The Case for Ending Half Your Promotional Plan

Half the events on your promo calendar do not earn their cost. You already know which half. The part nobody wants to do is decide, event by event, whether to kill the thing or just shrink it. This guide is the sequence most senior commercial teams end up walking anyway, written down so you can walk it before the calendar locks instead of after the writedown.

The calendar nobody wants to review

Every Fast-Moving Consumer Goods (FMCG) company has a promo calendar that has grown organically over a decade. Events were added in good years and never taken off in bad ones. Retailer asks turned into commitments. Seasonal mechanics became permanent. Somewhere along the way the calendar stopped being a plan and started being an inheritance.

The published numbers on what those calendars actually deliver are unflattering. NielsenIQ's 2022 industry note on trade promotional effectiveness summarised the pattern bluntly: "over half of all trade promotions result in little to no sales lift, meaning manufacturers are ultimately wasting time and money."¹ The exact share that fails depends on category, retailer mix, and how strictly you define incrementality. But the headline that more than half of trade promotions do not earn their cost is consistent enough that most senior commercial leaders have heard it three or four times at industry conferences and will not argue with the shape of it.

The reason calendars keep growing despite that benchmark is that ending a promotional event is politically harder than adding one. The retailer objects. The brand team fears the baseline. The sales team has the feature slot already on its spreadsheet. Each individual event looks too small to fight about, and the aggregate problem only becomes visible when Finance adds up what the full calendar costs and nobody can point to the incremental profit behind half of it.

So the first thing worth accepting is that if your calendar has 40 events this year, roughly 20 to 25 of them are not paying for themselves on their current shape. That does not mean 20 to 25 events should be deleted. About half of the loss-making events are loss-making because the event is fundamentally redundant, and those should be ended. The other half are loss-making because the current design of the event is too deep, too frequent, or too long, and those can be rebuilt into something smaller that does pay. The rubric below is the sequence that sorts the two.

What ending looks like, what shrinking looks like

Ending an event is exactly what it sounds like. The window comes off the calendar. No feature, no display, no price cut, no bonus pack. The Stock Keeping Unit (SKU) sits at its baseline shelf price in that window and the trade-spend line item drops to zero. Your risk is a short-term dip in baseline volume in the window the promo used to fill, usually 5 to 15 percent, followed by recovery over 2 to 4 cycles. Your upside is the full spend drops back into your Profit and Loss (P&L), plus whatever reinvestment you route to higher-ROI activities.

Shrinking an event means you keep the window but reshape what happens inside it. There are four main dials you can turn.

• Depth. Cut the discount from 30 percent off to 15 percent off. The mechanic still shows up on the feature ad and still gets shelf flagging, but the cost per unit drops by roughly half.
• Frequency. Cut the number of times the event runs in a year. A quarterly event becomes a semi-annual one. A monthly event becomes a quarterly one. Your trade-spend on the event drops proportionally, and each remaining window gets cleaner incrementality because cannibalisation falls when events are further apart.
• Duration. Cut the length of each window. A four-week feature becomes a two-week feature. You lose some of the pantry-loading late-window sales, but most of those were forward-buying anyway.
• Mechanic. Switch a Buy One Get One (BOGO) to a straight price cut of equivalent per-unit value. Switch a straight price cut to a loyalty-card-funded price cut that only reaches registered shoppers. Each swap usually takes cost out of the event without killing the visibility.

Shrinking is less dramatic than ending, and it is the right answer on events where the window is load-bearing for some reason that is not captured in the ROI alone. Retailer contract commitments. Habit windows in seasonal categories. Trial dynamics for entry-tier packs. The five questions below are the sieve that separates the two.

Question 1: is the event's promo ROI really below cost of capital?

The starting point is the hardest to do honestly. Pull the incremental profit for the event for the last 12 and 24 months. Divide by the total trade spend the event consumes. Compare it against your corporate cost of capital, usually 8 to 12 percent for a mainstream CPG company. Events that sit clearly below that line across both windows are candidates for action. Events above the line are not the calendar problem, even if your finance business partner would like them to be smaller for a different reason.

The honesty is in the word "total". Most brand teams count the on-invoice slice and maybe the main off-invoice pool and stop there. A full accounting includes the slotting fees tied to getting the feature, the in-store display labour your team pays the retailer to execute the display, the bonus-pack stock you wrote off when the event over-ran, the temp-label run at the distribution centre, and the brand-management time consumed in running the programme. Add those in and the ROI of many events drops by 15 to 25 percent against the simple calculation most dashboards surface by default. If your event is now sitting below cost of capital on the full accounting but looked fine on the narrow accounting, your calendar has been lying to you.

Question 2: how much of the promo volume is cannibalised from your own baseline?

Incrementality is the other half of the story, and it is the question that usually decides whether an event gets killed or just shrunk. A 20 percent price cut that delivers a 60 percent volume lift looks good until you decompose the uplift. If 45 of those 60 extra units would have sold at full price anyway, and 10 came from forward-buying your next month's demand, only 5 are genuinely new sales. Your cannibalisation rate on that event is 75 percent and your effective incrementality is 8 percent on the promoted volume, not 60.

Above 70 percent cannibalisation, you are running a discount disguised as a promotion. The event is not finding new shoppers. It is selling your own baseline at a cheaper price and calling the difference marketing. Kill it. The short-term dip when you end the event will be small because most of the uplift was pulled from baseline anyway, and the reinvestment you get from freeing up the spend will usually produce better ROI somewhere else in the calendar.

Between 40 and 70 percent cannibalisation, there is real incrementality in the event but it is not earning its keep on its current shape. Shrinking is the right call. Cut the depth in half. Halve the frequency. The incrementality usually compresses less than proportionally because the shoppers who respond to a deeper promo with less promo-induced cannibalisation are more likely to be genuine switchers rather than baseline buyers stocking up. You keep most of the incrementality, shed most of the cost.

Below 40 percent cannibalisation is where genuinely good promotional events live. Those events are doing work you cannot easily replicate with other levers. Do not touch them unless something external forces the question.

Question 3: has reference-price decay set in?

The third question is the one most commercial teams skip because it requires an experiment, and experiments on the promo calendar get political fast. Reference-price decay is what happens when a frequent deep promo trains shoppers to anchor their expected "fair price" on the promo price instead of the regular shelf price. Over time, the regular price starts to look expensive even though it has not moved. When you pause the promo, the baseline does not return to the pre-promo level. It settles at a lower level that reflects the new anchor.²

The experiment is to pause one cycle of the event and watch the baseline in the window and the two windows after it. If baseline holds within 3 percent of the rolling average, decay is mild and a clean ending is safe. If baseline drops by more than 10 percent and does not recover within a cycle, decay is live and ending the event cold will cost you structural volume that is hard to win back.

Events with live reference-price decay should be shrunk across 2 to 4 cycles rather than ended in one. The depth comes down in steps, typically from 30 percent to 20 percent to 15 percent to 10 percent across a year. Shoppers re-anchor gradually, the baseline holds, and the trade-spend reduction lands without the structural volume loss. It is slower, but the annualised profit math beats the stepwise walk.

Question 4: is the event tied to a JBP commitment?

The Joint Business Plan (JBP) with the retailer usually contains commitments on feature participation, event-calendar slots, and volume-tier thresholds. Some of your promotional events exist because those commitments require them. Ending such an event unilaterally is not a question of profit. It is a question of whether you are willing to break a contract line and accept whatever the retailer does in response, which usually lands on unrelated Stock Keeping Units (SKUs) through shelf-space reductions, feature-rotation disadvantages, or retaliation on the next negotiation.

Events that are JBP-contractual should not be ended until the next JBP cycle, when you can negotiate the commitment out properly. In the meantime, they should be shrunk inside the commitment's wording. A "feature-share" commitment usually lets you run a shallower feature without breaching the contract. An event-calendar slot can usually be filled with a smaller mechanic. The JBP conversation at renewal is where the real end happens, and it is worth going into that conversation with data showing exactly how much retailer margin the current event was delivering without incremental sales, so you have a counter-offer ready.

Events that are purely brand-initiated have no such protection. They exist because somebody added them last year and nobody took them off. Those are the cleanest kills on the calendar.

Question 5: is the category structurally promo-dependent?

The last question is about the category and occasion the event sits in. Some categories are structurally built around promotional visibility.

Ice cream lives on summer windows. Beer rotates around public holidays. Confectionery runs a gifting calendar. Entry-tier packs in many categories use promo windows as the primary trial engine, and killing the promo collapses trial. In these cases, the event is earning its keep on dynamics that do not show up in the standard promo ROI calculation. Brand trial is a long-cycle payoff. Habit reinforcement is a retention effect that the incrementality decomposition understates.

Categories that are steady weekly-shop items with no strong seasonal or trial dependency, like household cleaners, personal care staples, or basic condiments, do not carry the same structural need. Events in those categories live or die on their direct promo ROI.

If your event sits in a structurally promo-dependent category and an occasion that matters, shrinking is safer than ending. Take the depth out but keep the window. If the event is in a category that does not have the dependency, the direct ROI math is the whole story and the ending path is clean.

Kraft Heinz 2019 to 2023: the portfolio they had to rebuild

The canonical case for the end-or-shrink discipline at scale is the promotional reset Kraft Heinz ran after the February 2019 writedown.

The writedown was $15.4 billion against the Kraft and Oscar Mayer brand portfolios, and the stock dropped 27 percent the day it was announced. The public diagnosis at the time was that a decade of cost-out under the prior ownership group had starved the brands of advertising support and that the company had over-indexed on deep promotional pricing to hold quarterly volumes, building a cannibalisation problem that eventually ate the equity.

Miguel Patricio, formerly Chief Marketing Officer at AB InBev, was appointed Chief Executive in June 2019. The public diagnosis from Kraft Heinz across 2019 and 2020 named the problem directly. The portfolio had been over-spending on deep, frequent promotions without enough measurement of what was actually working. Patricio committed publicly to cutting promotional intensity across the portfolio and reinvesting the savings into brand advertising and product renovation.¹

The rebuild was not a blanket edict. That is the part of the story most worth studying. Some promotional programmes were ended cleanly, typically low-incrementality BOGOs and deep-depth price cuts on brands where cannibalisation had been measured above 75 percent. Other events were shrunk rather than ended, particularly retailer-facing feature programmes where the window was load-bearing in the JBP but the depth was not.

The numbers the company reported publicly across the window tell the rest. Consolidated gross margin held up across the rebuild period despite category-wide input-cost inflation of 10 to 15 percent. Heinz, Philadelphia, and Kraft Macaroni & Cheese revenue trajectories stabilised through 2021 and 2022, as visible in the brand-by-brand splits the company reported in its annual filings. Unit volumes held within roughly 2 percent of the prior year on the flagship Stock Keeping Units even as net revenue grew in double digits, which is the signature of a portfolio where the promo discipline has taken hold. The price side of the bridge does more work. The promo side does less. The net lands higher.²

The reset was not free. Volume share softened during the rebuild in some categories, particularly where retailers retaliated on unrelated SKUs or where competitors filled the feature windows Kraft Heinz had vacated. The company accepted that cost. The alternative, which was to keep firing unmeasured promotional spend in defence of share, had already been tried and had led to the writedown.

The Kraft Heinz case is worth holding in mind for three reasons. Event-by-event is the right level. Kill lists and shrink lists are different lists. And the short-term pain is almost always smaller than the long-term cost of not doing the work.

The Monday-morning move

Pull the last two years of promo ROI by event on your top 10 Stock Keeping Units. Rank them by profit drag. Walk the five questions through the bottom quartile before the next calendar lock, starting with the events that are clearly negative on the full accounting.

For each event, write down which path it belongs on, and why. End the ones with above-70-percent cannibalisation, mild reference-price decay, no JBP tie, and no structural category dependency. Shrink the ones with the opposite profile, and be specific about which dial you are turning. Depth, frequency, duration, or mechanic, and by how much.

Cycle the reinvestment. About 30 to 40 percent of the reclaimed spend back into higher-ROI promotional activity on SKUs with measured incrementality. Another 30 percent into above-the-line brand advertising, which is what most underperforming calendars are implicitly underfunding. Hold the remaining 20 to 40 percent back to Finance as margin recovery, which is the part that gets the first calendar review across the line and builds credibility for the second.

What you should not do is run a blanket cut across the calendar in defence of a round-number trade-spend target. That is the move Kraft Heinz already retired in 2019. The portfolio view, done event by event, is the only one that survives the retailer conversation and the internal commercial pushback at the same time.

Keep going:

• Walk the decision with a 5-question rubric and a worked example → The paired playbook runs the same sequence as an interactive decision tree, with a full worked example on a Tier 2 chocolate-biscuit SKU where a quarterly BOGO is failing ROI by 16 percent and cannibalisation sits at 78 percent.
• On-invoice or off-invoice: where smart trade money actually goes in 2026 → Companion article on the other half of the JBP conversation. The routing decision that runs in parallel with the end-or-shrink decision for every trade dollar.
• Why BOGO is dying, and what works instead → Goes deeper on the specific mechanic that shows up on more of your kill list than any other, and what the alternatives look like when you need to keep the window and change the mechanic.

Take the ideas in this article further, inside RGM Academy's lessons:

• Promo ROI Fundamentals. The lesson that trains the muscle behind Question 1 of this article. You walk through the full-accounting promo ROI calculation with an AI coach, surfacing the hidden cost lines most dashboards miss and ranking events by profit drag. Why this matters for this article: every kill-or-shrink decision starts with the honest ROI number, and the honest number is usually 15 to 25 percent worse than the dashboard shows. → Start the lesson

• Source of Volume. The lesson behind Question 2 (cannibalisation rate). Teaches you how to decompose promotional uplift into its four sources: cannibalisation from your own baseline, forward-buying, competitor switching, and genuinely new shoppers. The cannibalisation rate is the single number that decides whether an event is a discount in disguise or a real demand generator. → Start the lesson

• Baseline versus Incremental. The lesson behind Question 3 (reference-price decay). Walks the pause-cycle experiment for detecting decay and teaches you how to model the stepwise shrink that lets you pull depth out across 2 to 4 cycles without collapsing baseline in categories where decay is live. → Start the lesson

• Promo Calendar Optimization. The capstone lesson for the end-or-shrink discipline. Builds the event-by-event ranked list across a full calendar, flags the retailer-contractual events that need a JBP conversation rather than a unilateral kill, and walks the reinvestment split across higher-ROI promotion and above-the-line brand spend. → Start the lesson

References

1. NielsenIQ, "How to measure trade promotion effectiveness," 2022 industry note. Direct quote: "over half of all trade promotions result in little to no sales lift, meaning manufacturers are ultimately wasting time and money." https://nielseniq.com/global/en/insights/analysis/2022/how-to-measure-trade-promotion-effectiveness/. Kraft Heinz public commentary on the promotional reset post-2019 is from the company's 10-K annual filings and earnings releases for fiscal years 2019 through 2023, available via Kraft Heinz investor relations at https://ir.kraftheinzcompany.com and SEC EDGAR at https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001637459.
2. Reference-price decay is covered in the marketing-science literature on internal reference-price formation and decay constants. For a canonical treatment, see Kalyanaram and Winer, "Empirical Generalizations from Reference Price Research," Marketing Science, Vol. 14, No. 3, Part 2, 1995, https://pubsonline.informs.org/doi/10.1287/mksc.14.3.G161. Kraft Heinz consolidated gross margin and unit-volume trajectories cited above are from the company's 10-K filings for fiscal years 2019 through 2022.