What This Chart Shows
This chart breaks down total marketing impact: it starts with a baseline (outcome without marketing), then adds contributions from each channel, campaign, or lever, ending at a total. By the end, you see how those pieces add up to your total lift or marketing contribution.
Key Questions This Chart Helps Answer
- Which levers or channels contributed most to lift or revenue, and by how much?
- What is the baseline (outcome without marketing) and what is the total with marketing?
- How does my total marketing contribution break down by channel, campaign, or funnel stage?
- How does contribution change when I filter by date range, KPI, or outcome group?
Axes, Metrics, and Units
Element | Description |
|---|---|
| X-axis (vertical layout) | Category labels: Baseline, then each lever or group (e.g. channel, campaign, funnel), optional Other, and Total. |
| Y-axis (vertical layout) | Contribution (value) in the same units as the chosen metric (see below). |
| Horizontal layout | Same logic with axes swapped: categories on Y, contribution on X. |
| Bar colors | Green = positive (increasing) contribution. Red = negative (decreasing) contribution. Purple = Baseline and Total (absolute/total). |
| Baseline bar | Sum of the baseline KPI over the analysis window. |
| Total bar | Baseline plus sum of all contribution bars; should match the joint lift when the decomposition is additive. |
| Additivity annotation | If the sum of contributions differs from the joint lift by more than 5%, a red footnote-style annotation shows the discrepancy (e.g. sum vs joint and relative error). |
Control Options Reference
Control | Meaning |
|---|---|
| KPI classification / hierarchy / drilldown | Filter or drill into specific KPIs. |
| Plot type | Vertical (v) or Horizontal (h) orientation. |
| Grouping mode | Aggregate levers by: lever (default), channel, campaign, funnel, series, source_type, or topic. See reference page here. |
| Date range | Time window for attribution. |
| Event categories | Filter by specific event categories (include/exclude). |
| Funnel stages | Filter by specific funnel stages (include/exclude). |
| Search terms | Text search on lever/series labels. |
How to Interpret the Results
- Read left-to-right (or top-to-bottom): Baseline → each contribution → optional Model Residual → Total. Each step adds or subtracts the bar’s value.
- Largest positive bars: Main drivers of lift; focus optimization and storytelling here.
- Negative bars: Levers or groups the model attributes with negative incremental effect; review placement, creative, or audience.
- Other: Remaining levers/groups beyond the top 15; use grouping or filters to break out important segments.
- Total vs baseline: Difference = total attributed lift (or marketing contribution) in the period.
- Additivity note: If “Σ vs Joint” appears, the sum of bars doesn’t match the overall KPI delta (e.g., filtering, grouping, or per-target vs aggregate); use it as a sanity check, not to dismiss the chart.
Practical Applications for Marketers
Application | How to use this chart |
|---|---|
| Budget and planning | See which channels/levers contribute most; shift spend toward positive contributors. |
| Channel mix | Compare contribution by channel (use grouping by channel) and adjust mix. |
| Campaign evaluation | Group by campaign to see which campaigns drove lift. |
| Funnel and topic | Group by funnel or topic to see where in the journey or which themes drive impact. |
| Reporting and storytelling | Use baseline → contributions → total to explain “how we got to total lift.” |
| Validation | Use additivity note and baseline/total to check consistency with other reports. |
Common Mistakes and Misinterpretations
Mistake | Why it is a problem | How to avoid |
|---|---|---|
| Treating contribution as raw spend or revenue | Bars are modeled causal attribution (fair-share or lift), not actual spend or observed revenue. | Use axis title and metric label (Marketing Contribution or Lift); compare to cost/ROI charts for efficiency. |
| Comparing runs with different filters or date ranges | Changing date range, KPI, or outcome group changes the numbers. | Keep filters and date range consistent when comparing periods or scenarios. |
| Ignoring the additivity annotation | When shown, the bar sum does not equal joint lift; using the bar sum as total impact can be misleading. | Note the annotation; check filters and pipeline if you need the decomposition to match joint lift. |
| Using lever view when you care about channel | Lever-level bars can be noisy and hard to compare at a strategic level. | Switch grouping to Channel (or Campaign/Funnel) for a high-level read. |
| Assuming no data means no marketing effect | Empty or error states usually mean missing inputs or filters excluding all levers, not zero effect. | Check the on-chart message and ensure data has run and filters are on |
Caveats & Considerations
- Data dependency: Requires attribution data. If missing, the chart shows “Attribution data not available” and which family is missing.
- Empty or zero: If all contribution values are zero, the chart shows an explanatory message (e.g., marginal lifts not computed or zero effect).
- Filters: Event category, funnel stage, or search with no matches shows an empty-state message; no silent fallback to unfiltered data.
- Grouping: If you choose a grouping mode (e.g., topic) and the taxonomy isn’t available, the chart shows a short explanation instead of falling back to another mode.
- Top 15: Only the top 15 levers or groups by absolute contribution are shown as named bars; the rest are in “Other.” Use grouping or filters to see more detail.
- No-costs mode: When costs are absent or zero, the chart uses “Lift” and unit lifts; title may show “Lift Waterfall (Per Target).”
- Orientation: Vertical layout uses category on X and contribution on Y; horizontal swaps axes. Both show the same decomposition.