How to Score Your.

The fastest way to score AI marketing maturity is to answer 15 specific questions, each mapped to a weighted operating dimension. Most teams skip that step and hold a strategy meeting instead. A meeting produces narrative. The 15 questions produce a number between 0 and 100, a named tier, and three concrete moves that follow directly from where the gaps are.

Why does scoring AI marketing maturity take longer than it should for most teams?

Two patterns slow most teams down. First, they conflate AI maturity with AI adoption. A team using six AI tools may have lower maturity than a team using two, if the two-tool team has a named buyer path, a clear workflow owner, and a working measurement brief. Counting tools is faster than answering hard questions about who owns what. But it tells you nothing useful.

Second, most assessments are too long. Vendor assessments run to 40 or 60 questions across categories the team has never thought about. The exercise loses an afternoon and returns a PDF full of recommendations the team is not resourced to act on. Without a time constraint and a clear decision rule, the scoring session becomes another meeting about what AI should mean for the company.

The AI Marketing Maturity Benchmark strips the framework to 10 weighted dimensions and 15 questions. Each question has four answer levels: 0 means the capability is absent, 1 means it exists informally, 2 means it runs but is still fragile, and 3 means it is owned, documented, and visible in operating data. The weighted sum converts to a score from 0 to 100.

The useful reason to score the team is not to prove that the company is ahead. It is to choose where to inspect first. If the score points to workflow ownership, the next conversation is about the handoff. If it points to revenue measurement, the next conversation is about the metric, baseline, owner, result window, and caveat. If it points to attribution, the next conversation is about the fields that survive from click to CRM to opportunity.

What does the benchmark score across its 10 dimensions?

The benchmark scores 10 operating dimensions. Each carries a weight in the 0-100 total. The weights reflect the typical order in which capability gaps produce visible revenue drag. A broken workflow orchestration problem costs the team faster than a vendor consolidation gap.

The five path-moving dimensions

Workflow orchestration (15 points). Can a lead move from a named trigger to a booked next step without being carried by a person? A Level 3 means the trigger is named, the owner is assigned, the handoff rules are written, and exceptions have a destination. A Level 1 means AI helps one step but a person still connects the pieces.

Tool stack maturity (12 points). Do the AI tools in the stack pass the right fields to the right systems at the right time, or do humans bridge them? A Level 3 means the stack is orchestrated end to end without manual relay. A Level 1 means the team uses two to four specialized tools that do not talk to each other.

Revenue measurement (12 points). Can the team tie AI-assisted work to a named buyer path, a documented baseline, an owner, a result window, and a caveat? A Level 3 means those pieces are written and reviewable. A Level 1 means the team tracks activity counts but cannot say what changed for the buyer after AI entered the workflow.

Reporting automation (10 points). Does reporting give the team a signal early enough to act this week, or is the report rebuilt manually after the month closes? A Level 3 means reporting runs from owned fields and is ready before the next pipeline call.

Attribution fidelity (10 points). Can the CRM explain why a qualified opportunity moved without a separate spreadsheet story? A Level 3 means attribution rules are consistent enough to compare one path before and after a workflow change. A Level 1 means attribution depends on whoever built the last report.

The five system-support dimensions

Data integration (9 points). Do the fields the buyer path needs appear in the systems that act on them? A low score usually means the same buyer lives in several tools with different names, stages, or source records.

Team skills and training (8 points). Does the team run the workflow the same way under normal pressure? A written process matters only when people follow it without the most experienced person in the room.

AI governance and policy (8 points). Are approved use cases, customer-facing review rules, prohibited inputs, and escalation paths written down? A Level 3 means governance keeps useful work moving while stopping risky work early.

Budget discipline (8 points). Does each AI tool have a named owner, a renewal rule, and an evidence requirement? A tool should earn its renewal by helping a named path move, not by sounding plausible at the next finance review.

Vendor consolidation (8 points). Have overlapping tools been removed or merged into one workflow? A low score usually means the team has several subscriptions doing similar jobs while the actual handoff is still manual.

How do you run the 12-minute scoring session?

Open the benchmark at /benchmark. You can complete it alone in 12 minutes. For the three or four questions where your honest answer and your ops lead's honest answer will differ, invite them to sit next to you. Disagreements on those questions are usually the most useful data the session produces.

Before you start: two rules

Answer from the current state of the workflow, not the state it is intended to reach. If the process is documented but no one follows it consistently, that is a Level 1, not a Level 2. If the tool is integrated but someone still reviews the output before it moves downstream, that is a Level 2, not a Level 3. Optimistic scoring produces a flattering number and a useless map.

The credibility test for any Level 2 or Level 3 answer

Ask: if the most careful person on the team was unavailable for one week, would the workflow produce the same output? If the answer is no, lower your answer by one level. The benchmark is useful only when it shows where the system actually holds together and where it depends on a particular person's attention.

During the session: read all four options before selecting

Each question lists four answer levels. Teams most often overselect Level 2 ("exists, still fragile") when the honest answer is Level 1 ("informal, person-dependent"). Read all four options before selecting. The phrasing difference between Level 1 and Level 2 is usually the difference between "AI helps one step" and "some steps run end-to-end with checkpoint review." That is a meaningful operational difference and the score should reflect it.

When to lower your answer by one level

For any question touching attribution, ownership, or handoff quality, the honest answer comes from the CRM record, not from memory. If you cannot pull up the record in the room, use Level 1 as the working answer and verify it against the record before finalizing your score.

After the session: read the three gap items

The benchmark output shows your total score, your tier, and your three largest point gaps sorted by dimension weight. The first item in the gap list is the dimension where moving up one level adds the most points to your total. That is where to start.

What does your score tell you by tier?

Below 35: Behind

A Behind score almost always means at least two of the three highest-weight dimensions (workflow orchestration, tool stack maturity, revenue measurement) are at Level 0 or Level 1. The most common Behind pattern is a team using AI tools actively, but with no written workflow owner, no documented trigger, and no measurement that ties the work to a buyer-path movement.

The right move is not to buy another tool. It is to name one buyer path, assign one owner for that path, and write a measurement brief with five fields: the path, the baseline, the cost included, the result window, and the decision rule. That brief, written and shared, is what moves a Behind team to Catching Up faster than any new software purchase.

35 to 59: Catching Up

A Catching Up score means the infrastructure exists but is fragile. Workflows run but depend on a careful person staying in the loop. Reports exist but require manual assembly after the period closes. The most common gap at this tier is the distance between tool capability and workflow documentation: the tools can do more than the team has written processes for.

The right move is to take one workflow from informal to documented, then measure whether the documented version runs more reliably when the most careful person on the team is unavailable. If it does, the dimension moves from Level 1 to Level 2. If it does not, the documentation exposed a fragility that was already costing the team, only now it has a name.

60 to 79: On Pace

An On Pace score means the team has at least one well-owned path and a working measurement brief. The gap at this tier is usually in attribution fidelity or reporting automation: the team can see what happened, but the report is still built by a person rather than exported from a field. The team knows the workflow works; it just needs a cleaner record of why a deal moved.

The right move is to close the attribution gap first: check whether utm parameters are captured at form submission and passed to the CRM contact record.

80 and above: Ahead

An Ahead team has at least one fully owned, CRM-visible buyer path with a decision rule the team reviews on a named cadence. The point is not the tier label. The point is that another person can inspect how the path works, where the evidence comes from, and what decision the team makes when the result window closes.

The right move is to add a second path using the same measurement pattern rather than trying to improve dimensions that are already at Level 3. The Level 1 vs Level 4 breakdown describes what the behavioral signals look like when a team moves from one tier to the next.

Which dimension has the highest impact on a low score?

Workflow orchestration carries 15 points, the highest weight in the benchmark. A team stuck at Level 1 on workflow orchestration, meaning a person still carries the handoff between steps, leaves up to 10 points unclaimed on that dimension alone. Ten points is enough to move from Behind to Catching Up, or from Catching Up to within reach of On Pace.

Revenue measurement (12 points) and tool stack maturity (12 points) are the next highest weights. Teams that fix workflow orchestration first, then address measurement, tend to see the fastest score movement, because workflow ownership is a prerequisite for any measurement report worth reading. The full breakdown of all 10 dimensions explains what each level looks like in practice.

Where do teams consistently overrate themselves on the benchmark?

Workflow orchestration

The most common overrate is selecting Level 2 ("some workflows run end-to-end with checkpoint review") when the honest answer is Level 1 ("mostly manual, AI only helps one step"). The test is concrete: remove the most careful person from the workflow for five business days. If the handoff degrades, the workflow is at Level 1, regardless of what the documentation says. The benchmark catches this when the team uses the one-level-lower rule before confirming their answer.

Revenue measurement

Teams with dashboards often select Level 2 or Level 3 in measurement. The honest question is narrower: can the team state the buyer path, the documented baseline before AI entered the workflow, the cost included in the review (tool cost plus labor), the result window, and the decision rule? If any of those five pieces is missing, the dimension should stay at Level 1, regardless of how polished the charts look. A dashboard that cannot force a keep, repair, expand, or stop decision is still operating as decoration. See The AI Budget Cut Warning Signal for what a complete measurement brief includes.

What should you do with your score after the session?

Take the first item in your dimension gap list and ask one question: what is the smallest change that would move this dimension from its current level to the next level? The answer is almost never a new tool purchase.

For most Behind and Catching Up teams, the answer is a one-page workflow contract: a document that names the trigger, the owner, the field map, the exception route, and the stop rule. That document takes two hours to write. It costs nothing. And it is the thing that makes the next round of AI investment inspectable rather than invisible. Run the Revenue Audit if the gap list points to measurement or attribution and the team cannot yet name the specific path where the gap is costing them.

The score shows whether the team has a working buyer path or only AI-assisted tasks around one. That distinction matters more than how many tools are in use.

How often should a B2B SaaS marketing team rescore?

Rescore every 90 days, or immediately after you deploy a workflow change in one of your top three gap dimensions. Scoring more often than monthly produces noise without new underlying data. Scoring less often than quarterly misses the signal that a newly deployed workflow has quietly regressed to its previous fragile state.

If you run the benchmark before a planning review, the score gives you a specific number rather than a narrative. A 54 with a named gap in attribution fidelity is more useful than "we are actively investing in AI capabilities."

Methodology

The scoring framework in this post comes directly from the Conversion System AI Marketing Maturity Benchmark, an open diagnostic that scores 10 weighted operating dimensions across 15 questions. The tier thresholds (Behind below 35, Catching Up 35 through 59, On Pace 60 through 79, Ahead 80 and above) and dimension weights are defined in the production code at src/data/benchmark.ts, which is publicly inspectable in the repository.

The benchmark is a Conversion System rubric, not a third-party market claim. The dimensions, weights, question levels, tier thresholds, and recommendation logic are defined in the production source at src/data/benchmark.ts. To score AI marketing maturity for your team, run the benchmark at /benchmark. If the output points to a measurement or attribution gap, the Revenue Audit inspects the path before any sprint is planned.