"Our martech feels broken" is a feeling, not a finding. An audit turns it into a list: what is configured, how the team actually uses it, and what each gap costs. This is the framework I use. It works on any marketing automation platform because the failure modes are the same everywhere.
The six areas
1. Lead scoring and quality
Start with the MQL definition. Is it still what sales agreed to, and does anyone remember agreeing? Then the model itself: how much weight is behavioral versus demographic, whether scores decay, and whether the threshold has been touched since launch. The test that matters is simple: are MQLs turning into opportunities at a rate sales accepts? If not, the model is a wish list.
2. Lead and contact assignment
Routing is where good leads go to die. Check that leads reach the right owner, how long that takes, and what the logic is (territory, product line, round-robin). Then the question most teams have never answered: what happens when the owner is out? If the answer is "nothing," you have a backlog you cannot see.
3. Data integrity and sync
Recurring CRM sync errors, fields mapped wrong or not at all, duplicates with no merge rules, and nobody sure who is allowed to edit which field. This area usually produces the longest list and the highest-impact fixes, because every downstream report depends on it.
4. Campaign and program management
Naming conventions (consistent, and scalable past one person), program templates for the common campaign types, an archive strategy so the active folder is not five years deep, and whether programs are documented well enough to hand off.
5. UTM and source tracking
One UTM structure across every channel, parameters flowing into the analytics and attribution reports, first-touch and multi-touch sources that reconcile, and a central place links are built so nobody invents a new utm_medium on a Friday.
6. Operations and documentation
Written procedures for the common workflows, change tracking on critical assets, a user and permissions review, and integration health: API limits, rate limiting, error logs. This is the area that decides whether the other five stay fixed.
Sort findings by effort, not by how much they annoy you
An audit that ends in forty undifferentiated recommendations gets filed. Group them.
| Horizon | Typical items |
|---|---|
| Quick wins 1–2 weeks | Fix critical sync errors · update outdated scoring criteria · basic deduplication rules · a simple dashboard so stakeholders can see the instance at all |
| Medium-term 1–3 months | Persona-based nurture sequences · comprehensive reporting dashboards · standardized program templates · protections against hitting API limits |
| Strategic 3+ months | Full lead lifecycle redesign · advanced attribution modeling · predictive scoring · cross-platform integration architecture |
Attribution and ABM only work once the data underneath them is trustworthy. Strategic items built on an unaudited foundation get rebuilt within a year.
What the audit should hand over
- A current-state assessment: what is configured and how it is actually used, kept separate
- A gap analysis with prioritized recommendations
- A quick-win checklist the team can start on Monday
- A strategic roadmap with timelines and dependencies
- Documentation templates and the procedures the fixes depend on
How to run it yourself
- Walk the six areas in order. Write down facts first, interpretations second. Keep them in separate columns.
- For each gap, note the business impact, the effort, and what it depends on.
- Sort into the three horizons. Resist promoting items because they are interesting.
- Present the quick wins with owners and dates. Present the strategic items as decisions, not tasks.
If you would rather have it done for you, this framework is the basis of my Marketo Instance Audit, which scores an instance across ten dimensions and hands back the remediation roadmap.
Want this done in your instance? Most engagements start with an audit.
Book an intro call →