What This Chart Shows
This chart shows how much attributed impact (lift) each channel or group contributes and how efficient each is (lift per dollar spent). It has two panels: (1) Contribution — each bar is the signed attributed lift (or marginal lift in no-costs mode) for a marketing activity so you see both positive and negative contributions (e.g. cannibalization); (2) Efficiency — each bar is contribution per dollar spent (lift per $ invested) for the same groups.
A baseline and full outcome line can appear below the charts when the pipeline provides them.
Key Questions This Chart Helps Answer
- Which marketing activities contribute most to incremental lift, and which are negative?
- How does efficiency (lift per dollar) compare across channels or groups?
- Where might cannibalization or saturation show up as negative contribution?
Axes, Metrics, and Units
Element | Description |
|---|---|
| Left panel (Contribution) X-axis | Group labels (e.g. channel or series names). First bar = "Baseline" (baseline outcome). Then up to 15 groups, sorted by absolute contribution. |
| Left panel (Contribution) Y-axis | Contribution: Marketing Contribution or Marginal Lift. Modeled attribution; can be positive or negative. Same units as the selected KPI (e.g. conversions, revenue). |
| Right panel (Efficiency) X-axis | Same group labels as left (no Baseline bar on right). |
| Right panel (Efficiency) Y-axis | Efficiency: Lift per $ invested (contribution / cost). Costs mode: ratio, dimensionless. No-costs: efficiency panel shows "Efficiency (unavailable)" and gray bars. |
| Bar colors | Green = positive contribution or efficiency. Red = negative. Gray = Baseline bar (left only) or efficiency when costs unavailable. |
All contribution and efficiency values are modeled. They are not raw spend or raw conversions.
Control Options Reference
Control | Meaning |
|---|---|
| KPI classification / hierarchy / drilldown | Which KPI or outcome to use for attribution and baseline/full. |
| Grouping mode | Series / Channel / Campaign / Funnel / Source type / Topic. Determines how levers are aggregated into bars. |
| Date range | Time window for attribution when temporal data is used. |
| Event categories | Filter by specific event categories (include/exclude). |
| Funnel stages | Filter by specific funnel stages (include/exclude). |
| Search terms | Filter searches by text search. |
How to Interpret the Results
- Left Panel – Contribution: Tall green bars = large positive impact; red bars = negative impact (cannibalization, saturation, or other negative effects). Compare bar lengths to see who drives or hurts the outcome. Baseline (if shown) is the do-nothing level, not attributed to any group.
- Right Panel – Efficiency: When cost data exists, higher bars = more lift per dollar. Use to find the most efficient channels or groups; low or negative efficiency = poor or harmful return per dollar.
- No-costs mode: Right panel shows "Efficiency (unavailable)" and N/A bars; only the left panel is meaningful for contribution.
- Negative bars: Indicate the model attributes a negative effect to that group; interpret in context of the causal setup (e.g. saturation, cannibalization).
- Baseline and Full: Footer "Baseline: X | Full: Y" gives context: lift is on top of Baseline; Full is baseline + total attributed lift when provided.
Practical Applications for Marketers
Application | How to use this chart |
|---|---|
| Executive summary | One view of who contributes and who is efficient; use for CMO/leadership. |
| Budget allocation | Favor high contribution and high efficiency; reduce or investigate negative and low-efficiency groups. |
| Channel mix | Use Channel (or Campaign/Funnel) grouping to see which parts of the mix to scale or cut. |
| Cannibalization / saturation | Use red bars and negative efficiency to flag channels or tactics that may be hurting the outcome. |
| Storytelling | "These channels add lift; this one is negative; here's who is most efficient per dollar." |
Common Mistakes & Misinterpretations
Mistake | Why it is a problem | How to avoid |
|---|---|---|
| Treating contribution as total revenue | Contribution is attributed lift (incremental impact from the model), not total revenue or spend. | Use for incremental impact and efficiency, not P&L revenue. |
| Ignoring negative bars | Red bars show negative attribution (cannibalization, saturation, etc.); they are part of the story. | Always note red bars and use them in strategy, not only green. |
| Comparing efficiency when costs are missing | In no-costs mode the right panel is N/A; efficiency is not comparable. | Use Panel 2 only when cost data is present and the axis is "Lift per $ Invested." |
| Comparing across different grouping or filters | Bars and order depend on grouping, event category, funnel, search, and KPI. | Compare only with the same settings, or state how they differ. |
| Confusing Baseline with a channel | Baseline is the do-nothing outcome; it is not attributed to any channel. | Read Baseline as "starting point"; contribution bars are incremental to that. |
Caveats & Considerations
- Filter-empty: If event category, funnel stage, or search leave no levers, the chart shows the standard filter-empty state.
- Aggregation errors: If grouping fails (e.g. invalid mode or data), the chart shows an "Aggregation error" message.
- Taxonomy grouping: If channel/funnel/source type/topic returns no groups, the chart shows a short message that the grouping is not available and suggests trying another mode.
- No-costs mode: The right panel shows efficiency unavailable (gray N/A bars and annotation). Contribution in the left panel still uses the chosen metric (e.g. marginal lift).
- Top 15: Only the top 15 groups by absolute contribution are shown; others are not in the view.
- Click/drawer: Bars may be clickable for details; behavior depends on the app.