promotion, estimates the baseline, the uplift, the post-promotion dip and the share of uplift eaten by the dip. Shows how many weeks it took sales to return to baseline and suggests a gap before the next promotion.

Promotions: planning and evaluation For whom: Category managers and analysts 2–6 weeks after a promotion.

Calculate on your data →

At a glance

What question does it answer?
How much of the promotion uplift was pulled-forward demand and when sales returned to normal.
What data is needed?
Weekly sales for 4–8 weeks before, during and 4–6 weeks after the promotion.
How to read the result?
The baseline is 50.8 pcs a week. The promotion gave an uplift of 104.3 pcs and a dip of 18.3 pcs over 4 weeks after: the dip ate 17.6% of the uplift, net uplift 86 pcs. Sales returned to baseline in the fourth week after, so the gap before the next promotion is at least 3 weeks. In money at margins of 3,000 and 2,000 the net result is about +52,000. Example on the default values: Baseline: weekly sales — 50.8 pcs; Uplift over the promotion — 104.3 pcs; Post-promotion dip — 18.3 pcs; Net uplift — 86 pcs.

Why and how to reason

After a promotion sales often fall below normal: buyers have already taken what they would have bought later. So the gain from a promotion must be counted net: gain minus dip. First the baseline is estimated, the normal sales level: the average of weeks before the promotion, 50.8 units a week. Promotion gain: two weeks gave 110 + 96 = 206 against an expected 2 × 50.8 = 101.7, i.e. +104.3. The post-promotion dip: the weeks after fell below baseline by 18.3 units in total. Net gain 104.3 − 18.3 = 86 units, the dip ate 17.6% of the gain, and sales returned to baseline after 3 weeks. In roubles, with margins of 3,000 and RUB 2,000, the net result is RUB 52,000. The dip means part of the success of the promotion is moved sales, and it cannot be counted as success in full.

When to use it

It comes in handy after a promotion when you have weekly sales before, during and after, and before planning the next one to choose the gap. Assumption: baseline is stable, and season and weather do not move sales much. It does not fit goods with a sharp seasonal trend (the baseline will mislead) or promotions shorter than a week. For different goods you can set seasonality coefficients by week (column s).

How to read the result

Look at the share of gain eaten by the dip and at weeks until return to baseline. A share of 17.6% is moderate; above 40–50% means the promotion mainly moves purchases, and the next should be rarer or shallower. Weeks to return set the gap between promotions: if sales return after 3 weeks, a promotion every 2 weeks gives no time to recover. In roubles the dip is valued at full-price margin and the gain at promotion margin: this shows what was lost. Plan the next promotion no sooner than the weeks needed to return to baseline and check it does not overlap a seasonal rise: then the dip will not stack on expected growth.

Common mistakes

The usual mistakes: estimating baseline from weeks with a previous promotion; looking only at the week right after; ignoring stock: a dip can come from a lack of goods and not from demand; comparing with last year’s season without adjustment. If there are few post-promotion weeks, the dip is understated. Do not stop at one week after the promotion: the dip can be spread out.

What data is needed?

  • Weekly sales for 4–8 weeks before, during and 4–6 weeks after the promotion.
  • Optional: the seasonal index of weeks and unit margin at full price and in the promotion.

Inputs

FieldUnitValue in the exampleMeaning
Unit margin at full price₽3,000Excl. VAT. Empty: money not calculated.
Unit margin in the promotion₽2,000Gross margin per unit in the promotion, excl. VAT: the promotion price minus cost.
Threshold: share of uplift eaten by the dip%—If the dip ate more, the measure turns amber. Empty: no threshold, the color does not change. There is no single standard in Russia: enter your own.

How it is calculated

StepFormula
Baselineaverage (sold ÷ seasonal index) over the weeks before × the week’s index
UpliftΣ (actual − baseline) over the promotion weeks
DipΣ (baseline − actual) over the weeks after
Net upliftuplift − dip

Worked numbers

Default values from the app; in the interactive version you enter your own.

Baseline: weekly sales
50.8 pcsAverage before the promotion adjusted by the seasonal index.
Uplift over the promotion
104.3 pcs
Post-promotion dip
18.3 pcsSum of (baseline − actual) over the weeks after; negative means sales above baseline.
Net uplift
86 pcs
Share of uplift eaten by the dip
17.6 %
Weeks until back to baseline
3 · gap before the next promotion at least 3 wk
Net result in money
52,000 ₽Promotion margin − baseline margin − the dip at the full-price margin. Without cannibalisation and costs.

Promotion weeks and after

PeriodWeekSoldBaselineDeviation
PromotionPromotion 111050.859.2
PromotionPromotion 29650.845.2
AfterAfter 14150.8−9.8
AfterAfter 24450.8−6.8
AfterAfter 34850.8−2.8
AfterAfter 45250.81.2

Link to neighbouring calculations: this one measures the dip week by week after the promotion; “Promotion result: incremental profit” takes the dip as one figure (“Pull-forward”): enter the measured value there. “Elasticity from a past promotion” derives the dip from sales after the promotion over one period.

No visible dip does not prove there is none: other changes can mask it. For clothing the dip is weaker than for groceries, but shifting purchases into the promotion and waiting for the next discount remain.

Nuances and limits

  • No visible dip does not mean there is none: other changes can mask it.
  • Few weeks before the promotion make the baseline noisy.
  • For clothing the dip is weaker than for storable groceries, but shifting purchases and waiting for discounts exist.

Terms used

Calculate on your data →