Article
Understand first-touch and last-touch attribution — and why honest missing data beats invented numbers.
Last-click attribution is convenient but misleading. It credits the final touchpoint before conversion and ignores everything that built awareness, consideration, and intent. For lead-driven businesses with long cycles, last-click systematically under-invests in top-of-funnel and over-invests in bottom-of-funnel.
A typical B2B journey might look like:
1. **Display impression** — Brand awareness (no click) 2. **LinkedIn ad click** — First site visit (first touch) 3. **Organic search** — Researching category 4. **Email newsletter** — Case study read 5. **Google search (branded)** — Return visit 6. **Direct visit** — Pricing page 7. **Demo request form** — Conversion (last touch)
Last-click credits only step 7. First-touch credits only step 2. Reality is all of them.
| Model | Credit Assignment | Best For | |-------|------------------|----------| | First Touch | 100% to first interaction | Brand awareness investment | | Last Touch | 100% to last interaction | Short-cycle, transactional | | Linear | Equal across all touches | Balanced view | | Time Decay | More to recent touches | Long cycles, momentum | | Position-Based (U-shaped) | 40% first, 40% last, 20% middle | Considered purchases | | Data-Driven | Algorithmic (requires volume) | High-volume, mature programs |
**No model is "correct."** Each answers a different question. Use multiple.
**Qualified Cost Per Lead (qCPL)** Total spend ÷ qualified leads (not raw leads). Filters out junk.
**Cost Per Opportunity (CPO)** Total spend ÷ sales-accepted opportunities. Closer to revenue.
**Customer Acquisition Cost (CAC)** Total sales + marketing spend ÷ new customers. The north star.
**Return on Ad Spend (ROAS)** Revenue attributed ÷ ad spend. Channel efficiency.
**Payback Period** Months to recover CAC from gross margin. Cash flow reality.
**Pipeline-to-Spend Ratio** Pipeline generated ÷ marketing spend. Leading indicator.
Campaign → Lead → Qualified Lead → Opportunity → Customer → Revenue
**At each stage, ask:** - What source created this? - What touches influenced it? - What's missing from our tracking?
**Common gaps:** - Offline events (trade shows, dinners) - Dark social (Slack, WhatsApp, email forwards) - View-through (impressions without clicks) - Cross-device (mobile research, desktop convert) - Sales-sourced (outbound, referrals) not linked to marketing
**Don't fabricate attribution.**
If you can't tie a deal to a campaign, mark it "unknown" or "sales-sourced." Inventing attribution to make dashboards look complete destroys trust and misallocates budget.
**Better approach:** 1. Measure what you can (UTM persistence, CRM campaign fields, call tracking). 2. Report coverage: "Attribution known for 68% of pipeline." 3. Invest in coverage improvement: call tracking, offline import, sales process. 3. Make decisions on known data; acknowledge unknown.
1. **UTM governance** — Standardized parameters, enforced at form capture, persisted through session. 2. **CRM campaign linkage** — Every lead inherits campaign; opportunities inherit from lead. 3. **Call tracking** — Unique numbers per channel/campaign. 4. **Offline import** — Trade show lists, event scans, sales-sourced leads tagged. 5. **View-through window** — Define and document (e.g., 7-day impression window). 6. **Attribution reporting** — Multi-model dashboard (first, last, linear, time-decay). 7. **Regular audit** — Monthly: coverage %, model divergence, budget shifts.