# Refine Your Forecast

{% hint style="info" %}
**Quick Answer:** After generating a directional forecast, refine it using two approaches: **mass updates** for broad adjustments based on business rules, and **surgical edits** with Mo for targeted SKU-level changes. Most users start with mass updates, then fine-tune with surgical edits.
{% endhint %}

## Understanding Your Directional Forecast

When you generate a forecast in Moselle—whether top-down or bottom-up—the initial output is a **directional forecast**. This forecast provides a data-driven starting point based on historical patterns, trends, and the model you selected.

**Directional forecasts are:**

* Statistically grounded predictions based on your data
* Starting points that reflect general demand patterns
* Meant to be refined with your business knowledge

**Why refinement matters:** Your directional forecast doesn't yet include your institutional knowledge—upcoming promotions, product lifecycle changes, market shifts, or category-specific dynamics that only you understand. Refinement bridges the gap between statistical prediction and operational reality.

***

## Two Paths to Forecast Refinement

After reviewing your directional forecast, you have two complementary approaches:

| Approach           | Best For                                               | Method                       |
| ------------------ | ------------------------------------------------------ | ---------------------------- |
| **Mass Updates**   | Broad changes across many SKUs based on business rules | Mo Forecast Guidelines       |
| **Surgical Edits** | Targeted adjustments to specific SKUs                  | Conversational edits with Mo |

Most users follow this progression: start with mass updates to apply broad business logic, then use surgical edits for fine-tuning individual items.

***

## Mass Updates with Mo Forecast Guidelines

Mass updates are ideal when you need to apply consistent business rules across multiple products. Instead of editing SKUs one by one, you provide Mo with guidelines that describe how your products should behave.

### What Are Mo Forecast Guidelines?

Mo Forecast Guidelines are a set of business rules and product knowledge that you provide to Mo. These guidelines help Mo understand:

* How you think about different product categories
* Seasonal patterns specific to your business
* Promotional calendars and their expected impact
* Product lifecycle stages (launch, growth, maturity, decline)
* Category-specific demand drivers

### Creating Your Forecast Guidelines

{% stepper %}
{% step %}

### Review Your Directional Forecast

Start by examining your directional forecast to identify patterns that need adjustment:

1. Navigate to **Forecasting** in the left sidebar
2. Open your forecast scenario
3. Review the SKU-level projections
4. Note categories or product groups where the forecast doesn't align with your expectations

**Questions to consider:**

* Which product categories look too high or too low?
* Are seasonal patterns correctly reflected?
* Do new product launches have appropriate projections?
* Are promotional periods accounted for?
  {% endstep %}

{% step %}

### Document Your Business Rules

Capture how you think about demand for different parts of your catalog. Consider documenting:

**Category-level rules:**

* "Summer products should show 40% higher demand May through August"
* "Holiday gift sets peak in November with 3x normal volume"
* "Basics maintain steady demand year-round with minimal seasonality"

**Product lifecycle rules:**

* "New launches typically ramp to full velocity over 8 weeks"
* "Products in decline phase should forecast 15% lower than last year"

**Promotional rules:**

* "BFCM products see 2-3x lift during the promotional window"
* "Products featured in email campaigns see 25% lift for 2 weeks"

{% hint style="warning" %}
Focus on patterns and rules rather than individual SKU adjustments. Save specific SKU changes for surgical edits.
{% endhint %}
{% endstep %}

{% step %}

### Share Guidelines with Mo

Open the Mo chat interface and share your forecast guidelines. Mo will use these rules to suggest adjustments across your catalog.

**Example prompt:** "I've reviewed my directional forecast and want to apply these guidelines to my Spring 2026 forecast:

* Outdoor furniture should increase 50% from April through September
* Indoor furniture remains flat year-round
* New product launches should ramp to full demand over 6 weeks
* Products marked as discontinued should decrease 20% each month"

Mo will process your guidelines and propose mass updates that align with your business rules.
{% endstep %}

{% step %}

### Review and Apply Mass Updates

Mo will present the proposed changes based on your guidelines. Review the adjustments to ensure they match your intent:

1. Check sample SKUs from each affected category
2. Verify the magnitude of changes aligns with your expectations
3. Confirm the timing of seasonal adjustments
4. Approve the changes to apply them to your forecast

{% hint style="success" %}
Mass updates are applied across your catalog while preserving any individual item settings you've configured (launch dates, comparables, etc.).
{% endhint %}
{% endstep %}
{% endstepper %}

***

## Surgical Edits with Mo

After applying mass updates, use surgical edits for targeted adjustments to specific SKUs. This approach is ideal for:

* Individual products with unique circumstances
* Corrections to specific items after reviewing mass updates
* One-off adjustments based on specific market intelligence
* Testing different scenarios for key products

### How to Make Surgical Edits

{% stepper %}
{% step %}

### Identify SKUs Needing Adjustment

Review your forecast after mass updates to find items that need individual attention:

* SKUs where the forecast still doesn't match your expectations
* Products with unique circumstances not covered by your guidelines
* High-value items that warrant closer scrutiny
  {% endstep %}

{% step %}

### Use Natural Language with Mo

Open the Mo chat and describe the specific changes you need:

**Example prompts:**

* "Increase the forecast for SKU-12345 by 30% for Q2—we're running a major promotion"
* "Set the forecast for our new protein bar line to match the chocolate bar performance from last year's launch"
* "Reduce the forecast for winter accessories by 25% in March—we're exiting the category"

Mo will make the targeted adjustments while preserving your other forecast settings.
{% endstep %}

{% step %}

### Verify Changes

After each surgical edit, verify the change was applied correctly:

1. Check the updated forecast values for the affected SKUs
2. Confirm the timing of changes aligns with your request
3. Review any downstream impacts (inventory projections, reorder points)
   {% endstep %}
   {% endstepper %}

***

## Forecast Revision History (Coming Soon)

{% hint style="info" %}
**What's This?** Moselle automatically tracks every forecast change behind the scenes. While the UI for browsing revision history isn't available yet, your forecast edits are being captured with full audit trails—and we're building tools to make this history accessible to you.
{% endhint %}

### What's Being Tracked

Every time Mo makes bulk edits to your forecast, Moselle captures a complete revision snapshot:

**Change Details:**

* Before/after values for every modified cell
* Number of items and channels affected
* Total magnitude of changes across your forecast

**Context & Reasoning:**

* Your original prompt to Mo
* Mo's assumptions and reasoning for the changes
* Thread ID linking to your full conversation
* Timestamp and user information

### Upcoming Capabilities

We're actively building UI features to make this revision history accessible:

**Revision History Browser:**

* View a timeline of all changes to your forecast
* See who made changes and when
* Filter by date range, user, or change magnitude
* Click into any revision to see detailed before/after comparisons

**Change Comparison:**

* Side-by-side view of forecast values before and after each edit
* Visual diff highlighting exactly what changed
* Mo's reasoning displayed alongside the changes
* Export revision details for your records

**Revert Functionality:**

* Undo problematic bulk edits with one click
* Preview revert impact before applying

**Audit & Compliance:**

* Export complete audit trails for compliance reporting
* Share revision history with stakeholders
* Track forecast evolution over time
* Understand the full story of how your forecast was built

### Need Access Now?

If you have an urgent need to review forecast revision history or revert changes before the UI is available:

{% hint style="success" %}
**Contact your Moselle success team** and we can provide engineering support to:

* Generate detailed revision reports
* Review Mo's reasoning for specific changes
* Revert problematic edits
* Export audit trails for your analysis
  {% endhint %}

### Why This Matters

**Transparency:** See exactly what Mo changed and why, building trust in AI-assisted forecasting.

**Accountability:** Track who made changes and when, essential for team collaboration and compliance.

**Safety Net:** Know that you can always review or revert changes if something doesn't look right.

**Learning:** Understand how your prompts translate to forecast adjustments, helping you work more effectively with Mo over time.

***

## Recommended Workflow for New Users

If you're new to Moselle or working with a fresh forecast, follow this recommended workflow:

{% stepper %}
{% step %}

### Generate Your Directional Forecast

Create your initial forecast using top-down or bottom-up methods. This gives you a data-driven starting point.

{% content-ref url="/pages/XFbBg0MSQpAiqVkBJo75" %}
[How to Create a Forecast](/planning-and-execution/forecasting/how-to-create-a-forecast.md)
{% endcontent-ref %}
{% endstep %}

{% step %}

### Review and Identify Gaps

Examine the directional forecast and document what doesn't align with your business knowledge. Focus on:

* Categories that need adjustment
* Seasonal patterns that are missing
* Product lifecycle considerations
  {% endstep %}

{% step %}

### Build Your Mo Forecast Guidelines

Create a document capturing your business rules. This becomes your playbook for forecast refinement that you can reuse and refine over time.
{% endstep %}

{% step %}

### Apply Mass Updates

Share your guidelines with Mo and apply broad adjustments across your catalog.
{% endstep %}

{% step %}

### Make Surgical Edits

Fine-tune individual SKUs that need specific attention beyond the mass updates.
{% endstep %}

{% step %}

### Finalize Your Forecast

Review the complete forecast, make any final adjustments, and lock it when ready.

{% hint style="info" %}
You can lock your forecast to prevent further changes, including changes by Mo. This is useful when you've finalized planning and want to preserve the forecast for execution.
{% endhint %}
{% endstep %}
{% endstepper %}

***

## Frequently Asked Questions

### When should I use mass updates vs. surgical edits?

**Use mass updates when:**

* You need to apply consistent rules across many products
* You're adjusting entire categories or product groups
* You want to encode business logic that applies broadly

**Use surgical edits when:**

* You're adjusting specific individual SKUs
* The change is unique to one product or situation
* You're fine-tuning after mass updates

### Can I undo forecast changes?

Yes. You can regenerate your directional forecast at any time to start fresh, or use Mo to reverse specific changes. Moselle automatically tracks all forecast changes with detailed audit trails—see the [Forecast Revision History](#forecast-revision-history-coming-soon) section above to learn about upcoming features for browsing and reverting changes.

### How do I know if my forecast guidelines are working?

After applying mass updates, spot-check representative SKUs from each affected category. The changes should align with the rules you provided. Over time, track your forecast accuracy using the Forecast Performance Report to see if your guidelines improve predictions.

### What if Mo's suggestions don't match my expectations?

Clarify your guidelines with more specific instructions. Mo works best with clear, concrete rules rather than vague direction. For example, instead of "increase summer products," specify "increase outdoor furniture by 50% from April through September."

### How often should I refine my forecast?

Most users refine their forecast:

* **Monthly**: Review accuracy and make minor adjustments
* **Quarterly**: Update guidelines based on performance data
* **Seasonally**: Major updates before peak periods (BFCM, summer, etc.)
* **As needed**: When significant business changes occur

***

## Related Guides

{% content-ref url="/pages/XFbBg0MSQpAiqVkBJo75" %}
[How to Create a Forecast](/planning-and-execution/forecasting/how-to-create-a-forecast.md)
{% endcontent-ref %}

{% content-ref url="/pages/m7P5Z47OAkdJCWudsFQK" %}
[Customize a Forecast](/planning-and-execution/forecasting/customize-a-forecast.md)
{% endcontent-ref %}

{% content-ref url="/pages/uiyBUBi8r56D1ZueVTqo" %}
[Import and Export Forecast](/planning-and-execution/forecasting/import-and-export-forecast.md)
{% endcontent-ref %}

{% content-ref url="/pages/sb2SV5PNVIJy5jUlhar2" %}
[Meet Mo](/mo/meet-mo.md)
{% endcontent-ref %}


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://learn.moselle.io/planning-and-execution/forecasting/refine-your-forecast.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.