Inventory Is Short. Which Channel Gets It?

Here is a problem worth having, and it is the one on my desk in August 2026. Every channel is growing. The website, the marketplaces, the retail account. And there is not enough stock to serve them all. This matters because the meeting that follows is almost always the wrong meeting: everyone argues about who gets the units, when the actual question is which channel you are trying to win in. Here is the framework I use, what the public filings say about where consumer brands really sell, and what to watch once you have made the call.

The question is not "who gets the units"

A live example, with the brand kept out of it. Four channels: an owned store, two marketplaces, and a large retail account. Two hero SKUs are short. Demand is running around 4,000 units a month against no restock scheduled for the month. There are roughly 3,500 units of the best-selling SKU sitting in the warehouse, and every one of them is allocated to the owned store. The marketplaces got zero.

Why? Because the priority order says owned store first, then marketplace, then the second marketplace. And that order is defensible. The owned store moves about 60 units a day of that SKU, so it burns through inventory fast and visibly.

But here is what had changed. The retail account used to be served by its own dedicated production run, so it never appeared on the priority list at all. That dedicated stock ran out and its replacement is late. For the first time, retail is competing for the same pool. Nobody reopened the priority order, because on paper there was no decision to make. The rule just kept running.

That is the real failure mode, and it is almost never described as a strategy failure. The pattern we see again and again is a channel priority that was set 12 months ago, under supply conditions that no longer exist, and has been inherited since as a rule rather than a choice.

The moment supply is short, allocation stops being arithmetic. Whichever channel gets the units is the channel you have decided to win in. You are making that decision either way. The only question is whether you make it on purpose.

What Playing to Win actually says

The cleanest framework for this is Playing to Win, written by A.G. Lafley and Roger Martin. Lafley ran Procter and Gamble twice. Martin was Dean of the Rotman School. The book exists because P&G, with an enormous house of brands, had to decide which categories, which shoppers and which channels it was actually going to win in, and stop pretending it could win everywhere.

The framework is five linked choices:

The choice	The question it answers	What it looks like for a channel decision
Winning aspiration	What does winning mean for us?	Are we building enterprise value for a sale, or maximising cash now? These point at different channels.
Where to play	Which customers, categories and channels?	Owned site, marketplace, mass retail, specialty. Pick. You cannot fund all four.
How to win	What makes us win there specifically?	Availability and shelf position win in retail. Assortment and brand story win on an owned site.
Capabilities	What must we be great at?	Retail needs forecast accuracy and fill rate. Owned site needs creative and retention.
Management systems	What keeps it running?	The allocation rule itself, and a standing review of it.

The middle two carry the weight, and Martin is specific that they are a matched pair. Writing in Harvard Business Review, he argues these choices cannot be made sequentially or in isolation: no how-to-win choice can be made outside the context of a given where-to-play choice. You do not pick a channel and then work out how to win in it. You pick the pair, because a channel you cannot win in is not a channel you should be in.

Applied to a shortage, the cascade does something useful and slightly uncomfortable. It forces you to say which channel you are conceding.

The replenishment test

The cascade tells you to choose where to play. It does not tell you how to know. Here is the test I use first, and it takes about a minute.

Ask what your customer does on the morning they run out.

If they need the product that day, and will buy a competitor's version off whatever shelf is nearest rather than wait, you are in a replenishment category. Hair product, deodorant, toothpaste, coffee, diapers, pet food. In a replenishment category your own website is structurally capped, because shipping time is the binding constraint. No amount of paid social fixes a two-day delivery window against a customer who needs it before work. The category will always route most of its volume through places that solve immediacy: shelves and marketplaces with fast fulfillment.

If your customer will happily wait four days for the specific item they wanted, you are in a considered purchase. Skincare regimens, supplements on subscription, apparel, specialty tools. Here an owned site can hold real share, because the customer is choosing you, not choosing what is in front of them.

This one question resolves most channel arguments, because it converts "where should we win" from an opinion into something you can observe. And it explains why so much DTC advice ages badly when it crosses categories. The advice was written by someone in a considered-purchase category and read by someone in a replenishment category.

When I talk to founders in this spot, the instinct is always to protect the channel they can see in real time. The owned store has a live dashboard. The retail account reports weeks later through a portal nobody enjoys opening. Visibility is not the same as value.

What the filings actually say

I did not want to argue this from instinct, so I went to the disclosures of the two largest publicly traded hair and personal care pure-plays and read how they describe their own channels.

The Honest Company is the sharper case, because it started as a direct-to-consumer brand. Its FY2025 Form 10-K states plainly that "effective December 31, 2025, we have transitioned away from Honest.com as a shipping and fulfillment channel, while maintaining Honest.com as a resource for educating consumers, showcasing our complete product portfolio, and driving consumers to purchase through our leading retailers and their websites, and third-party ecommerce sites."

They kept the website. They stopped shipping from it. The site is now a marketing asset that routes buyers to other people's checkouts.

The concentration that replaced it is stark. The same filing discloses that in 2025, Amazon, Target and Walmart accounted for approximately 39%, 28% and 8% of total revenue. Three partners, three quarters of the business. FY2025 revenue was $371.3 million, down 1.9% from $378.3 million the year before, so this is not a story about a channel exit fixing growth. It is a story about a company deciding where it can actually win and stopping the pretence elsewhere. We covered the earlier stage of that shift in our Honest Company teardown.

Olaplex is the more interesting read, because of what it does not tell you.

Its FY2025 net sales split three ways with no dominant channel: professional (salons) $153.3 million, specialty retail $130.4 million, and DTC $139.3 million, on $423.0 million total. A third of the business in DTC looks like a genuine counterexample until you read the definition. The filing describes selling "through professional distributors in salons, directly to retailers for sale in their physical stores and e-commerce sites, and DTC through sales to third party e-commerce customers and through our Olaplex.com websites."

Third-party e-commerce customers sit inside the DTC line. Marketplace revenue is in there, and the company never breaks out what its own website does.

Neither of the two biggest public hair care pure-plays can show you a large owned-site business. One shut the channel down. The other folds it into a number that includes marketplaces and declines to separate it. If you are benchmarking your Shopify store against a public company's "DTC" line, you are comparing yourself to something bigger and different.

That is a negative result, and negative results are useful. If you believe your category wins on your own site, you should be able to point at somebody doing it at scale. In this category, from the filings, you cannot.

Exhaust the supply moves before you cannibalize anything

None of the above means you start starving channels this week. Before you take a single unit from anywhere, there is a sequence that usually creates more inventory than the allocation fight was going to save.

1. Ask the supplier to ship the existing order sooner. Plainly, with no negotiation attached. This works more often than it should.
2. Ask for partial, rolling fulfillment. Most suppliers quote a full purchase order on a three-month lead time. Many will ship 10% now and the balance weekly if you ask, because it smooths their production too. This is the request that frees up the most units.
3. Steal capacity from your own SKUs that have slack. You have products with comfortable cover. Push their production back and pull the constrained SKU forward. You are reallocating your own manufacturing line, not asking anyone for a favour.
4. Pay for timing. If a 2% price bump moves the delivery date, take it. Compare it against the margin on the units you would otherwise not sell, and against a retail partner's opinion of you.
5. Check the SKU is actually fungible. Confirm the box and barcode are identical across channels before any of this. If retail requires different packaging, the units cannot move and the whole debate is theoretical.

One more thing on the retail side, which is not about inventory at all. If you are going to miss, tell the partner before they find out. A retail buyer treats a surprise short shipment as a reliability failure, and reliability failures cost you shelf space that takes a very long time to win back. A stockout on your own site costs you a sale. A stockout at a major retailer can cost you the account.

This is how we advise clients

I am not quoting Playing to Win because it is a famous book. It is the book we hand clients, and the cascade is the structure we run the conversation through whenever a brand has to decide where the next constrained thing goes, whether that is inventory, headcount or ad budget. It works because it is uncomfortable in exactly the right place. You cannot answer "where will we play" honestly without naming what you are giving up.

The version we run in a working session is four questions, in this order:

1. What does winning look like in the next 24 months? Building enterprise value for a sale and maximising cash this year are different answers, and they point at different channels.
2. Which channel are we trying to win in, and which ones are we willing to serve with whatever is left? The second half of that sentence is the part nobody says out loud, and it is the part that actually allocates the units.
3. What do we have to be good at to win there? Fill rate and forecast accuracy in retail. Creative and retention on an owned site. If we are not good at it, and we are not funding getting good at it, we have not chosen that channel. We are hoping in it.
4. What would have to change for this answer to change, and when do we look again?

Twenty minutes of that beats a week of arguing about units, because it moves the decision up to the level it actually belongs at. If you want the source, it is Playing to Win by Lafley and Martin, and Martin keeps a summary of the cascade on his own site. This is the work we do as a client's fractional CFO team, and a channel call under a shortage is one of the more common reasons a founder picks up the phone.

Make the call, then write down what would change it

Once you have chosen, record two things: the choice, and the conditions that produced it.

The conditions are the part everyone skips, and skipping them is what created the stale priority order in the first place. "Owned store first" was correct when the retail account had dedicated supply. It became wrong the moment that stopped being true, and it stayed in force because it had been written down as a rule with no expiry attached.

So write it as a conditional. Something like: retail gets first call on this SKU while its dedicated production is delayed, reverting when the delayed purchase order lands. Then put the review on the calendar. When we have struggled with this ourselves, the mistake was treating it as an allocation problem when it was a strategy problem, and the tell was that nobody could say what would make us change our minds.

Two practical notes to finish. First, sort your channels by contribution margin per unit, not revenue. The channel with the biggest top line is frequently the one with the worst unit economics once you load in acquisition cost, and a shortage is the one moment when that difference is unavoidable. If you have never built that number properly channel by channel, our multi-channel profitability playbook has the method. Second, before you decide anything off a spreadsheet, check the spreadsheet is calculated rather than typed. The allocation sheet in the example above was entirely hand-entered. That is a calculation error waiting to become a channel strategy, and it is the same discipline problem we wrote about in diagnosing a rising CAC.

Sources and methodology

Both company disclosures were read from the filings themselves on SEC EDGAR, not from press coverage of them. Each 10-K was downloaded and text-extracted, and every quoted sentence in this article appears verbatim in the source document. Channel figures were cross-checked against the two places they appear in the Olaplex filing (the management discussion and the revenue note), which agree.

The Honest Company figures come from its FY2025 Form 10-K. The filing carries the December 31, 2025 exit from Honest.com as a shipping and fulfillment channel, the retailer concentration of approximately 39% Amazon, 28% Target and 8% Walmart, and revenue of $371.3 million against $378.3 million in the prior year.

One correction worth flagging for anyone researching this company. Several secondary summaries in circulation report Honest's FY2025 revenue as $378.3 million and describe it as roughly 10% growth. Both are wrong: $378.3 million is the FY2024 figure, and FY2025 revenue declined 1.9%. This is a good argument for reading the filing rather than a summary of it.

The Olaplex figures come from its FY2025 Form 10-K. The filing reports net sales of $153.3 million professional, $130.4 million specialty retail and $139.3 million DTC on $423.0 million total, and defines the DTC channel as sales to third-party e-commerce customers plus its own websites.

The strategy framework is Lafley and Martin's, cited from Martin's own publications rather than a summary. The book is Playing to Win: How Strategy Really Works, Lafley and Martin, Harvard Business Review Press. The five-choice cascade is set out on Roger Martin's site, and the specific point that where-to-play and how-to-win must be chosen together rather than in sequence comes from Martin writing in Harvard Business Review.

The operator example is a live client situation, anonymized. No brand, retailer or individual is identified. Unit volumes, daily sell-through and channel counts are real figures from the engagement; the identifying details are not included.

Frequently asked questions

how do i decide which sales channel gets inventory when i am short?

Decide where you intend to win before you split a single unit. Rank channels by contribution margin per unit and by how recoverable a stockout is in each one, then protect the channel that scores highest on both. Revenue is the wrong sort key, because your biggest-revenue channel is often your worst-margin one.

what is the playing to win framework in plain english?

Five linked choices from A.G. Lafley and Roger Martin: what winning looks like, where you will play, how you will win there, what capabilities that needs, and what systems keep it running. The middle two do the work. Strategy is choosing where to compete and how, which means choosing where not to compete.

should i short ship a big retailer to protect my own website?

Usually no. A retail partner treats repeated short shipments as a reliability failure and reduces your space, which is slow to win back and sometimes permanent. Your own site is the one channel where a stockout costs you a sale rather than the relationship. If you must miss, tell the retailer before they discover it.

how do i know if my category is a replenishment purchase?

Ask what your customer does the morning they run out. If they need it that day and will buy whatever is on the shelf, you are in a replenishment category and your own website is capped by shipping time. If they will wait several days for the specific item, it is a considered purchase and your own site can hold real share.

is dtc or retail better for a physical product brand?

It depends on the category, not on the business model. Two of the largest public hair and personal care pure-plays both sell mostly through other people's shelves. One shut its own website as a fulfillment channel entirely at the end of 2025. That is category evidence, not a universal rule, but it should make you check your assumptions.

what should i ask my supplier when i am going to run out?

Four things, in order: can you ship the existing order sooner, can you ship it in partial rolling batches instead of one drop, can you move production capacity from one of my SKUs that has slack, and what would it cost to expedite. Ask about the price last. Timing is usually worth a small premium.

do public companies actually report their own-website revenue?

Often not. Olaplex reports a DTC channel but defines it as third-party e-commerce customers plus its own sites, so marketplace revenue is inside that number. If you are benchmarking your Shopify store against a public company's DTC line, you are comparing against a bigger and different thing.

how often should i revisit a channel priority decision?

Whenever the conditions that produced it change, and at minimum every quarter. The most common failure we see is a priority order set when one channel had dedicated supply, then inherited as a rule long after that stopped being true.