Sales Lift compares purchase outcomes for an exposed group and a control group within an agreed observation window. Pharma studies may use available anonymised pharmacy purchase data and partner matching. The result is a difference within that study design; it should not automatically be extrapolated to the whole market or treated as causal when groups are not comparable.
How it works
- 01Agree KPI and study design
- 02Build comparable groups
- 03Run campaign and observe purchases
- 04Calculate difference and limitations
One eligible audience, two comparable groups
Study design
Agree on purchases, buyers or sales value as the KPI. Set baseline and follow-up windows, test and control criteria and minimum sample size before launch.
Reading incrementality
Distinguish percentage points, relative uplift and absolute purchases. A simple exposure-purchase match shows association, not necessarily added effect.
What is fixed, and when
- 01Before the campaign
Set the KPI, baseline, test and control groups, matching rules and purchase window.
- 02During the campaign
Show ads to the test group and record exposure. Keep control free from planned exposure.
- 03After the campaign
Compare purchases in the agreed window and examine external factors.
Published example
In a 2023 medical device case, purchase conversion in the test group was 34% higher than control. This is one campaign result, not a forecast for other brands.
“+10%” and “+1 pp” describe different quantities
What Sales Lift measures
Before launch, choose the primary metric: buyer share, purchase count, frequency, packs or monetary value, where the data support it. Higher frequency among existing buyers is not the same as a higher buyer share. Fix group definitions, the period and the metric denominator in advance to avoid choosing the most favourable wording after seeing the result.
Why control is designed in advance
Before placement, we define whom to expose to advertising and whom to keep in the control group. We record the baseline period and matching rules. After the campaign, results are compared within those conditions. If a control group was not planned, a retrospectively selected group cannot be guaranteed to have been comparable and unexposed.
What the report contains
To review the report, check both group sizes, underlying values of the primary metric, the purchase window, source coverage and matching rules. Then examine the observed difference and its uncertainty. A small difference in a large sample and a substantial difference in a small sample call for different decisions. Note promotions, product availability and other events that may have affected purchases.
Where the study is useful
Sales Lift can test offline impact from a launch, competitor-buyer strategy, repeat-purchase effort or audience comparison. Feasibility is checked anew for every category, purchase volume, reach and campaign schedule.
Choosing an outcome measure
| Metric | What is compared | What to check |
|---|---|---|
| Buyer share | Buyers among the participants of each group | The same rules for calculating the share and the purchase observation period. |
| Purchase frequency | Purchases per person within the agreed base | All participants or buyers only: the bases differ. |
| Product quantity | Packs or product units in the available sample | Product definitions and counting units. |
| Monetary value | Purchase amounts in the available data | Price changes, promotions and source coverage. |
Limitations
Sales Lift requires sufficient purchases, reliable matching, comparable groups and permission to use the data. Seasonality, distribution and promotions can affect the result.
Questions
How does Sales Lift differ from Brand Lift?
Sales Lift measures purchase behaviour using sales data. Brand Lift typically surveys awareness, perception or intent. They answer different questions and can be used together.
Can Sales Lift be ordered after a campaign has ended?
Standard Pharmatic Sales Lift must be designed before launch: define test and control groups, the baseline period, KPI and purchase window in advance. Selecting a group after the campaign does not guarantee comparability or the absence of advertising exposure. Retrospective sales analysis is a different format with its own limitations.
What data are required?
The study needs ad exposure data, available anonymised purchase data and an agreed partner matching route. Group size, purchase frequency and category coverage are checked before launch.
What if most sales happen in offline pharmacies?
Partner purchase data can make this measurable. We check category coverage and whether enough purchases are observable for a group comparison.
Does the result equal total brand sales growth?
No. It is a difference in the agreed sample and metric. Wider extrapolation requires additional assumptions and calculations.
Can promotions or seasonality affect the result?
Yes. We account for promotions, distribution changes and seasonality in study design and explain material external events in the report.