Case: compare email and social traffic without confusing clicks with leads

Keep the offer and time window comparable

Use Example Domain for a harmless routing rehearsal. For a real campaign, use your registration page with website analytics and a successful-form event already configured. Keep the offer, form, and reporting period consistent.

All campaign numbers below are training inputs. First verify the plumbing; a comparison is only useful when each source reaches the intended destination.

Create two complete tagged destinations

https://example.com/?utm_source=newsletter&utm_medium=email&utm_campaign=workshop_2026
https://example.com/?utm_source=social_post&utm_medium=social&utm_campaign=workshop_2026

1. Create two account-owned 302 links named “Workshop — email” and “Workshop — social,” each with its corresponding complete destination.
2. Record baseline human and bot counts, the time, and the time zone. Keep the rehearsal private.
3. Open the first short URL manually three times and the second twice. Reopen the short URL each time; refreshing the final page does not pass through SOSU again.
4. Check the final utm_source values. After the processing queue runs, refresh each link’s statistics and compare changes from baseline.
5. A clean rehearsal should produce three and two human requests, not three and two unique people. Investigate extra requests or bot classification if your observations differ.

Read a sample campaign report correctly

Sessions and registrations in this table must come from the destination analytics for the same period, not from SOSU. Calculate registrations ÷ sessions × 100%. Here the email example brings fewer sessions but more registrations.

This small invented dataset does not prove that email is inherently better. Audiences and placement exposure differ. Repeat the comparison over comparable periods and account for costs separately.

Avoid these measurement traps

• Do not share one short URL across both sources; query strings added to SOSU short URLs are not forwarded.
• Record the boundary between rehearsal and launch rather than presenting your own clicks as campaign results.
• Keep message-preview bots out of human results.
• Do not silently substitute short-link clicks for website sessions in a conversion rate.
• Forwarded links, consent choices, and blockers can change attribution; never put personal data into campaign tags.