It sounds like a simple question.
It isn’t.
Inventory decisions are irreversible. Once you commit cash to product, you can’t unwind it the way you can pause ads or tweak pricing. That’s why inventory planning quietly separates amateur operators from disciplined ones.
Most businesses get this wrong not because they’re careless, but because they stop one layer too early.
This is the framework I use.
The most common answer to “how much should I order?” is:
“Enough for 90 days.”
That logic usually looks like this:
30 days to produce.
30 days to ship.
30 days of buffer for delays.
So you look at how much you sold in the last 30 days, multiply it by three, and place the order.
Some operators take this a step further and plan for days on hand on arrival. Instead of ordering 90 days of inventory, they order enough so that when the inventory arrives, they still have 90 days left—effectively planning 150 days in advance.
This is conservative. It’s also clean.
And it works—until demand changes.
Now introduce growth.
Assume you spent $100 per day on ads over the last 30 days. Next month, you plan to spend $200 per day.
If you simply multiply last month’s sales by three, you will under-order.
Your coverage math was built for a $100/day business. You are now trying to operate a $200/day business with the same inventory assumptions.
At this stage, many brands do the obvious thing: they multiply their order by the increase in spend.
That gets you closer.
But it still isn’t enough.
Because spend does not translate to sales in a straight line.
This is where inventory planning stops being mechanical and starts being strategic.
If you double your ad spend, do your unit sales double?
Sometimes yes. Often no. Occasionally, they grow by more than 2×.
That relationship is elasticity.
Elasticity explains how sensitive your unit sales are to changes in marketing spend.
An elasticity of 1.0 means sales scale perfectly with spend.
An elasticity of 0.6 means doubling spend only increases unit sales by 60%.
An elasticity above 1.0 can happen during seasonal demand shifts or strong product-market fit.
The only honest way to estimate elasticity is to look at historical data—prior spend increases, corresponding unit growth, and seasonal context.
If you’re early and don’t have clean data, this is where operator skill matters. You’re making an informed assumption, not pretending the future is predictable.
At this point, your forecast is no longer just “last 30 days × time.” It’s last 30 days × intent × response.
Even with spend and elasticity modeled, you are still making assumptions.
You’re assuming channels behave similarly, creative holds, conversion rates don’t collapse, and external conditions stay stable.
This is where most models lie to themselves.
The final lever I use is demand confidence.
Demand confidence answers one question:
How much of this forecast am I actually willing to fund?
A confidence of 1.0 means you fully believe the future you’re modeling.
A confidence of 0.8 means you believe it, but want margin for error.
A confidence of 0.6 means you think it might happen—but you’re not paying full price for the risk.
This isn’t vibes. It’s capital discipline.
You are explicitly discounting your own optimism.
Inventory forecasting isn’t about predicting the future.
It’s about deciding which version of the future you’re willing to pay for today.
That’s why I separate maintenance inventory (what I need if nothing changes) from growth inventory (what I need if my assumptions are right).
Blending those two is how businesses quietly overextend.
Separating them is how you stay solvent while still scaling.
Most brands fail at inventory not because they lack data, but because they collapse reality, intent, assumptions, and confidence into a single number.
Inventory deserves more respect than that.
When you treat it as a system—not a rule—you stop guessing and start making deliberate bets.
That’s the difference between running a store and operating a business.