The BOGO trap: how one promotion wrecked a brand's year-2 forecast

Year 1 was a great year.

The brand had a strong product, growing distribution, and a smart promotional strategy. They ran buy-one-get-one deals at their major drug and mass retail partners — the kind of high-visibility programming that moves product off shelves and builds household penetration. Consumers responded. Sell-through looked strong. The retailer re-ordered. Everyone went into the annual planning cycle feeling confident.

The year-2 forecast reflected that confidence. Sales had been strong. The product was proven. The team modeled 25 percent growth on top of an already good year.

Twelve months later, they were trying to explain why revenue had fallen off a cliff.

What a BOGO actually does to demand

A buy-one-get-one promotion does exactly what it's designed to do: it gets consumers to buy more product in a single trip than they otherwise would.

A consumer who normally buys one bottle of moisturizer goes to CVS, sees the BOGO offer, and buys two. She didn't need two. She now has two. The extra bottle goes under the bathroom sink.

From the brand's perspective, that transaction looks like strong demand. Two units moved. Revenue recognized. Sell-through data shows the product flying off the shelf.

What it actually is: one purchase that satisfies two future purchase occasions. The consumer just bought herself out of the market for the next six to twelve months.

Multiply that by thousands of transactions across CVS, Walgreens, and Walmart, and you've just moved a significant chunk of year-2 demand into year 1. Consumers are stocked up. Their cabinets are full. They don't need to repurchase — not in month three, not in month six, maybe not until the following year.

This is pantry loading. And it's almost completely invisible in sell-in data.

The forecast mistake

The year-2 plan was built from year-1 shipments.

The team looked at what the brand had shipped to retailers in year 1 — strong numbers, driven by the BOGO lift — and extrapolated forward. If we shipped this much in year 1, and the brand is growing, surely year 2 should be higher. Twenty-five percent growth felt reasonable. Ambitious but achievable.

What the model didn't ask was: of all the product that shipped in year 1, how much of it has actually been consumed?

That's a different question from how much sold. Sell-in is what you shipped to a retailer. Sell-through is what consumers actually took off the shelf. And in year 1, a significant portion of the sell-through was driven by a BOGO that loaded consumers up with more product than they'd normally buy.

The 25 percent growth forecast assumed year-1 velocity was real and sustainable. It wasn't. It was inflated by a one-time promotional mechanic that pulled forward demand from the following year.

What happened in year 2

Retailers placed smaller orders. Then smaller again. By mid-year, some accounts had gone quiet entirely.

The team started working through explanations. The category was softening. Competition had increased. The marketing spend wasn't as effective as it had been. Each explanation was plausible. None of them was the actual cause.

Here's what was happening at shelf: consumers were still working through the extra bottles they'd bought during the BOGO. They hadn't repurchased because they didn't need to yet. Retailer shelves were still well-stocked from year-1 inventory that hadn't fully cleared. The retailers had no reason to reorder — their own inventory levels were healthy, even as the brand's shipments had dried up.

The demand signal the brand was watching — their own shipments to retailers — had gone quiet. But consumer demand hadn't disappeared. It had been borrowed. It was sitting under a bathroom sink somewhere, waiting to be used up.

The brand had shipped the product. The retailer had received it. The consumer had bought it. But the consumption hadn't happened yet. And until it did, there was nothing to replenish.

The gap between sell-in and sell-through

This is the core of the pantry loading problem, and it's one that catches brands at every stage.

Sell-in is the number most brands track because it's the number they control and the number that shows up in their revenue. When a retailer places an order and you ship it, that's revenue. It's real.

Sell-through is what consumers actually buy off the shelf. It's the number that determines whether the retailer will reorder.

In a healthy, steady-state business, these two numbers stay roughly in sync over time. Sell-in refills what sell-through depletes. The channel stays balanced.

A major BOGO promotion breaks that balance. Sell-through spikes artificially in year 1 as consumers pantry load. The retailer's shelves clear, which triggers reorders, which creates strong sell-in. Everything looks great.

In year 2, consumer sell-through normalizes — or more accurately, it drops below normal because the pantry-loaded supply hasn't been worked through yet. Retailer shelves stay full. Reorders stop. Sell-in craters.

The brand experiences the year-2 drop as a sudden change in the market. It isn't. It's the mathematical consequence of what happened in year 1.

What the data would have shown

The information needed to forecast this correctly existed. It just wasn't being looked at.

Retailer inventory data. Most major retailers will share warehouse inventory levels with their vendor partners, either directly or through paid data services. A brand that had pulled retailer inventory data at the end of year 1 would have seen elevated stock levels — product that had been bought by consumers but not yet consumed, or product that was still sitting in the retailer's DC. That's a direct signal that year-2 reorders would be slow.

Loyalty card data. Drug retailers like CVS and Walgreens have detailed purchase history through their loyalty programs. Aggregate purchase data — how often the average cardholder repurchases a given category of product — tells you the real consumption rate. If the average repeat purchase cycle for a moisturizer is five months and consumers bought two during a BOGO, the next purchase cycle is ten months out, not five. That math belongs in the year-2 forecast.

Category repeat rate benchmarks. Even without retailer-specific data, most categories have known repurchase rates. A moisturizer that lasts six months has a different demand profile than a serum used daily. Understanding the consumption cycle for your product and building that into your forecast — particularly in the year following a heavy promotional period — is the correction the model needed.

The rule of thumb the forecast ignored

Promotional lifts in year 1 are real. The units moved. The revenue happened.

But promotional lifts driven by pantry loading don't represent new demand — they represent borrowed demand. Year-1 volume that was inflated by a BOGO needs to be interpreted as: some of these units have created forward demand and some have replaced it. The forecast for year 2 needs to reflect both.

A simple adjustment: before applying any growth assumption to year-1 actuals, strip out the estimated promotional volume and ask what the underlying baseline was. The brand's non-promoted velocity — the months without BOGO programming — is a better starting point for year-2 planning than the promoted peak.

Then build the BOGO back in explicitly: if you're running the same programming again, model the same kind of spike. If you're not, don't assume the year-1 spike will repeat.

What the team should have done differently

Three things, none of which required new software or a larger planning team.

First, get retailer inventory data before finalizing the year-2 plan. One conversation with each major retail partner asking for their current inventory position on your SKUs would have surfaced the problem.

Second, model year-2 demand starting from underlying velocity, not from the promotional peak. Strip the BOGO lift out of year 1 before extrapolating.

Third, apply a basic consumption model. If your product is used once a day and lasts 60 days, a consumer who bought two during a BOGO has 120 days of supply. The next purchase occasion is four months out, not two. Build that lag into the repurchase assumption for the following year.

None of this is complicated in retrospect. It's the kind of thing that gets missed when the plan is built around sell-in data and revenue targets rather than consumer behavior and real demand.

The expensive part

The brand eventually found its floor, rebuilt from a clean baseline, and stabilized. But it took the better part of a year — a year spent explaining underperformance to retail partners, managing excess inventory, and rebuilding a forecast that should have been realistic from the start.

The BOGO promotion that drove the spike was a smart piece of marketing. It worked. It built household penetration. Done again with the right year-2 forecast in place, it would have been a clean win.

What made it expensive wasn't the promotion. It was the forecast that treated the promotional spike as the new normal.

If you're heading into a planning cycle after a heavy promotional year and want to make sure the year-2 forecast reflects what's actually happening in the channel, we're happy to take a look.