1) Meaning

Total Incremental Benefit (TIB) is the overall net gain achieved by applying a treatment (e.g., marketing campaign, discount, medical intervention) compared to if the treatment had not been applied at all.

It measures the absolute business or outcome improvement driven by an uplift strategy.
In uplift modeling, this is the final value of the Cumulative Incremental Gain (CIG) curve at 100% of the targeted population.

2) Formula

$\text{Total Incremental Benefit} = \sum_{i=1}^{N} \big( y_i^{\text{treat}} – y_i^{\text{control}} \big)$

Where:

  • $N$ = total population
  • $y_i^{\text{treat}}$​ = outcome if treated (e.g., purchase, conversion)
  • $y_i^{\text{control}}$​ = outcome if not treated

If expressed in monetary terms:

$\text{Total Incremental Benefit (Value)} = \text{Incremental Conversions} \times \text{Profit per Conversion}$

3) Example

A retailer runs a promotion campaign on 10,000 customers:

  • Treatment group (5,000 customers): 700 purchases (14% rate)
  • Control group (5,000 customers): 500 purchases (10% rate)

Incremental conversions: $700 – 500 = 200$

If profit per purchase = $50: $\text{Total Incremental Benefit} = 200 \times 50 = \$10,000$

The promotion generated 200 additional purchases worth $10,000 net benefit compared to no campaign.

4) Relation to Other Metrics

  • Incremental Gain (per decile/segment): local effect in a portion of the population.
  • Cumulative Incremental Gain (CIG): running total as more customers are targeted.
  • Total Incremental Benefit: the final cumulative value (total effect over all customers).
  • Qini Curve: plots CIG; the top of the curve shows the TIB.

5) Why It Matters

  • Directly quantifies business value of a campaign.
  • Allows ROI calculation:
    • $ROI = \frac{\text{Total Incremental Benefit} – \text{Campaign Cost}}{\text{Campaign Cost}}$
  • Provides a benchmark for comparing uplift models or campaigns.
  • Supports budget allocation by showing the true added impact.

Bottom line:
Total Incremental Benefit is the overall additional gain (conversions, revenue, or profit) produced by a treatment compared to no treatment. In uplift modeling, it’s the final cumulative benefit seen at the end of the Qini curve.