Skip to main content

Reading your data

Interpreting the Data Table​

The Forecast table displays the lifecycle of your cohorts. Depending on how much time has passed since a user installed your app, the numbers you see will represent different states of data completeness.

  • Forecasted Metric Selector: Use the dropdown at the top left of the chart to toggle which metric you are currently analyzing.
  • Supported Forecast Metrics: The machine learning model currently supports forecasting the following key performance indicators. Note: With the exception of Retention Rate, all metrics are calculated cumulatively, meaning their values are summed up and grow as the cohort progresses through milestone days (e.g., from D3 to D7).
    • ARPI (Average Revenue Per Install): The cumulative average revenue generated per user.
    • Payout: The cumulative payout generated by the cohort.
    • Retention Rate (Non-cumulative): Forecasts the percentage of users who will continue to open the app on specific future milestone days.
    • Revenue: The total cumulative revenue generated by the cohort.
    • ROAS (Return on Ad Spend): Forecasts the cumulative gross revenue generated per dollar spent on advertising.
    • ROI (Return on Investment): Forecasts the cumulative net profitability of your campaigns over time.
  • Actuals: These are real, recorded metrics based on actual spend and revenue/events.
  • Forecasts: These are the machine-learning based estimations for future milestone days.

Understanding Matured vs. Partial Actuals​

Because users in a selected date range install your app on different days, a cohort takes time to "mature." You can check the exact maturity status of your data in two places:

  1. In the Chart: Hover over any data point on the Forecast Trajectory chart to view maturity details in the tooltip.
  2. In the Data Table: Look directly beneath the daily metric columns (e.g., D3, D7) to see the exact completion percentage (e.g., "76% matured").
  • Mature Actuals: This state is only reached when 100% of the users in your selected date range have lived past a specific milestone day.
  • Partial Actuals: You may notice that recent milestone days (e.g., D14) show a lower Actual percentage than expected. This happens when only a portion (less than 100%) of your selected users have reached that specific day. The system will rely on Forecasts to fill in the rest of this data.

Forecast Confidence Levels​

Not all forecasts are created equal. Brand-new campaigns have less historical data to draw from than mature campaigns. To help you understand how much trust to place in a specific forecast, the system assigns a Confidence Level to every forecasted metric.

You can view these confidence indicators in two places:

  1. On the Forecast Trajectory Chart: The data points plotted along the dotted forecast curve visually change based on their confidence level.
  2. In the Data Table: A confidence circle icon appears directly next to the forecasted ("FCST") percentages in your campaign rows.

What the Confidence Icons Mean:​

  • High Confidence (⬤): Based on stable, consistent data from this campaign. The forecast should stay close to this number as more data arrives.
  • Medium Confidence (◐): Based on limited data for this specific slice, or drawn from a broader view across the campaign. Expect some movement as the campaign builds up more data.
  • Low Confidence (◯): The data so far is sparse or volatile, so the forecast covers a wide range of outcomes. Expect larger swings until the campaign matures.

Setting Your Targets​

You can visualize your break-even goals directly on the chart and table.

  • Click Edit targets on the right side of any campaign row to input your specific metric goals for various milestone days.
  • These targets will appear as diamond markers on the Forecast Trajectory chart, allowing you to visually see exactly when a campaign is estimated to hit your profitability goals.

Why do I see a dash (—) instead of a forecast?​

Occasionally, you may see a dash (—) in the data table instead of a percentage. This indicates that the system does not have the required data baseline to calculate or forecast that specific metric.

For return-based metrics like ROAS and ROI, this typically happens when there is zero recorded ad spend for the users in your selected cohort. This is completely normal in a few common scenarios:

  • Zero-Spend Days: If a campaign was paused or did not utilize any budget during the exact dates you selected in the date picker, no return can be calculated for those specific users.
  • Delayed Network Reporting: If an ad network is experiencing a temporary delay in reporting spend data to the platform, the dashboard will display a dash until the spend data successfully arrives.
Jump to