> ## Documentation Index
> Fetch the complete documentation index at: https://docs.spherecast.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# A day in the life: Demand planner

> Follow Maria through a full demand-planning day, from the morning forecast run to pushing consensus down to supply.

Meet **Maria** — a demand planner for a mid-sized consumer-goods company. She owns the forecast: the number the whole supply chain plans against. Here's how a typical Tuesday goes.

## Morning: what happened overnight

Maria starts in her **Inbox** — the feed of tasks and notifications in the left sidebar. Two items are waiting under **Today**: **"New forecast run is ready"** and **"Review forecast validation results."** Good. The scheduled statistical run finished on time, so she has fresh numbers to work with.

Before she dives in, she glances at the **Cockpit**, her health dashboard. She isn't looking at the three KPI cards this morning — she's scanning the **Exceptions** queue for two types that are hers: **Sales deviation** (actual sales drifting away from the forecast) and **Demand exception** (demand behaving oddly). A couple of Sales deviation rows catch her eye — a beverage SKU (Stock Keeping Unit, one sellable item) sold far above plan last week. She notes it and moves on. More on the Cockpit in [Exceptions](/guide/home/exceptions).

## Reviewing the baseline run

Next she opens the **[Baseline forecast](/guide/demand-planning/baseline-forecast)** — the statistical starting point Spherecast projects from sales history. The run comes with a **"Products to review"** list: the items whose recent sales deviated enough that the model flagged them. That beverage SKU from the Cockpit is right at the top, alongside a snack item that dropped off a cliff.

Maria works the list. For the products that genuinely spiked or dropped — not just noise — she **re-runs the baseline** so the model incorporates the new reality instead of anchoring to a stale pattern. For a couple of others, she decides the blip was a one-off promotion last year and leaves the baseline as-is. This is judgment work, and it's why a human sits in the loop.

## Midday: into the S\&OP grid

With the baseline cleaned up, Maria opens the **S\&OP (Sales & Operations Planning) grid** — her main planning canvas. First she narrows it down: she filters to **her market and her category** so she's only looking at products she owns. (Markets and Channels are terms your company may have renamed.) See [Filter and display](/guide/demand-planning/filter-and-display).

Now she reads the metric ladder across each row:

* **Target** — the top-down goal from leadership.
* **Baseline** — the statistical forecast she just reviewed.
* **Baseline Adj.** — the Baseline plus any included S\&OP adjustment lines. **This one is derived — Maria never types it.** It moves only when she changes the adjustment lines feeding it.
* **Consensus** — the final number everyone agrees to plan on.

She eyeballs the **Forecast Gap** — the distance between Consensus and Target. Where the gap is wide, leadership's goal and the bottom-up plan disagree, and that's where she spends her attention.

## A request from Marketing

Mid-morning, Marketing pings her: they're running a **promotion on a whole product family next quarter** and expect a real lift. Maria doesn't overwrite any cells by hand. Instead she models the uplift as an **[adjustment](/guide/demand-planning/adjustments) line** — a deliberate, documented change on top of the baseline — and she attaches a **reason** ("Q3 promo, Marketing request") so anyone reviewing later understands why the number moved. Because the promotion has a seasonal shape rather than a flat bump, she reaches for a **[demand profile](/demand-profiles)** to spread the lift across the weeks the way the campaign will actually land.

Before she commits, she wants to see what this does downstream. She spins up a quick **[scenario](/guide/scenarios/overview)** — a safe what-if copy of the plan — bumps the demand there, and previews the supply impact without touching anything live. It confirms the promo is orderable in time. Reassured, she comes back to the real plan.

## Afternoon: consensus and push-down

Her teammates have been busy too. In the adjustments sidebar, Maria reviews the changes they've **proposed** — each with its own required reason — and works through them one by one, **Accepting** the sound ones and **Rejecting** the two that double-count a promotion she's already modeled. This back-and-forth is the collaborative heart of the process; see [Building consensus](/guide/demand-planning/building-consensus).

With adjustments settled, she sets **Consensus** across her products — the number the business will actually plan on. Then she **pushes it down**, handing the agreed forecast to supply planning and kicking off a fresh supply re-run so the buyers see the new demand.

Last thing before she logs off: she takes a **[snapshot](/guide/demand-planning/forecast-accuracy)**, freezing today's plan so that months from now she can measure how close her forecast came to reality. Accuracy tracking is how she gets better every cycle.

That's Maria's day — a mix of statistics, judgment, and negotiation, all pointed at one honest number.

> **Tip:** New to the vocabulary? Start with [Key concepts](/getting-started/key-concepts), then see the [Demand planning overview](/guide/demand-planning/overview).