Last-touch budget review with messy UTMs

Use a multi-touch view for budget discussions and keep last-touch as a diagnostic, not the decision rule. Last-touch overcredits paid search because it often captures the final high-intent click after webinars, organic research, and partner introductions have done earlier work. For a 90-day cycle with five or more touches, a linear or position-based model better reflects influence, but it is still an approximation. I would present three views side by side: first-touch, last-touch, and multi-touch or campaign-influence. Add self-reported attribution from sales calls for deals where digital tracking is incomplete. The tradeoff is that multi-touch depends on clean contact and deal association, and it can overstate low-cost touches if you weight every interaction equally. Position-based weighting can reflect discovery and closing more realistically. The executive recommendation should be based on influenced pipeline, cost per opportunity, close rate, and deal velocity by channel, not on one attribution number.

Fix tracking and deal association before debating model choice. Standardize UTM naming for source, medium, campaign, and content, and apply the same convention to partner links so partner referrals are not misclassified as direct or organic. Audit existing campaigns for missing or inconsistent values. Require sales to associate deals with the originating or influencing campaign when creating or updating a deal; make this a pipeline hygiene check. Define lifecycle stages and campaign membership rules so a webinar registrant, a content download, and a partner referral are counted consistently. Deduplicate contacts and companies, especially where partners introduce contacts already in the CRM. Document which reports use campaign association versus contact properties. Run a weekly exception report for deals without a campaign, missing UTMs, and suspicious direct traffic spikes. Until these fixes are in place, any attribution model will produce confident but unreliable conclusions.

Do not cut webinars or partners based on last-touch alone. Paid search may deserve credit for capturing demand, but the CFO needs to know what creates pipeline. For webinars, compare 900 registrants and 4 percent opportunity rate against cost per opportunity, opportunity quality, and influenced closed-won revenue. If webinars mostly influence deals that close through search, cutting them could reduce pipeline next quarter even if last-touch looks weak. Optimize before cutting: tighten registration targeting, add qualification questions, follow up within one day, and measure webinar-influenced opportunities. For partners, low volume but a 35 percent close rate is promising; calculate cost per partner-sourced opportunity and whether partner volume can scale. A defensible recommendation is to protect partner spend, run a webinar optimization test with a holdout or cohort comparison, and shift a small portion of paid search budget only if incremental tests show it is saturated. Present the decision with influenced pipeline and deal velocity, not just last-click share.