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

# Clever AI Forecaster

> Simulate cashback campaign outcomes over 30, 60, or 90 days before launching: compare scenarios, estimate uplift, and commit budget with confidence.

The Clever AI What-If Forecaster is an intelligent planning tool built into your Hello Clever dashboard. This page explains what the Forecaster does, how to run both organic and cashback-driven forecasts, how to interpret the outputs you receive, and best practices for getting the most out of scenario planning before you launch a campaign.

The Forecaster uses machine learning trained on your historical sales, seasonal trends, and customer transaction behaviour to project where your revenue is heading, and how much a cashback campaign could move that trajectory.

## Getting started

<Steps>
  <Step title="Navigate to Forecaster">
    From your Hello Clever dashboard, go to **Site Overview → Forecaster**. You will land on a view showing your sales trend over the last 30 days and a prompt to begin forecasting.
  </Step>

  <Step title="Run an organic sales forecast">
    Click **Forecast** to project your baseline sales without any cashback incentive. Select a forecast duration:

    * **30 days**: best for tactical, near-term campaign planning
    * **60 days**: balanced view for medium-term decisions
    * **90 days**: strategic planning and quarterly marketing calendars

    The forecast result includes:

    * Forecasted sales value (for example, `$11,873.02`)
    * Percentage change over the previous comparable period
    * Confidence range (the upper and lower bounds of the projection)
    * Forecast reliability score (for example, `85%`)

    <Tip>
      The longer the forecast range, the broader the confidence interval. Use 60–90 day forecasts for strategic planning and 30-day forecasts for tactical campaign decisions.
    </Tip>

    **How the forecast is calculated:**

    Clever AI combines the following data sources to build your baseline projection using a trained time series model with variance bands:

    * Historical sales volume
    * Seasonal and weekday trends
    * Customer transaction behaviour
    * Cashback engagement data (where applicable)
  </Step>

  <Step title="Simulate a cashback campaign">
    Scroll to the **Want to boost your sales?** card. Configure your scenario:

    * Select a cashback percentage (for example, `10%` or `20%`)
    * Choose a campaign duration (`30`, `60`, or `90` days)
    * Click **Forecast**

    Clever AI recalculates your projection and overlays the cashback scenario on the organic forecast. You receive:

    * Forecasted sales with cashback applied
    * Revenue uplift over the organic baseline
    * Estimated cashback cost
    * Visual overlays comparing both outcomes

    **Example output:**

    |                  | Without cashback | With cashback |
    | ---------------- | ---------------- | ------------- |
    | Forecasted sales | \$11,873.02      | \$411,400.10  |
    | Uplift           | +95.9%           | +223.23%      |
    | Cashback cost    | \$0.00           | \$82,280.02   |

    The forecast graph displays:

    * Solid blue line: historical sales
    * Dotted blue line: forecast without cashback
    * Dotted green line: forecast with cashback
    * Shaded range: confidence interval
  </Step>

  <Step title="Test different scenarios">
    Scroll to **Forecast a different scenario** and adjust the cashback percentage (5–50%) and campaign duration, then click **Forecast** again. Clever AI instantly recalculates new forecasted sales, projected uplift, cashback cost, and confidence range for each scenario you test.

    Run multiple simulations side-by-side to find the cashback rate that delivers the best return for your budget.
  </Step>
</Steps>

## Interpreting forecast outputs

### Forecast reliability score

Each forecast includes a reliability score (for example, `85%`). This reflects the model’s confidence based on:

* **Data consistency**: how uniform your historical transaction patterns are
* **Pattern strength**: how clearly seasonal or weekday trends emerge in your data
* **External volatility**: exposure to holidays, economic shifts, or other external factors

A lower reliability score means wider confidence bands. In those cases, treat the forecast as a directional signal rather than a precise target.

### Sales uplift vs. cashback cost

Use the forecast to evaluate return on cashback spend before committing budget.

**Example ROI calculation:**

* Forecasted uplift: `$399,527.09`
* Cashback cost: `$82,280.02`
* ROI: **4.85× return on cashback spend**

<Note>
  Forecasts are estimates, not guarantees. Accuracy improves as Hello Clever captures more transactional history for your store. Stores that have recently onboarded may see higher variance in early forecasts.
</Note>

## Use cases

<CardGroup cols={2}>
  <Card title="Campaign planning" icon="calendar">
    Simulate before launching to validate ROI and avoid under- or over-investing in cashback.
  </Card>

  <Card title="Budget allocation" icon="chart-bar">
    Identify which weeks or product categories benefit most from cashback incentives.
  </Card>

  <Card title="Quarterly planning" icon="calendar-days">
    Use 90-day forecasts to inform your broader marketing calendar and spending cycles.
  </Card>

  <Card title="A/B scenario testing" icon="arrows-left-right">
    Run two simulations with different cashback percentages to find the optimal rate before launch.
  </Card>
</CardGroup>

## Best practices

* Start with 30-day forecasts to build familiarity with the tool and calibrate your expectations.
* For broader strategic decisions, run 60–90 day simulations to account for seasonality.
* Simulate multiple cashback percentages (for example, 5%, 10%, 20%) to locate the sweet spot between uplift and cost.
* Combine Forecaster outputs with [Clever AI Actionable Insights](/clever-ai/actionable-insights) to focus cashback on the customer segments or regions where impact will be greatest.
* Account for external events (holidays, competitor activity, or economic shifts) when interpreting results, as the model does not automatically factor in events outside your historical data.

## Forecast limitations

<Warning>
  Forecasts are statistical estimates based on your store’s historical data. They are not guarantees of future performance. Always apply business judgement alongside forecast outputs when making campaign or budget decisions.
</Warning>

* High variance may occur in low-volume stores or stores that have recently onboarded to Hello Clever.
* Accuracy improves over time as more transactional history is captured.
* The model does not automatically account for sudden external events such as major holidays, economic downturns, or significant competitor actions.


## Related topics

- [What is Hello Clever?](/getting-started/introduction.md)
- [Merchant Portal Overview](/portal/overview.md)
- [Clever AI Actionable Insights](/clever-ai/actionable-insights.md)