Vantage Health: Four Numbers That Should Agree
Vantage Health runs quarterly reporting for a CMO dashboard. For the Q2 webinar campaign, the marketing email report shows 4,820 clicks on the registration email; the traffic analytics report shows 3,090 sessions with source = email. The campaign report states the campaign influenced $410,000 in closed-won revenue; the revenue attribution report on closed-won deals with an associated campaign contact says $182,000. The 'Q2 Webinar Registrants' active list shows 12,400 members, but the follow-up email was delivered to 9,870 contacts, and the team is being asked why 2,530 people were skipped. Separately, the LinkedIn campaign shows 1,900 contacts but only 640 new contacts in the same period. Nobody has documented which attribution model the dashboard uses; the previous analyst used last touch and the current analyst switched to first touch in April without telling anyone.
1. Explain the gap between 4,820 email clicks and 3,090 email sessions, and state how you would report clicks from now on.
Read the worked answer
The two numbers measure different objects. Email clicks are events on links; a session is a tracked visit. Gaps come from, in rough order of size:
- Multiple clicks per person. One contact clicking three links in one email produces three clicks and at most one session. The email report should be read as unique clicks (contacts who clicked) when comparing to sessions.
- Non-visit links. Unsubscribe, preference-centre and 'view in browser' clicks register as clicks but do not generate a session on your site.
- Link scanning. Corporate email security gateways and some mailbox providers prefetch links, registering clicks no human made.
- Consent and blocking. Where a cookie banner is configured and consent is declined, or where an ad blocker or tracking prevention is active, the visit happens but HubSpot cannot attribute a session to the email source.
- Session definition. Sessions expire after a period of inactivity, so a return visit counts as a new session while the click is counted once - which pushes the other way and makes the ratio non-constant.
Reporting fix. Report contacts-who-clicked alongside sessions, label the analytics figure 'sessions attributable to email', and stop presenting the raw click count as traffic. If the gap moves outside roughly 20–40%, investigate tracking or consent changes before celebrating.
2. Reconcile $410,000 influenced revenue against $182,000 attributed revenue. Which should the CMO use, and what has to change in the reporting?
Read the worked answer
They answer different questions. Influenced revenue is an ecosystem number: it counts deals that have at least one associated contact touching the campaign. A deal with contacts in five campaigns is counted in all five, so influenced totals across campaigns exceed company revenue and should never be summed. Attributed revenue applies a weighting model to the same deal-contact relationships, so a deal's value is split rather than repeated - which is why it is smaller here.
Which to use. For budget defence and channel comparison, attributed revenue under a documented model - it is additive and comparable across campaigns. For 'is this programme part of the conversation?', influenced revenue with a clear note that it overlaps. The CMO's single dashboard number should be attributed revenue, with influenced revenue as context beside it, not beneath it in the same column.
What must change. Document the model on the dashboard itself. Freeze it for at least four quarters - the April switch from last touch to first touch means Q2 and Q1 are not comparable, and any trend line drawn across them is fiction. Publish the date basis too (deal close date versus contact creation date), because a deal closing in Q2 from a Q1 campaign lands differently under each. Then re-state Q1 under the new model for the comparison.
3. Explain the 2,530 skipped contacts and the 1,900-versus-640 gap, and say which one is actually a problem.
Read the worked answer
The 2,530 are mostly legitimate exclusions. An active list counts members; a send counts recipients. Between the two sit: contacts who unsubscribed from that subscription type or globally, hard-bounced addresses, contacts marked non-marketing, suppressed or blocked domains, contacts with invalid or missing email, and anyone added to the list after the send was queued. Duplicates created since the send also inflate the list. So the first job is a breakdown by reason, not an explanation - pull counts per exclusion category and only then answer the CMO. If one category dominates, that is the story; if they are spread evenly, the list is simply normal and the dashboard should show delivered, not list size.
The 1,900-versus-640 gap is the more interesting one, and probably not a problem. LinkedIn campaign contacts include existing CRM records reached by retargeting and matched audiences, while new contacts counts records created in the period. 1,260 people who already existed being re-engaged is exactly what retargeting is for. It becomes a problem only if the campaign was sold internally as an acquisition play with a new-contact target - in which case the reporting, or the audience, was mis-specified.
Recommendation. Report paid social on two lines: new contacts acquired, and existing contacts engaged. Neither number alone tells the CMO whether the spend worked.