Free resource · Self-assessment
The marketing systems scorecard
Twenty checks across five layers of your stack. Score each one right here, watch your total update live, and walk away knowing exactly where your operations are leaking.
20
Checks across data, routing, lifecycle, attribution & AI
5
Layers, scored 0 to 4 each
10 min
To run it against your own stack
How it works
Score each check 0 to 4. Be honest.
This is the diagnostic we run at the start of an audit, productized so you can run it yourself. Tap a score on every check below and your layer subtotals, running total, and result band update as you go. Score each check: 0 = does not exist, 1 = exists but broken, 2 = works but manual, 3 = automated but untrusted, 4 = automated and trusted. Max score is 80.
The temptation is to grade generously. Resist it. The score is only useful if it reflects the system you actually have on a bad Tuesday, not the one you described in a deck. A 2 you can prove beats a 4 you are hoping for. Prefer paper? Download the printable worksheet →
Layer 1 · Data model 0 / 16
Objects and relationships are defined
Contacts, companies, deals, and any custom objects associate correctly. You can answer "which accounts have open pipeline" without exporting to a spreadsheet.
Properties have one owner and one meaning
No two fields competing to mean the same thing. No free-text field where a picklist belongs. Required fields are actually required.
Lifecycle stages are defined and exclusive
Every stage means one specific thing, and a record can only sit in one. "MQL" means the same thing to marketing and sales.
Data hygiene runs on a schedule
Duplicates, missing fields, and formatting drift are caught weekly by automation, not discovered during a quarterly fire drill.
Layer 2 · Lead routing 0 / 16
Routing logic is documented and current
You can name the rules without opening the tool, and the doc matches what is live.
Strategic accounts bypass the generic pool
Named accounts and high-fit segments route to the right rep before round-robin ever runs.
Speed-to-lead is measured, not assumed
You know your median first-touch time from a real timestamp, not a guess.
No-touch leads get re-routed
An SLA breach reassigns the lead and alerts a manager automatically.
Layer 3 · Lifecycle & nurture 0 / 16
First-touch fires in minutes, not hours
New MQLs get an automated first response and a rep task immediately on stage change.
Nurture branches on behavior, not a calendar
Sequences react to what the prospect did, and booked meetings suppress active nurture.
Closed-lost has a real second act
Lost deals re-enter nurture or get a re-engagement task based on the loss reason.
Customer lifecycle is automated too
Onboarding handoff, renewal warnings, and expansion flags fire without anyone remembering to.
Layer 4 · Attribution & reporting 0 / 16
Attribution is stamped at conversion
First- and last-touch source write to the record automatically. You are not reconstructing it later.
Time-in-stage is captured on every move
SLA and conversion reporting run on real timestamps.
Numbers match across tools
The figure in the CRM matches the figure in the dashboard matches the figure in the board deck.
Leadership trusts the dashboard
Nobody rebuilds the "real" numbers in a side spreadsheet before the meeting.
Layer 5 · AI & tooling 0 / 16
AI tools are real software, not pasted prompts
Anything "AI-powered" in your stack runs as an instrumented workflow you can monitor, not a prompt living in someone’s doc.
Enrichment runs automatically and is trusted
New records get enriched on create, and the enriched fields are clean enough to segment and route on.
There is a human gate where it matters
AI drafts and classifies; a person approves anything that sends externally or changes a deal. The gate is defined, not improvised.
You own and can run the tooling
No black-box vendor lock-in. If the vendor vanished tomorrow, you could still operate.
How to read your score
Total it up. Then read the band.
65–80 — Properly built. Your operations layer is doing its job. The work now is optimization at the margins, not rescue. Focus on the lowest single layer and tighten it.
45–64 — Working, but manual. The system runs because people are holding it together. That works until someone leaves or volume doubles. Find the layer scoring lowest and automate it before it becomes the bottleneck.
25–44 — Leaking. You are losing leads, time, and reporting confidence in ways you can feel but cannot yet name. The data model is almost certainly the lowest layer. Start there; everything above it depends on it.
Under 25 — Rebuild territory. Patching individual checks will not hold because the foundation is not there. This is a model-first rebuild, scoped tightly so it does not become an open-ended project.
One rule that overrides the total: your real score is your lowest layer, not your average. A stack that scores 4s on reporting and 0s on data is not a 2. It is a 0 with expensive dashboards on top.
LSR Marketing
Pipelines you can run. Systems you can read. A remote-first consultancy for B2B teams that need their stack to actually work.
Remote-first · Global clients
Services
Marketing Operations & CRM
Growth & Demand Generation
Websites & Search
AI Business Solutions
Fractional CMO
Get Started
Free 30-minute audit. We map your stack and tell you where to start.