Salesforce campaign attribution: prove which marketing drives pipeline

You prove which marketing drives pipeline by connecting three things in Salesforce: campaigns with consistent types, campaign members who are real buyers, and opportunities with accurate contact roles. Campaign Influence then credits campaigns to opportunities using a model you choose. Most attribution problems are data problems, so fix the records before arguing about models.

Why can't we answer which campaigns created pipeline?

Usually the attribution logic is fine and the inputs are missing. Salesforce can only credit a campaign when the people on a deal are also members of that campaign.

When marketing ops leaders ask us why their influence reports look thin, the causes repeat across orgs:

• Opportunities have no contact roles, so there is nobody to match against campaign members.
• Campaign members are added late, in bulk, or with a single status that never marks who responded.
• Reps convert leads without a campaign, or create opportunities straight from an account.
• Events, webinars, partner referrals and outbound sequences never get loaded as campaigns at all.
• Campaign names and types drift, so a report cannot group like with like.

Each gap quietly removes credit from marketing. The result is a report that undercounts everything, and a leadership team that stops trusting it.

How should we structure campaigns, types and names?

Use a shallow hierarchy, a short controlled list of campaign types, and one naming pattern everyone follows. Structure decides what you can roll up later.

Salesforce supports campaign hierarchies up to five levels deep, and hierarchy statistic fields roll child results up to the parent. Most teams need three levels at most: a program or fiscal theme, a channel or initiative, and the individual tactic.

• Type: a picklist of channels such as Webinar, Event, Paid Search, Content Syndication, Partner and Outbound. Keep it under a dozen values.
• Name: a fixed pattern, for example year-quarter, type, region and short description.
• Member statuses: define which status counts as a response for each campaign type, and tick Responded only for those.
• Costs: fill Budgeted Cost and Actual Cost on the lowest-level campaign so totals roll up correctly.
• Owner and dates: every campaign has an accountable owner, a start date and an end date.

Lock the type and status values with picklists and validation rules. Free text is where attribution goes to die.

What does Salesforce Campaign Influence actually measure?

It creates a record linking a campaign to an opportunity and assigns that campaign a share of influence. The model you choose decides the share.

The default model is Primary Campaign Source. It gives all of the credit to the single campaign in the opportunity's Primary Campaign Source field. During lead conversion, Salesforce fills that field with the lead's most recently associated campaign, and the user can change it before saving.

Customizable Campaign Influence goes further. Its models scan active campaigns for members who also hold a contact role on an open opportunity, then create influence records. You can limit that matching with an auto-association time frame and campaign-type filters. Once an opportunity is closed won or lost, no new influence records are created for it.

You can also build custom models. Influence for those is assigned by hand on the opportunity, or by Apex or the API. That suits teams with their own weighting rules, but it means someone owns the code.

Which attribution model should we report on?

Report on more than one, and say which question each answers. No single model is true; each one is a lens with a known bias.

A practical pattern is sourced pipeline on Primary Campaign Source or first touch, plus influenced pipeline on an even or custom model. Present both on the same page so nobody adds them together.

Where do Account Engagement and Marketing Cloud fit?

They feed campaign data into Salesforce; they do not replace the CRM data model. Attribution still runs on campaigns, members and opportunities.

In Account Engagement, Connected Campaigns tie each marketing campaign to a Salesforce campaign so engagement lands on the CRM record. Our account engagement setup guide covers the connector and campaign setup in detail. On top of that, B2B Marketing Analytics Plus includes a Multi-Touch Attribution dashboard. Salesforce Help says it requires Customizable Campaign Influence and Connected Campaigns. It also lists Account Engagement Plus or Advanced editions with the B2B Marketing Analytics Plus add-on; confirm current packaging before you plan around it.

Marketing Cloud Engagement is more often used for consumer and high-volume programs. Through Marketing Cloud Connect, sends can target Salesforce campaigns and tracking can flow back to the CRM. Salesforce also offers attribution inside Marketing Intelligence for channel and spend analysis. Which pieces you own depends on your contract, so check with your account team.

How do we make opportunity contact roles reliable?

Make adding them part of the sales process, not a favor to marketing. Without contact roles, influence models have nothing to match.

• Require at least one contact role, with a primary, before an opportunity passes an early stage.
• Define a short role list that matches how you sell, such as decision maker, evaluator and champion.
• Let lead conversion create the first role, then prompt reps to add the rest of the buying group.
• Report weekly on open opportunities with no contact roles, by owner, and review it in pipeline meetings.
• Enforce it with a Flow or trigger that counts contact roles on the opportunity, plus a validation rule on that count.

Explain the payoff to sales leaders in their terms. Contact roles also show who is engaged on a deal, which helps the forecast conversation as much as the marketing report.

How do we capture offline and sales-led touches?

Treat every offline source as a campaign with members and statuses. If it is not in Salesforce as a campaign member, it cannot earn credit.

• Events and trade shows: load attendee lists within days, with statuses for registered, attended and met at booth.
• Webinars: sync registrants and attendees through your webinar connector, or load them on a fixed schedule.
• Partner referrals: create one campaign per partner program, and a lookup to the referring partner account.
• Outbound: give each sequence or play its own campaign, so sales-generated pipeline is counted honestly.
• Direct mail and field marketing: use response codes or landing pages that write back to the campaign.

The building-products manufacturer in one of our case studies had exactly this gap. Trade-show leads arrived with minimal information, and the team could not measure cost per lead. We moved it to Account Engagement integrated with Sales and Service Cloud, with campaign tracking and attribution for trade shows and other initiatives.

Is self-reported attribution worth adding?

Yes, as a complement. A required free-text or picklist question on demo and contact forms catches channels your tracking misses.

Podcasts, peer recommendations, communities and private messages rarely leave a clickable trail. Store the answer on the lead and carry it to the contact and opportunity at conversion. Compare it quarterly with system-recorded sources, and investigate the channels where the two disagree most.

What reports and dashboards prove marketing's contribution?

Start with standard campaign reports and a small influence dashboard. Add tools only when native reporting cannot answer an agreed question.

• Campaign ROI Analysis report, using hierarchy statistics for program-level totals.
• Campaigns with Influenced Opportunities, split by model.
• Sourced and influenced pipeline by campaign type and quarter.
• Cost per response and cost per opportunity, using Actual Cost.
• A hygiene panel: opportunities without contact roles, campaigns without type or cost, and leads converted with no campaign.

Keep marketing views separate from sales leadership dashboards, which have their own design rules. Our guide to dashboards leadership will use explains how to keep a board focused.

A dedicated attribution tool or CRM Analytics earns its place when you need weighted multi-touch across web sessions, ad platforms and offline data. It also fits when data volumes outgrow standard reports. Both still depend on the same clean campaigns and contact roles, so fix those first.

How do we keep attribution data clean after launch?

Give it an owner, written standards and a recurring check. Attribution decays quietly when nobody watches it.

• Name one marketing ops owner for campaign standards and influence settings.
• Publish the type list, naming pattern and member statuses where campaign creators work.
• Restrict campaign creation to trained users, with a request path for everyone else.
• Review the hygiene panel monthly and fix gaps at the source, not in spreadsheets.
• Change models or time frames only with a dated note, so trend lines stay explainable.

What belongs in a phase-one attribution plan?

Fix the data foundation and agree on two numbers before anything else. Multi-touch modeling can wait until the basics hold.

• Audit current campaigns, types, statuses and contact role coverage.
• Agree definitions for sourced and influenced pipeline with sales and finance.
• Standardize campaign types, naming, statuses and the hierarchy.
• Require contact roles and make Primary Campaign Source dependable at conversion.
• Load the main offline sources as campaigns going forward.
• Enable Customizable Campaign Influence with one additional model and a sensible time frame.
• Ship one dashboard and one hygiene report, then review them after a full quarter.

Lead scoring and forecast accuracy are related but separate projects. If those are on your list too, see our guides on lead scoring and on why forecasts go wrong.