Salesforce Campaign Influence: How It Works, Pros and Cons

Campaign Influence links Campaigns to Opportunities and assigns each linked campaign a share of the opportunity's amount. That share is what turns a campaign from an activity into a number: instead of reporting that a webinar had 240 registrations, you can report that it is credited with a proportion of a pipeline figure, and roll that up across every campaign in the quarter.

Salesforce stores this on the Campaign Influence object, one record per campaign and opportunity pair, carrying the model it belongs to and the revenue share it holds. Because those are ordinary records, everything downstream is ordinary Salesforce work: reports, dashboards, list views and formula fields all behave the way your team already expects.

Campaign Influence does not observe behaviour. It reads relationships that already exist in the org, and there are exactly two of them.

• The Campaign MemberA Lead or Contact has to be a member of the campaign. That is the record saying this person was reached by this campaign, and it is created by a marketing tool, an import, an automation or a person.
• The Opportunity Contact RoleThat Contact has to be attached to the Opportunity as a Contact Role. It is the link between the person marketing reached and the deal sales is working, and it is the one most orgs are missing.

This is the distinction behind most of the confusion. Every Opportunity has a Primary Campaign Source field, which points at one campaign and gives it the whole amount. Customizable Campaign Influence is the version that shares the amount across several campaigns, and it has to be switched on in Setup.

If your pipeline reporting still shows exactly one campaign per opportunity, this is almost certainly why. The fix is a setup change rather than a purchase.

Customizable Campaign Influence ships with straightforward models: first touch, last touch and an even split. You can also write your own influence percentages, manually or through automation, which is how teams build weighted models when they want them.

The clearest way to see the difference is to put one deal through all three. Take a €40,000 opportunity whose contacts touched three campaigns: a LinkedIn Ads campaign in March, a webinar in April, and a branded Google Ads campaign in June, days before the deal was signed.

None of these is the true answer, and that is the point of having several. First touch flatters demand generation, last touch flatters the campaigns closest to the signature, and an even split flatters whichever team runs the most campaigns. Reading two of them side by side is more honest than arguing about which single one is correct.

Attaching campaigns to opportunities by hand does not survive contact with a real pipeline, so Salesforce can add them automatically. Auto-association creates the influence record when a campaign meets the rules you set, and the most consequential of those rules is the time frame: how far back before the opportunity was created a campaign membership still counts.

There is no correct number here, only a number that matches your sales cycle. Too short and the campaigns that opened the relationship fall out of every report, which quietly defunds top of funnel work. Too long and campaigns get credit for deals they had nothing to do with. Teams with a six month cycle usually end up somewhere near a year, and then adjust once they have seen a few quarters of reports.

Campaign Influence is Sales Cloud. Salesforce also sells marketing products, and they add attribution of their own: Account Engagement, formerly Pardot, carries the B2B Marketing Analytics dashboards with weighted multi-touch models, and the marketing platform has added multi-touch attribution dashboards of its own. So it is not true that Salesforce has no multi-touch attribution. It has several, in different products, licensed separately.

What they share is their input. Every one of them attributes what has been represented inside Salesforce as a campaign, a member or an activity. That is a boundary rather than a flaw, and knowing where it runs makes the rest of this article, and any buying decision, much simpler.

• It costs nothing extra. The feature is in the editions most B2B orgs already run, so the only investment is configuration time.
• It reports natively. Campaign Influence records behave like any other object, so your existing dashboards, sharing rules and exports all work.
• It credits offline work properly. Events, webinars, conferences and outbound sequences are first class here, which is more than most web analytics tools can say.
• It is flexible where it matters. Predefined models cover the common cases, and custom percentages let an operations team encode a model the business actually agreed on.
• It aligns two teams around one record. When the campaign credit sits on the opportunity sales is working, marketing and sales are reading the same number rather than two exports.
• It scales with the CRM. No separate data model to maintain, no second system of record, no reconciliation meeting.

None of these are bugs. They follow from what the feature is: a model over records that already exist in Salesforce.

• It starts at the record, not at the clickInfluence begins with a Campaign Member. The ad click, the organic session and the pricing page visit that happened before anyone filled in a form were never records, so they cannot be credited.
• It only knows what is a CampaignSearch, direct, referral, review sites and AI assistants have no campaign to join. Unless somebody creates and maintains a campaign to represent them, they are invisible to the model.
• Campaign Cost is typed in by handNo spend is pulled from Meta, Google, LinkedIn or Microsoft. Without live cost there is no ROAS and no cost per closed deal, only revenue divided between campaigns.
• It depends on human diligenceContact Roles have to be attached, campaign memberships have to be current and the time frame has to suit the cycle. The report degrades quietly when any of that slips.
• Nothing leaves SalesforceThe credit stays in the report. Closing a deal changes nothing in any ad account, so the platforms keep optimising toward the events they can see, which are usually form fills.

It is usually enough when

• Most of your pipeline comes from events, webinars, outbound and partnerships, all of which are naturally Salesforce Campaigns.
• You have an admin or operations person who keeps Contact Roles and campaign membership honest.
• Budget decisions are made quarterly, at the level of programmes rather than individual ads.
• The question you are answering is which campaigns influenced revenue, not what a lead cost.

It stops being enough when

• A meaningful share of pipeline starts with a paid click, because that click is not in the model at all.
• Somebody asks what a closed deal cost, which needs spend joined to the pipeline rather than a manual cost field.
• Your reports keep showing Direct, Other or a blank source for deals you know came from advertising.
• You want the ad platforms to learn from your outcomes, which requires sending the closed deal back out.

1. In Setup, find Campaign Influence and enable Customizable Campaign Influence. Until this is on, every opportunity has exactly one Primary Campaign Source.
2. Choose the model or models you want available, and decide whether influence percentages will be set by a model, by automation or by hand.
3. Configure auto-association, including the time frame that decides how far back a campaign membership counts before the opportunity was created.
4. Make sure Opportunity Contact Roles are actually being created. A flow, a required step in the sales process or a validation rule is what keeps this true six months from now.
5. Add the Campaign Influence related list to the Opportunity page layout so the credit is visible on the record, not only in a report.
6. Build the reports: campaigns with influenced opportunities, influence by model, and a dashboard your marketing and sales leads both look at.

• Name campaigns to a convention that survives a year, because the report is only as readable as the names in it.
• Keep campaign hierarchies shallow, and use the parent campaign for the programme rather than for every asset in it.
• Fill in Campaign Cost, even roughly, or half of the ROI questions cannot be answered at all.
• Attach Contact Roles at the point the opportunity is created rather than at the end, when nobody remembers who was involved.
• Review the time frame once a quarter against your actual average sales cycle.
• Report two models side by side, so nobody has to defend a single model as the truth.
• Record offline touches as campaign memberships if they matter, because a conference conversation that closed a deal deserves to exist in the data.
• Decide who owns the model. Attribution without an owner drifts back to whatever the last admin configured.

Most teams that outgrow Campaign Influence do not replace it. They keep it for the campaigns they run inside Salesforce, and they add a measurement layer in front of the CRM to cover what the model cannot observe: the sessions before the record existed, the click IDs, and the spend.

That is what LeadJourney does. Every visit is tracked server-side on your own domain before a Lead exists, the anonymous first click is stitched to the Lead when the form is submitted, and first and last touch are written onto the Lead, the Contact and the Opportunity as fields your existing Salesforce reports can use. Spend is pulled from Meta, Google, LinkedIn and Microsoft Ads, so cost per closed deal becomes a live number rather than a manual one, and when the opportunity is marked Closed Won the amount goes back to the platform that produced it as an offline conversion.

The two layers answer different questions and both are worth having. Campaign Influence answers which of our campaigns influenced this revenue. A measurement layer answers where this revenue came from in the first place, and what it cost to buy. If you want the longer comparison, we ranked twelve attribution tools for Salesforce, and the Salesforce integration page lists exactly which fields get written.

Related guides: multi-touch attribution explained, how the gclid travels into the CRM, why GA4, Meta and Google never match your CRM and attribution over a long sales cycle.