Demand Planning: Who Owns the Number

Your forecast says one thing, sales swears the launch will double it, and finance wants the number lower so the cash plan holds. Someone has to reconcile this to one number the company actually buys against. Welcome to demand planning.

Demand planning is the process of turning a demand forecast into an agreed, business-ready plan by layering in promotions, launches, and market knowledge. It's owned by the demand planner, who reconciles inputs from sales, marketing, and finance into one number the company uses to make their inventory buy.

Key takeaways

• The forecast is the start, not the plan: planning adds the context data can't see (promotions, launches, market shifts) and then gets agreement on the result.
• The output is consensus, not just a number: a plan nobody committed to is a forecast with a fancier name.
• One person owns the process, many feed it: the demand planner runs reconciliation; sales, marketing, finance, and supply all bring inputs.
• The hard part is people, not math: every role sees a different version of demand, and landing one number is a negotiation discipline.

What is demand planning?

Demand planning is the process of refining a statistical forecast with business judgment (promotions, launches, market shifts) into one plan teams commit to. It sits between the raw math and the buying decision, and its product is agreement: a single demand number per SKU per period that the whole business treats as the plan.

Forecast + business overlay = the plan

The statistical forecast knows the history; the business knows the future the history can't see. The overlay is where those meet: marketing's campaign calendar lifts specific SKUs in specific weeks, the sales team's new retail account adds a step change, finance's constraint trims the tail. Each adjustment should be explicit and owned, so when the plan misses, you can see which overlay was wrong rather than re-litigating the whole number.

How is demand planning different from demand forecasting?

Forecasting produces the baseline number from data; planning adjusts that number using context the data can't see, then gets agreement on it. The distinction sounds academic until a miss happens: forecast errors are data problems, planning errors are judgment problems, and consensus failures are process problems. Naming which one bit you is how you fix the right thing. (The forecasting discipline itself is owned by what is demand forecasting.)

• Input. Demand forecasting: Sales history + demand signals; Demand planning: The forecast + business knowledge
• Output. Demand forecasting: Baseline estimate; Demand planning: Agreed, business-ready plan
• Who. Demand forecasting: Analyst / model; Demand planning: Demand planner + cross-functional inputs
• Adds judgment?. Demand forecasting: Minimal, by design; Demand planning: Deliberately, and on the record

Who owns demand planning?

The demand planner owns the process, but the plan is cross-functional: sales, marketing, finance, and supply all feed it. Ownership means running the calendar, holding the standards for what counts as an input, and having the final call when functions disagree, not producing every number personally.

The demand planner role

In a growing brand this is rarely a dedicated hire at first; it's a hat worn by an ops lead, a founder, or an analyst. What defines the role isn't the title, it's three responsibilities:

• Own the baseline: keep the statistical forecast honest and current.
• Run the overlay: collect, challenge, and record the business adjustments.
• Land consensus: get one number agreed on a fixed calendar, and publish it as the plan.

Cross-functional inputs and consensus

Each function's input has a known bias worth naming out loud: sales trends optimistic (the pipeline always looks great), marketing overweights its own campaigns, finance anchors to the budget, supply prefers the number that keeps orders smooth. Consensus doesn't average the biases away; it exposes them, argues them against evidence, and lands a number each function will act on. That agreed number is exactly what the supply side then plans against (demand planning vs supply planning covers that handoff).

What makes demand planning hard?

The hard part isn't the math; it's reconciling people who each see a different version of demand and landing on one number. The math finishes in an afternoon. The consensus takes the rest of the month, every month, and it's the part that decides whether the plan means anything.

Competing inputs and how consensus resolves them

The working pattern is a fixed monthly rhythm: baseline forecast published, overlays submitted with reasons attached, one review where disagreements get argued against evidence, and a published plan with a change log. Two rules keep it honest. Adjustments must carry a reason ("sales feels good" is not an input), and the plan's misses get reviewed next cycle against who adjusted what, so the process learns. The step-by-step version of this cycle lives in the demand planning process, and the habits that keep it healthy in demand planning best practices.

Frequently asked questions

What's the difference between a demand planner and an inventory planner?

The demand planner owns the demand number: forecast plus business overlay, landed as consensus. The inventory planner converts that number into stock decisions: buys, targets, reorder triggers, balanced against cash. In small brands one person wears both hats; the questions still separate cleanly into "what will they buy?" and "so what do we stock?"

Does demand planning require a dedicated hire?

Not at first. Most brands run it as a hat on an ops lead or founder, with a monthly rhythm and a simple overlay sheet. The dedicated hire earns itself when SKU count, channel count, or the volume of promotions makes reconciliation a real workload, typically well into the growth stage.

What skills does a demand planner need?

Three clusters: enough analytics to keep a baseline forecast honest, enough business fluency to challenge sales and marketing inputs credibly, and enough facilitation skill to land consensus without owning every disagreement personally. The rarest skill is the third one. Tools can carry much of the first.

How does consensus demand planning work?

On a fixed cycle: the baseline forecast publishes, each function submits adjustments with reasons, a single review meeting argues the disagreements against evidence, and one agreed plan publishes with a change log. Next cycle, the plan's misses get reviewed against who adjusted what, which keeps the inputs honest over time.

What department does demand planning sit under?

It varies: operations or supply chain most often, sometimes finance, occasionally commercial. Where it sits matters less than what it can do: see all the inputs, challenge any of them, and publish a number every function accepts. Burying it inside one function tends to bias the plan toward that function's view.

Can AI handle demand planning, or just forecasting?

AI-powered platforms now carry more than the baseline: they maintain the forecast, flag where actuals diverge from plan, and quantify how past overlays performed. The consensus itself stays human, because it's an agreement between functions. The practical split: AI owns the number's starting point and the scorekeeping; people own the commitments.