> For the complete documentation index, see [llms.txt](https://learn.moselle.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://learn.moselle.io/planning-and-execution/forecasting/phase-3-refine-and-maintain/setting-up-forecast-guidelines.md).

# Setting Up Forecast Guidelines

Learn how to write Mo Forecast Guidelines in Moselle: what to include, what to leave out, and a worked example of rules that improve your forecast.

{% hint style="info" %}
**Quick Answer:** Forecast guidelines are the things you know about your business that Mo can't see in your sales data — which history to trust, when your promotions run, and why some past months look wrong. Save them once in **Mo > ⋮ > Settings > Forecast Guidelines** and Mo applies them every time it builds or adjusts a forecast.
{% endhint %}

**Time Required:** 15–30 minutes for your first draft\
**Difficulty:** Intermediate

## What Are Forecast Guidelines?

**Forecast guidelines** are short, written rules Mo follows when it builds or adjusts your forecast. Mo already learns trend and seasonality from your sales history. Guidelines fill in what the history can't explain on its own, and Mo applies them across your entire catalog instead of you editing SKUs one by one.

Guidelines typically cover:

* **History to trust** — Which periods reflect how your business runs today
* **Recurring events** — When your promotions and seasonal events happen, and how big they are
* **Unusual periods** — Past dips or spikes, and what caused them
* **New products** — Only when a new SKU should take volume and seasonality from different products (a single comparable goes in [Item Comparables](https://learn.moselle.io/planning-and-execution/forecasting/phase-3-refine-and-maintain/special-situations/item-comparables.md))
* **Upcoming changes** — Launches, phase-outs and shifts your history can't predict
* **Sell-through** — Where Mo finds your retailer sell-through data and how to fold it into your sell-in forecast (see [Sell-In vs. Sell-Through](https://learn.moselle.io/planning-and-execution/forecasting/phase-1-understand-forecasting/sell-in-vs-sell-through.md))

{% hint style="warning" %}
**The one test for every rule:** *Does this tell Mo something it can't see in my data or my Moselle setup?* If the answer is no, leave it out — it adds noise without changing the forecast.
{% endhint %}

## Where Do I Save My Forecast Guidelines?

{% stepper %}
{% step %}

### Open Mo

Click **Ask Mo** in the left sidebar, or click the Mo chat widget in the top-right corner of any page.
{% endstep %}

{% step %}

### Open Settings

Click the three-dot menu **(⋮)** in the top-right corner of the Mo chat, then select **Settings**.
{% endstep %}

{% step %}

### Fill in Forecast Guidelines

In the **Mo Configuration** window, scroll past **Company Description** to **Forecast Guidelines**, enter your rules, and click **Save Settings**.

**Expected Result:** Every new forecast you create gets a guidelines pass from Mo. Mo changes only what a rule asks for, and tells you about any rule it couldn't apply with the data available.
{% endstep %}
{% endstepper %}

{% hint style="info" %}
**Guidelines are shared across your company.** Everyone on your team works from the same saved rules, and any change you save applies to all future forecast work with Mo.
{% endhint %}

### Saved guidelines vs. a quick tweak in chat

|                         | **Saved in Forecast Guidelines**                                                                      | **Asked in a Mo chat**                                                                               |
| ----------------------- | ----------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------- |
| **Where**               | Mo > ⋮ > Settings > Forecast Guidelines                                                               | Typed or attached in a single Mo conversation                                                        |
| **How long it applies** | Every forecast Mo builds or adjusts, until you change it                                              | That conversation only                                                                               |
| **Best for**            | Standing rules — history to trust, recurring events, unusual periods, new-product and phase-out plans | One-time or surgical changes — a single promotion, raising one SKU for one month, a quick correction |
| **Example**             | *"Prime Day, mid-July: about 4x normal daily sales on Amazon."*                                       | *"Add 500 units to SKU-410 in March for a retailer launch."*                                         |

***

## What Should I Include?

Good guidelines usually cover a few of these. Each one is knowledge only you have.

### 1. Which history to trust

Tell Mo which periods reflect how your business runs today.

| Example guideline                                                                                                      | What it tells Mo                                  |
| ---------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------- |
| *"Use sales from July 2023 onward. Before that, our order data was split across two stores."*                          | Where reliable history starts, and why            |
| *"Amazon grew sharply after we added Subscribe & Save last spring. Weight the last 6 months most heavily for Amazon."* | Which recent period best reflects a channel today |

### 2. Recurring events, with a date window and a lift

Give each promotion or seasonal event a **date window** and a **lift relative to normal sales**. A multiplier ("about 4x normal daily sales") stays meaningful year to year, even as your volume changes.

| Example guideline                                                                                          | What it tells Mo                       |
| ---------------------------------------------------------------------------------------------------------- | -------------------------------------- |
| *"Prime Day, mid-July (2 days): about 4x normal daily sales on Amazon. DTC is unaffected."*                | Event window, lift size, which channel |
| *"Anniversary sale, second week of September: about 2x normal sales on DTC. Next year it runs Sept 8–14."* | A date that moves each year            |
| *"National Pet Day, April 11: one-day email promo, about 1.5x normal DTC sales that day."*                 | A small one-day lift                   |

### 3. Unusual periods and their causes

Mo can see a dip or a spike, but not *why* it happened. Explain the ones that shouldn't shape your future forecast — and the ones that should.

| Example guideline                                                                                          | What it tells Mo                                |
| ---------------------------------------------------------------------------------------------------------- | ----------------------------------------------- |
| *"DTC sales were low for two weeks in March 2025 because our checkout was down, not because demand fell."* | A dip that isn't real demand                    |
| *"A creator video drove a 3-week spike on SKU-220 in August 2025. Don't carry it forward."*                | A one-off spike to leave out                    |
| *"Pet store reorders arrive quarterly as large orders. Those are normal demand, not spikes."*              | Large orders that are normal and should stay in |

### 4. New products with a split comparable

Set a single comparable for a new SKU in your catalog (see [Item Comparables](https://learn.moselle.io/planning-and-execution/forecasting/phase-3-refine-and-maintain/special-situations/item-comparables.md)). Use guidelines only when **volume** and **seasonality** should come from different products.

| Example guideline                                                                                                                      | What it tells Mo                                                              |
| -------------------------------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------- |
| *"New Travel Water Bottle (SKU-480): take volume from our Collapsible Bowl (SKU-410) and seasonality from our Cooling Mat (SKU-455)."* | Which product sets the size of demand, and which sets its shape over the year |

### 5. Changes the history won't show

Things you know are coming that past sales can't predict.

| Example guideline                                                                                          | What it tells Mo                              |
| ---------------------------------------------------------------------------------------------------------- | --------------------------------------------- |
| *"We're raising prices on all leashes by about 10% in May. Expect unit sales to dip for the first month."* | A change that will move demand                |
| *"We're launching on Chewy in March. Expect it to add demand, not shift it from Amazon."*                  | Whether a new channel adds demand or moves it |

### 6. Where your retailer sell-through data lives

If you sell wholesale, tell Mo where to find retailer sell-through and how to use it. See [Sell-In vs. Sell-Through](https://learn.moselle.io/planning-and-execution/forecasting/phase-1-understand-forecasting/sell-in-vs-sell-through.md) for the full setup.

***

## Template

Copy this into **Forecast Guidelines** and fill in what applies. Delete any section you don't need.

```
# History
- [Which period of sales best reflects your business today, if any, and why
  earlier data is less reliable]
- [Anything about growth or recent changes that should change how history is
  weighted — e.g. a channel that grew quickly, a new fulfillment setup]

# Recurring events
- [Event], [start date] – [end date]: about [X]x normal sales on [channel].
- [Events whose dates move each year, and their next dates]

# Unusual periods
- [Why demand was low or high in a past period — e.g. a stockout, outage,
  viral moment, one-time order — and how Mo should treat it: ignore it,
  keep it as normal demand, or give it less weight]

# Upcoming changes
- [Shifts you know are coming that your history can't show — e.g. a new
  channel, a price change, a big retailer launch — with dates, and how they
  should change the forecast]
```

{% hint style="success" %}
**Tip:** Not sure how big a lift was? Ask Mo: *"How much did Amazon sales lift on Prime Day last year compared with the two weeks before?"* Then put the answer in your guidelines.
{% endhint %}

***

## Frequently Asked Questions

### How many rules should I start with?

Start with the few that matter most — usually your biggest events and any periods that distort your history. Five to ten lines is plenty. Add more as you review forecasts.

### Can I change my guidelines later?

Yes. Edit **Forecast Guidelines** in Mo Settings and click **Save Settings**. Mo uses the new version for every forecast from then on. To apply the change to a forecast you've already built, ask Mo to re-apply your guidelines to it.

### What happens if Mo can't apply a rule?

Mo skips it and tells you why — for example, if the rule refers to a channel that isn't in the forecast. Check Mo's summary after each forecast is built.

### What shouldn't go in my guidelines?

Some things people write into guidelines already have a place in Moselle. Set them there instead, so they stay accurate and visible to your team.

| If you want to tell Mo…                                  | Set it here instead                                                                                                                                                                          |
| -------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| When a new product starts selling                        | **Launch Date** on the item's **Adjust Forecast** tab (see [Product Setup](https://learn.moselle.io/operations/catalog/product-setup.md))                                                    |
| When a product is being discontinued                     | **Phase Out** on the item details page (see [Product Setup](https://learn.moselle.io/operations/catalog/product-setup.md))                                                                   |
| Which existing product a new SKU should be forecast like | **Product Comparables** on the item (see [Item Comparables](https://learn.moselle.io/planning-and-execution/forecasting/phase-3-refine-and-maintain/special-situations/item-comparables.md)) |
| How bundles or multipacks count toward demand            | [Bill of Materials](https://learn.moselle.io/operations/catalog/product-setup/bill-of-materials.md)                                                                                          |
| Revenue goals for a channel or month                     | Revenue targets in a [Top-Down Forecast](https://learn.moselle.io/planning-and-execution/forecasting/phase-2-build-your-forecast/how-to-create-a-forecast/generate-a-top-down-forecast.md)   |
| Supplier lead times                                      | The supplier's settings (see [Add a Supplier](https://learn.moselle.io/operations/suppliers/add-a-supplier.md))                                                                              |

### How do I know if my guidelines improved the forecast?

Compare forecast accuracy before and after in **Reports > Forecast Performance**. See [What is MAPE?](https://learn.moselle.io/faq/forecasting/what-is-mape-and-good-mape-score.md) for what a good score looks like.

***

## Related Guides

{% content-ref url="/pages/UUTQWdGmVeuEJedst19p" %}
[Item Comparables](https://learn.moselle.io/planning-and-execution/forecasting/phase-3-refine-and-maintain/special-situations/item-comparables.md)
{% endcontent-ref %}

{% content-ref url="/pages/9zQ3PsTwvdrNQhe08C8X" %}
[Sell-In vs. Sell-Through](https://learn.moselle.io/planning-and-execution/forecasting/phase-1-understand-forecasting/sell-in-vs-sell-through.md)
{% endcontent-ref %}

{% content-ref url="/pages/Ig32ZSeUite8Tj50RFbE" %}
[Refine Your Forecast](https://learn.moselle.io/planning-and-execution/forecasting/phase-3-refine-and-maintain/refine-your-forecast.md)
{% endcontent-ref %}

{% content-ref url="/pages/1dxratABFxJY8eSDfvto" %}
[How Mo's Forecasting Works](https://learn.moselle.io/planning-and-execution/forecasting/phase-1-understand-forecasting/how-mos-forecasting-works.md)
{% endcontent-ref %}
