advancement-playbook.md
Advancement playbook
The actions that move a company from each level to the next, in dependency order, and how the scope of the work scales with maturity. Plan one level at a time.
Actions by level
From Level 1 to Level 2
- Write the ICP down with firmographic criteria, personas and disqualifiers.
- Make the CRM the single source of truth; retire side spreadsheets.
- Define basic qualification criteria (budget, authority, need, timing, or an equivalent).
- Create the first outreach templates.
From Level 2 to Level 3
- Add technographic and behavioral signals to scoring.
- Add a second outreach channel, such as social or phone.
- Define the intent signals and triggers that should move an account up the list.
- Start a weekly optimization cadence.
From Level 3 to Level 4
- Process intent signals as they arrive rather than in batches.
- Deploy trigger-based workflows, each with a named owner.
- Track performance by persona.
- Feed sales feedback and win/loss outcomes into scoring calibration.
From Level 4 to Level 5
- Deploy lead-level messaging, with human approval on top-tier accounts.
- Implement cross-channel attribution.
- Track win rates by persona and segment.
- Make scoring recalibrate itself from outcomes.
Scope by starting level
| Starting level | Focus | Typical duration |
|---|---|---|
| Level 1 | Foundation build: ICP, CRM, basic sequences | 8 to 12 weeks |
| Level 2 | Operationalization: scoring, second channel, optimization cadence | 6 to 10 weeks |
| Level 3 | Automation: workflows, triggers, advanced analytics | 4 to 8 weeks |
| Level 4 | Compounding: AI integration and self-improving models | Ongoing |
Most companies assessed honestly start at Level 1 or 2. Level 4 within six months of starting is a realistic target; skipping a level is not.
Sequencing judgment
- Fix the binding gap first, even when it is unglamorous. A data cleanup that unblocks three dimensions beats a new channel that improves one.
- Match investment to level. Predictive scoring on a Level 2 data foundation is over-engineering; the model learns from bad inputs and its confidence hides that.
- Use maturity as a readiness test for any purchase. Before a tool or service is bought, identify the level it assumes. If the company is not there, the purchase waits or the plan closes the gap first.