← Selected work
Work
Customer Operations

Customer Experience Operating Model

What best-in-class should actually mean operationally, how it would be measured, and where the investment should go.

Operating ModelService StandardsMeasurement

Sanitized for publication. Company names, customer identities, and proprietary figures have been removed.

01

Purpose

The company talked about best-in-class customer experience without a shared definition of it. Everyone agreed it mattered. Nobody could say what it required, what it cost, or how we would know when we had it. This work existed to turn a phrase into an operating model.

02

The problem

Support was being judged on responsiveness alone. That left most of the experience unmeasured: what happened before a customer ever contacted us, what onboarding taught them, whether the product explained itself, and whether anyone noticed when a customer quietly stopped using it. The question was not how to answer faster. It was which parts of the experience deserved to exist at all, and who owned each one.

03

The thinking

  • Start with definition before measurement. If accessibility, accuracy, speed, personal attention, and consistency are the qualities that matter, each one needs an operational meaning, not an adjective.
  • Separate the experience into proactive, reactive, and self-service. Most organizations only staff the middle one, then wonder why volume never falls.
  • Treat onboarding as the highest-leverage moment. The first two weeks decide whether a customer ever builds a habit, and habits decide retention.
  • Look outside the category. Retail, streaming, and banking had already solved parts of this. Borrow the mechanics rather than the language.
  • Tie every proposed capability to a metric that already existed or could reasonably be collected. A capability nobody can measure will not survive its first budget cycle.
04

What was created

  1. 01

    A working definition of service quality

    Five dimensions with operational meaning behind each one, so that quality could be discussed without relying on adjectives. Accessibility meant every channel usable by everyone. Accuracy meant correctness in troubleshooting and follow-up, not just tone.

  2. 02

    Channel and response standards

    Target response and resolution windows by channel and by complexity, along with first-contact resolution targets. Stated as commitments the team could staff against rather than aspirations.

  3. 03

    A measurement set

    Satisfaction, effort, promoter sentiment, first-contact resolution, self-service rate, adoption, and churn, framed as one instrument panel instead of separate departmental scoreboards.

  4. 04

    An onboarding model

    A short in-product survey feeding a tailored path, a product walkthrough, an equipment walkthrough, and opt-in accountability nudges when activity stalled. Modeled on how well-run apps in adjacent categories handle a first session.

  5. 05

    A self-service architecture

    Knowledge base, guided troubleshooting, video, and in-product help positioned as the first layer rather than the overflow layer, with topic usage data deciding what gets rewritten or retired.

  6. 06

    A proactive layer

    Order status communication, inactivity outreach, scheduled check-ins, and post-purchase follow-up. The intent was to reach customers before they had a reason to reach us.

05

Selected artifacts

Figure 01 — Experience operating model
ProactiveORDER STATUS · INACTIVITY OUTREACH · CHECK-INS · POST-PURCHASE FOLLOW-UPSelf-serviceKNOWLEDGE BASE · GUIDED TROUBLESHOOTING · VIDEO · IN-PRODUCT HELPReactiveCHAT · EMAIL · PHONE · ESCALATION PATHMeasurementONE PANELVOLUME REMOVED UPSTREAM NEVER BECOMES A CONVERSATION
Three layers of service, each with its own staffing model. Proactive work reduces the volume the reactive layer has to absorb. Self-service decides how much of the remainder ever becomes a conversation.
Figure 02 — Service standards, by channel
CHANNELFIRST RESPONSERESOLUTIONLive chatEmailPhoneComplex caseSTRUCTURE SHOWN. PUBLISHED TARGETS WITHHELD.
Standards defined per channel and per complexity, so a commitment could be staffed rather than promised. Values shown are illustrative of the structure, not the published targets.
06

What it informed

  • Gave the support organization a definition of quality it could hire, train, and report against.
  • Set the frame for later work on self-service and in-product help.
  • Shaped the onboarding conversation with product, particularly around what a first session should teach.
  • Made the case for measuring effort and self-service rate alongside satisfaction.
07

Reflection

The strongest part of this work was the definition. The weakest was assuming that agreement on a definition would produce funding for it. If I did this again I would pick two capabilities, prove them with a small measurable pilot, and let the results argue for the rest. A complete model is easy to admire and easy to defer.

See the health and retention half of this in practice
Creating Customer Health & Retention Visibility →
08

Related work