baseline (in marketing measurement)
Also called: marketing baseline
What a metric such as leads, deals or margin would have been without the campaign, estimated from history, a comparable period or a control group that did not get the campaign. Only the difference above the baseline can be credited to marketing.
A baseline is an estimate of what a metric (leads, opportunities, deals, margin) would have been without the campaign. Marketing can only claim the difference between what happened and the baseline. Without one, every bit of growth from seasonality, word of mouth, a market upturn or the sales team's own work ends up credited to the last campaign.
There are three common ways to build one. History: the same period last year, which handles seasonality but not this year's trend. Trend: last year adjusted for how the business has been growing anyway. Control group: a comparable region, segment or set of accounts that did not get the campaign; last year's figure for the campaign group is multiplied by the growth the control saw. A control chosen before the campaign starts, ideally at random, is the strongest of the three.
Every baseline rests on an assumption (that the control behaves like the campaign group, that last year was normal), so state it next to the result. Small numbers make baselines noisy, and B2B effects often arrive months later, when deals close. A baseline doesn't make measurement precise; it makes it honest about what would have happened anyway.
Example
In autumn 2026 (September–November) Klasna for Teams spent 325,000 UAH on campaigns in Kyiv and Lviv only; other cities got none and served as the control. Opportunities, compared with the same months of 2025:
Baseline for Kyiv and Lviv: 15 × 1.2 = 18 opportunities would have come anyway. Extra opportunities: 31 − 18 = 13, not 37 (all of them) and not 17 (the year-on-year growth).
The same logic by channel, using the control's 20% growth: referrals went from 10 to 12, exactly the baseline of 10 × 1.2 = 12, so 0 extra; content went from 5 to 7 against a baseline of 6, so 1 extra. The channels that did not exist in 2025 (events, outbound, the partner) have a baseline of 0, and their 3 + 6 + 3 = 12 opportunities plus content's 1 give the 13.
Common mistakes
- Year-on-year growth as the effect. Last year's figure ignores the growth that would have happened anyway; adjust it with a control or trend.
- Picking the control after seeing the results. Choose the control regions or accounts before the campaign starts, ideally at random.
- Crediting organic channels. Referrals and unpromoted content that grow exactly like the control are baseline, not campaign results.
- Ignoring small numbers. With 5 or 6 opportunities in a control, one deal changes the baseline noticeably; give a range.
- Measuring too early. B2B results land when deals close, months after the campaign; set the measurement window in advance.
Learn it in the course
- Measuring marketing with a baseline · Why "the webinars returned twelve times their cost" is wrong: a baseline from the same months last year and a control group, incremental margin instead of revenue, and an honest ROMI.
- Case: a quarter's budget and what really moved · Leads more than doubled this autumn. Split the credit between marketing and the baseline, channel by channel, and recommend where next quarter's 325,000 UAH should go.
- Case: Veloria is 38% of Teams revenue · Check a concentration headline against a year of data, turn the share into money, see where concentration moves next, and write the key account plan and the answer to the CEO.