Why Salesforce campaign influence shows zero revenue for events (and what to report instead)

Salesforce shows zero revenue next to an event because the default campaign influence model gives the whole revenue share to one primary campaign source. Every other campaign the contact belongs to still appears on the opportunity as an influence record, but with no revenue attached. The event touched the deal and the model did not pay it.

The primary campaign source is entered on the opportunity, usually when a contact is converted or added to a new opportunity. Whatever sits in that field collects the dollars. A webinar that ran six weeks earlier and a dinner that ran last month can both be on the deal, and only one of them gets credit in a revenue report.

A marketing lead usually finds this out in a meeting. She opens a report, sees the event on the opportunity, sees nothing next to it, and concludes the attribution is broken. In the orgs we work in, it is usually doing exactly what it was configured to do. The configuration is what deserves the question, and that is a different conversation from a data fix.

Open the campaign influence related list on the opportunity. You will see one row per campaign that registered, each with a revenue share percentage. Under a primary source model, one row says 100% and the rest say 0%. The zero is the model's answer to how much of this deal's revenue belongs to that campaign, and under that model the answer is none.

Two kinds of report give opposite impressions here. A yes/no report, which opportunities did this campaign touch, counts every row, including the zero-share ones, so it shows the event on the deal. A revenue report, how much attributable revenue did this campaign produce, sums the share, so the event shows nothing. Teams build one and read the other, then argue about which is wrong. Both are correct about the question they answer.

Influence is calculated through contacts. For a touch to register, the person who attended the campaign has to be a contact role on the opportunity. If someone scans a badge at a booth and a different person from the same company is added to the deal, the opportunity gets no link to that event.

In one org we worked in, an account page showed no influence while one of its contacts had several campaign touches, and we had not verified why by the time we left it. Test account-level views in your own org before you promise anyone a number from them.

This is the short circuit behind most missing-event complaints. The deal is real, the event was real, and the person on the deal was never the person at the event. Sales tends to add whoever is easiest to reach, so the opportunity picks up a champion or a finance contact while the attendee stays off it. The fix sits at opportunity creation: require a contact role, and require that the contact who prompted the opportunity is on it. We usually build that as a validation rule or a required step in the conversion flow. Check how reps create opportunities too, because some creation paths, cloning among them, may not carry contact roles across. Treat cloning as a suspect to test.

It is the explanation we hear most often, and it did not hold up when we checked it. The theory says that if an SDR emails a contact after the event, the SDR gets the last touch and the event loses credit. Influence in a contact-based setup does not depend on touch order. If the contact attended the campaign and sits on the opportunity as a contact role, the influence record exists whoever contacted them afterward.

A report can still leave the deal out. Event performance reports often carry filters on opportunity source or owner team, and one filter on an SDR-sourced field is enough to drop the deal. In the org where this came up, the opportunity everyone thought was missing was present in the right reports once we read the filter line by line. Read the filters before you touch the model.

Open a view of contacts on deals grouped by campaign member status and you will often find "Booth visit" next to "Visited booth", along with other pairs that mean the same thing. Each spelling counts separately, so an honest summary of booth engagement is split across rows, and any filter built on one spelling misses the other. It is workable, and it gets worse with every event.

Map the variants to a short set, roll them up in reporting, and agree the set before the next event. The same statuses later become inputs to lead scoring, so the cleanup pays twice. Decide what each status means in terms of intent. A booth scan, a conversation at a hosted dinner and attendance at a broad webinar are different signals and should not land in one bucket by accident.

Report what the model measures reliably. Count the opportunities with campaign influence in the period, split by in-person and webinar, and show the trend. That number does not depend on who got the revenue share, so the zero-share rows count. Total the dollars separately, by joining opportunity, account and campaign influence in a BI tool or a custom report type, and label the figure as dollars on influenced deals rather than dollars attributed to events.

Keep the recipe consistent. A measure that is slightly off but built the same way every quarter tells you whether events are trending up or down. Budgets and targets were usually set against a particular recipe, and changing it mid-year makes this quarter incomparable with the last four. Resist filtering to primary influence only. That narrows the view to the deals where the event happened to be recorded as the source and hides most of what events touch.

Marketing teams often ask for a weighted split, a slice to an ebook and a bigger slice to a dinner. We have not seen those weights structured in a way that finance and sales will defend. Argue over the count and the dollars on influenced deals. Those are numbers both sides can reproduce.

Changing the revenue model is a wholesale change. It rewrites how every historical and future opportunity is credited, and it lands on every dashboard that reads campaign revenue. We recommend leaving it alone until two things are true: contact roles are reliably present on opportunities, and someone with budget authority has said which question the new model has to answer. A model change on top of missing contact roles gives you new numbers built on the same gaps.

A related decision is the lookback window for auto-associating contacts and campaigns to an opportunity. In one engagement the team floated a shorter window while the previous consultant had suggested a longer one, and nobody had decided. The window is a business choice about how far back a touch still counts as part of the story. Pick it deliberately, write it down, and test it against a handful of closed deals you know well.

Pull ten closed-won opportunities from last quarter. For each one, list the people who attended your events in the months before it closed and check whether any of them are contact roles on the deal. If most are, you have a model question and the reporting recipe above is your answer. If most are not, you have a contact-role problem, and no influence model will fix it. Do that count before your next planning meeting.