Key Takeaways
The assessment takes twelve minutes. Carmen answers 30 questions about onboarding, support, customer success and knowledge. The vendor's form returns a verdict: Stage 2 of 4, “Operationalizing.” A PDF follows with a roadmap to Stage 3. It recommends a health score, a digital program and a dedicated operations hire. Carmen forwards it to her VP with a note saying it looks about right. Then she tries to answer the question her VP will ask first. If the team reaches Stage 3, what happens to renewals? Nothing in the PDF says. Nothing in any maturity model she can find says either.
That gap isn't an oversight by one vendor. It's how maturity models were built from the start. The customer operations maturity model most teams meet today is a 1990s software procurement tool with new labels and without the caveats its authors attached. This post traces where the stages came from and scores the published customer-side models on their evidence. Then it replaces the stage question with five capabilities you can check in an afternoon.
Where the customer operations maturity model came from
The five-level maturity model was written at Carnegie Mellon's Software Engineering Institute for the US government. It judged whether software contractors could deliver. The 1993 version, CMM v1.1, is still online, and its first level describes a familiar place. “Success in Level 1 organizations depends on the competence and heroics of the people in the organization.”
Anyone who has run a support team through a product launch knows that sentence. It's the part of the CMM that customer operations borrowed, and it's accurate. The two other things the authors wrote got left behind.
Customer operations is a younger discipline, and its scope is still settling. Our explainer on what customer operations is describes three competing definitions. A maturity model inherits whichever definition its author picked, so two assessments can score the same team differently without either being wrong.
What the evidence behind maturity models shows
If any maturity model has proven that moving up a level improves results, it's the CMM's successor, CMMI. In 2006 the institute published its performance results. The median gains looked strong: 34% on cost, 50% on schedule, 61% on productivity and 48% on quality, with a 4.0 to 1 return on investment.
The report is also candid about what those numbers can't carry.
This was the strongest evidence the genre ever produced, and its authors treated it as a proof of concept. Academic reviewers have gone further. A 2011 review counted “hundreds of maturity models” across domains. It summarized the criticism of them as “step-by-step recipes” that oversimplify reality and lack empirical foundation. The same review notes that stage models “tend to neglect the potential existence of multiple equally advantageous paths.” Two support organizations can reach the same result in a different order. A single ladder can't show that.
The customer-side maturity models, scored on their evidence
Several maturity models now target customer-facing teams. Each one describes stages clearly. The question for a leader who has to defend a budget is what each one can prove. Here's the scorecard, built only from what each publisher states on its own pages.
| Model | Borrowed from | Stages | Stated time to move | Outcome evidence published |
|---|
| SEI CMM (1993) | Original | 5 levels | Years per level | None in the model itself |
| CMMI results (2006) | CMM | 5 levels | Not stated | 35 self-selected organizations; 7 data points on customer satisfaction |
| TSIA Customer Success Maturity | A team development model | 4 phases | Not stated | None in the public description |
| TSIA LAER path | TSIA's LAER model | 4 phases | Not stated | “Few companies reach” the last phase |
| KCS v6 adoption | KCS practices | 4 adoption phases | Benefits in 3 to 30 months | Benefit windows, tied to how much work depends on knowledge |
Sources: SEI 1993, SEI 2006, TSIA customer success maturity, TSIA LAER phases, KCS v6 Practices Guide.
Three patterns stand out. First, the stage names are borrowed. TSIA says openly that its customer success phases adapt “a well-known team development model”: forming, storming, norming and performing. That's a model of how small groups learn to work together, not of how service improves.
Second, even the upper stages can describe weak operations. In TSIA's third phase, a CSM team “may be in place, though often working reactively and focusing on escalation management.” A team can score “norming” and still spend its week firefighting.
Third, the stages are mostly about customer success alone. Customer operations covers support, onboarding, customer success and the knowledge under all three. None of the published models spans that full scope, which is part of why our guide to customer operations team structure treats scope as the first design decision.
The two maturity signals with a published link to outcomes
Strip out the stage names and two capabilities remain that carry some published connection to results. Neither one is a level. Both are things you can inspect.
Signal 1: Customer data is unified
TSIA's 2026 customer success report is direct: “In 2026, the ability to unify customer data is one of the strongest predictors of renewal performance.” The public summary doesn't publish the statistic behind it, so treat it as TSIA's finding rather than a benchmark. It points the same way as Salesforce's 2025 State of Service survey of 6,500 service professionals. There, companies that integrated their service channel data on one platform were 1.4 times more likely to call their AI implementation very successful.
Both findings are correlations, and both are self-reported. They still say something no stage chart says: the thing that tracks with results is whether one person can see an account's requests, onboarding progress and product questions together. We cover the mechanics in unifying knowledge sources for AI support.
Signal 2: Answers are captured once and reused
Knowledge-Centered Service is the one model on the scorecard that says when benefits should appear. KCS v6 puts direct operational improvements at 3 to 9 months, with deeper benefits over 6 to 18 months and 12 to 30 months. It also states a working assumption that 80% of requests have already been answered somewhere in the organization. The guide calls that “a good average,” not a measured benchmark. Pair the scorecard with customer operations KPIs you can track.
KCS is also honest about fit. Its key factor for value is “the degree to which knowledge or experience plays a role in responding to user's or customer's requests.” For a high-tech product with a long setup and a steady stream of how-to questions, that degree is high. That's why knowledge-driven support tends to pay off first in exactly the teams that feel most stuck.
A customer operations maturity model built on capabilities
If stages don't predict outcomes, measure what does. The model below keeps the useful part of the maturity idea, a shared picture of where the team stands, and drops the ladder. Each capability is a yes-or-no test you can run on your own systems.
Capability 1: One account view
Pick a customer at random. Can one person see its open requests, its onboarding status, its last renewal note and the articles its users searched, without opening four tools? If the answer takes an hour of exports, this capability is missing. It's the capability TSIA ties to renewal performance.
Capability 2: Answers captured at the point of resolution
When an agent or CSM solves a new problem, does the fix become findable the same day? Or does it live in one person's inbox? Check the last 20 resolved requests and count how many produced or updated an article.
Capability 3: Handoffs that carry context
Follow one customer from the signed contract to the first support request. Did anyone ask the customer to repeat what they told sales or onboarding? Every repeated question is a handoff that dropped its data.
Capability 4: Measures that tie to money
Can you report gross and net revenue retention next to support volume and onboarding duration for the same accounts? A team that only reports activity counts can't show the link. Our guide to customer success metrics that change a decision covers which measures to join.
Capability 5: Changes without a ticket to IT
When the process changes, can the operations team change the form, the routing rule or the onboarding checklist itself? If every change waits in an engineering queue, the model's other capabilities decay while they wait.
Each capability is binary on purpose. Five yeses don't make a team “Level 5.” They make a team that can see its customers, reuse its answers and prove its value. And the order doesn't matter. The academic critique was right: there's more than one good path, so start with whichever capability costs your team the most today.
How to run the assessment in one afternoon
- Pull five real accounts. Mix a new customer, a renewal due this quarter and one that escalated recently.
- Run the five tests on each account. Record yes or no, and note which tool held the missing piece.
- Count the gaps by capability, not by account. A capability that fails on four of five accounts is systemic. One failure is noise.
- Price the biggest gap in hours. Estimate the weekly hours your team spends working around it, such as exports, re-asked questions or duplicate answers.
- Fix one capability, then retest in a quarter. Use KCS's 3 to 9 month window as the honest expectation for the first visible change, not a stage promotion.
A team that runs this twice a year has a maturity model it can defend. Every score points at a real system, and every change can be checked against the same five accounts.
Where the capabilities live
Most teams fail capabilities 1 to 3 for the same reason. The account record sits in the CRM, onboarding tasks sit in a project tool, requests sit in a help desk and answers sit in a wiki. A stage-based model treats that as a people problem, and prescribes a hire. The capability view shows it's a data problem.
MatrixFlows puts customer operations on one foundation. Accounts, onboarding projects, requests and knowledge articles are linked records in Matrix, so the one-account view is a filter rather than an export. Conversations in the Inbox stay attached to those records, and the question that started a thread can become an article in the same place. Operations teams change forms, fields and routing themselves in no-code apps. That covers capabilities 1, 2, 3 and 5 without a new stage label.
Back to Carmen's PDF. The better reply to her VP isn't “we're Stage 2.” It's “we pass two of five capability tests.” Then she adds the hour count on the biggest gap. That answer survives the follow-up question.
Create a Free Workspace → and build the one-account view for your five test accounts this week.
.svg)
.svg)
.svg)
