Unifying pipeline attribution across 14 regional marketing teams
Ashford's 14 regional marketing teams ran on 14 different spreadsheets. Meridian gave them one attribution model the whole company could trust.
- Client
- Ashford Logistics Group
- Duration
- 5 months
- Team
- 6 Meridian implementation specialists, 3 Ashford stakeholders
- Role
- Platform implementation, data migration, attribution model design
100%
Pipeline visibility
31 hrs/mo
Reporting time saved
+64%
Attribution accuracy
14
Regional teams unified
Overview
Ashford Logistics Group operates 14 semi-autonomous regional marketing teams across North America, each running its own campaigns, budgets, and — critically — its own spreadsheet-based attribution model. Corporate marketing leadership had no reliable way to compare regional performance or defend total marketing spend to the board. Meridian was brought in to replace all 14 models with one governed, company-wide standard.
The Problem
Each region attributed pipeline differently: some used last-touch, others first-touch, one region simply divided credit evenly across every campaign a contact had ever touched. When corporate tried to roll these numbers up for a quarterly board review, the totals didn't reconcile with Salesforce's own pipeline figures by nearly 40%. Marketing's credibility with the executive team was eroding quarter over quarter.
Research
Our implementation team spent three weeks auditing all 14 regional models, interviewing regional marketing directors, and reconciling every model's logic against actual Salesforce opportunity data. We found that no single existing model — first-touch, last-touch, or linear — matched Ashford's actual multi-region, multi-touch buying journeys, which routinely spanned 6 to 11 touchpoints across 90+ days.
Discovery
Workshops with regional stakeholders surfaced a shared, unstated requirement: regions wanted credit preserved for early-funnel brand campaigns, not just the last email that triggered a demo request. This ruled out simple last-touch models and pointed toward a time-decay, multi-touch model weighted toward moments of genuine engagement rather than raw click volume.
Competitive Analysis
Ashford had evaluated two other attribution vendors previously. Both required a rip-and-replace of the regional teams' existing ad tooling, which stalled adoption for over a year. Meridian's non-destructive, read-layer integration model — sitting alongside existing tools rather than replacing them — was the deciding factor in vendor selection.
Goals
Three measurable goals were set at kickoff: (1) a single attribution model adopted by all 14 regions within two quarters, (2) monthly reconciliation variance versus Salesforce under 5%, and (3) a 50% reduction in the manual hours regional analysts spent building spreadsheet reports.
Wireframes
Low-fidelity wireframes for the unified regional dashboard were sketched collaboratively with three pilot regions, focused on a single 'campaign to closed-won' waterfall view that every region — regardless of size or channel mix — could read the same way at a glance.
User Flow
The final flow let a regional director land on their own scoped dashboard by default, drill into any campaign to see full touchpoint-level detail, and roll up seamlessly into the corporate view without needing separate logins or exported spreadsheets — solving the exact hand-off problem that caused the original reconciliation gap.
Visual Design
Dashboards were designed around a restrained, high-contrast style intentionally different from Ashford's cluttered legacy BI tool — large single-number KPIs at the top of every view, with detail progressively disclosed rather than front-loaded, reducing the cognitive load on non-technical regional stakeholders.
Design System
We extended Meridian's core design system with an Ashford-specific theme (their brand navy as the primary accent) while keeping every interaction pattern — filters, date ranges, export actions — identical to the standard platform, minimizing training overhead for a 40-person rollout.
Development
The attribution engine was implemented as a nightly batch reconciliation job against Ashford's Snowflake warehouse, with an incremental streaming layer for same-day campaign performance. Historical data going back 3 years was backfilled and re-attributed under the new model for like-for-like year-over-year comparison.
Architecture
Data flows from Google Ads, LinkedIn Ads, and 11 regional ad accounts into a staging layer, is deduplicated against Salesforce contact and account IDs, and is scored using a configurable time-decay weighting function before landing in the reporting layer that powers every regional and corporate dashboard.
Performance
The reconciliation job processes roughly 2.1 million touchpoint events per month across all 14 regions in under 11 minutes, with dashboard queries returning in under 400ms at the p95 percentile thanks to pre-aggregated rollup tables refreshed on each batch run.
Accessibility
All dashboard components meet WCAG 2.1 AA contrast requirements, are fully keyboard-navigable, and every chart ships with an accessible data-table fallback view for screen reader users on the regional analyst team who rely on assistive technology.
SEO
Not applicable to this internal reporting engagement — the attribution platform is authenticated and not publicly indexed. Documentation for the rollout was published internally on Ashford's intranet with structured, searchable headings for the regional support team.
Testing
Every region's historical numbers were shadow-run against the new model for six full weeks before cutover, with discrepancies above 2% flagged for manual review by both Meridian's data team and the regional analyst. This shadow period was the single biggest driver of stakeholder trust ahead of go-live.
Deployment
Regions were migrated in three waves of roughly five regions each, two weeks apart, rather than a single cutover — allowing the implementation team to resolve data-quality issues discovered in wave one before they affected the remaining nine regions.
Results
Within two quarters, all 14 regions were reporting from the same attribution model. Reconciliation variance against Salesforce dropped from 40% to under 3%. Regional analysts collectively reclaimed roughly 31 hours per month previously spent building manual spreadsheet reports, redirected into campaign strategy work.
Lessons Learned
The biggest lesson was organizational, not technical: giving regional teams a voice in the attribution model's design — rather than mandating a corporate standard top-down — was what actually drove adoption. The wave-based rollout, born from necessity, became the template we now recommend for any multi-region deployment.