Key takeaways
  • CX is the whole relationship: marketing, sales, product, support, billing, churn, win-back.
  • It's broader than UX, and product teams routinely forget half of it.
  • A fragmented owner is the core problem; CX needs a named function.

UX gets the books, UI gets the dribbble shots, and CX gets the quarterly board paper that nobody reads carefully. That ranking is roughly the inverse of the size of the territory each one covers. UX is the experience of using the product. UI is the surface that experience is rendered on. CX is the entire relationship. It covers every touchpoint from the moment the customer first hears about you to the moment they renew, refer or churn. It is the largest of the three, and the one product teams systematically underplay because most of the surface area sits outside the product itself.

If you only own the in-product experience, you only own the middle third of CX. The first third (how the customer found you, what they were promised, what state they arrived in) sits with marketing and sales. The last third (onboarding follow-up, billing, support, renewal, win-back) sits with operations, finance, customer success and support. And because no single function owns the full arc, the seams are where customers leak. That is the central CX problem, and naming it is half of fixing it.

CX, distinguished from UX and Service Design.

Three overlapping disciplines, with different emphases:

  • UX. The experience of the user with the product. Stops at the edge of the product surface. A great UX team can build an excellent in-product experience that exists inside a chaotic wider relationship and never sees the disconnect.
  • CX. The experience of the customer with the organisation. Includes UX as a subset, plus marketing, sales, support, billing, renewals, win-back. Concerned with the lifecycle, not just the product.
  • Service Design. It designs the whole service, including the parts the customer never sees (back-stage operations, staff scripts, system handovers). Most disciplined of the three. Strongest in regulated, multi-channel, multi-stakeholder services: public sector, healthcare, banking.

The disciplines overlap, the labels travel between markets, and the borders are political more than intellectual. The useful move is to stop arguing which one is which, and start asking, at any given handover in your customer's life, who is owning this seam? Most of the time the honest answer is "nobody." That's the gap CX exists to close.

The customer lifecycle, in seven stages.

Different writers slice it differently. A defensible seven-stage version covers the territory most product teams will work with:

  • Aware. The customer learns the product exists. Channels: ads, content, word of mouth, search. Owner: marketing.
  • Consider. The customer evaluates against alternatives. Touchpoints: website, reviews, demos, sales conversations. Owner: marketing and sales.
  • Buy. The customer commits. Touchpoints: sign-up flow, contract negotiation, payment. Owner: sales (B2B), product (B2C self-serve).
  • Onboard. The customer's first experience using the product. Touchpoints: welcome emails, in-product onboarding, first-week guidance. Owner: usually contested. Product has the surface, success has the human, marketing has the email.
  • Use. The customer's ongoing relationship with the product. Touchpoints: the product itself, support tickets, status pages, release notes. Owner: product (mostly).
  • Renew. The customer continues or expands. Touchpoints: renewal emails, account reviews (B2B), upgrade prompts (B2C). Owner: success, sales, product, depending on motion.
  • Refer or Churn. The customer either advocates or leaves. Touchpoints: cancellation flows, exit surveys, referral programmes, win-back campaigns. Owner: a smear of marketing, success, and product.
Diagram · The seven-stage customer lifecycle, with ownership and the gaps
The seven-stage customer lifecycle ribbon, with ownership labels and three gap markers at handover points A horizontal ribbon of seven ink blocks, each labelled with one customer lifecycle stage: Aware, Consider, Buy, Onboard, Use, Renew, Refer or Churn. Connecting arrows run between blocks. Above each block sits an ownership label in monospace type — Marketing, Marketing and Sales, Sales, Contested, Product, Success, Mixed. Below the ribbon, three dashed marigold GAP markers sit at the handover points where ownership changes hands and customers most often leak — between Buy and Onboard, between Onboard and Use, and between Renew and Refer or Churn. A caption beneath the ribbon reads: The seams are where customers leak. CUSTOMER LIFECYCLE · 7 STAGES MARKETING MKT + SALES SALES CONTESTED PRODUCT SUCCESS MIXED Aware stage 01 Consider stage 02 Buy stage 03 Onboard stage 04 Use stage 05 Renew stage 06 Refer or Churn stage 07 GAP handover GAP handover GAP handover The seams are where customers leak. CX is the entire arc. UX is one block in it. Product teams who own only Use forget Onboard and Renew — which is where the churn maths actually lives. Aware → Consider → Buy → Onboard → Use → Renew → Refer or Churn.
Seven stages, three big gaps. The product team usually owns Use cleanly and contributes to Onboard and Renew. Marketing owns the start. Nobody, in most companies, owns the seams between them. That is where the customer's experience actually compounds.

Where product teams forget CX.

Three predictable places. None of them are in the product itself. All of them are where the product team's writ runs out and somebody else's begins, and where neither side has written down the handover.

  • The post-purchase email. The customer pays. The product team has an account creation flow ready. But the actual confirmation email (the one in the customer's inbox sixty seconds later) is owned by marketing's email tool, written in a different voice, designed in a different template, possibly contradicting the in-product copy. The customer's first impression after spending money is "I'm not sure I bought what I think I bought."
  • The renewal nudge. A B2B customer is six weeks from renewal. The product team has analytics showing the customer is healthy. The success team has a call scheduled. Marketing has an automated renewal email queued. Nobody has compared the three messages. The customer receives three contradictory things in a fortnight and reads it, correctly, as organisational confusion.
  • The cancellation flow. The customer decides to leave. They open Settings, click "Cancel subscription," and are taken to a flow nobody on the current team wrote. It's a defensive dark-pattern from three years ago, designed by someone who has since left, that talks down to them. It is the last thing the company says to that customer. It almost guarantees they will not come back.

These aren't UX problems. The screens, individually, may be fine. They are CX problems. The experience across the system, owned by nobody in particular, has degraded.

Marketing, product and support all touch CX. Often, no one owns it. That's the gap.

The "fragmented owner" problem.

Most organisations have a CRO who cares about acquisition, a CTO who cares about reliability, a CPO who cares about the product, a head of customer success who cares about renewal, and a head of support who cares about ticket volume. Each is rational inside their scope. Each is partly responsible for CX. None of them is single-threadedly accountable for the customer's experience of the company.

This is the structural problem CX exists to surface. The functional org doesn't map cleanly onto the customer's experience. The customer moves through the arc once. The organisation hands them off six times. Unless someone is watching the seams, the customer's experience is whatever the cracks add up to.

The role of CX as a function.

The fix that big organisations land on is to create a CX function: a Chief Customer Officer (CCO), a CX team, or at minimum a CX lead. Its remit is exactly to own the arc and the seams. Done well, the CCO doesn't try to take over the touchpoints (marketing keeps marketing, product keeps product). Instead, they own three things: the customer journey map (see N°12), the cross-functional metrics (NPS, CSAT, churn cohorts), and the seam contracts — who is responsible for the message at each handover, what the standards are, and what triggers escalation.

Smaller organisations can't justify the function and often don't need one. What they do need is the same job done by somebody — usually the CEO at very small scale, the CPO at medium scale, a head of CX or success at growth scale. The job exists whether the title does or not. When the job is unowned, the seams degrade by default.

Where CX fails most often.

A short, predictable list:

  • The onboarding seam. Marketing promises X. Sales sold Y. The product delivers Z. The first email says W. The customer is told four different things in their first week.
  • The first support interaction. The product is great. The support response is slow, generic, or contradicts the in-product help. The relationship trips before it has started.
  • Billing. Pricing change, upgrade, downgrade, accidental over-charge. Billing is the touchpoint product teams understand least and the touchpoint customers feel most viscerally. If the billing flow is opaque, every other strength of the product is discounted.
  • The cancellation flow. Already named. The last impression is the longest-lived.
  • Win-back. The customer left. Sixty days later, somebody sends a generic "we miss you" email. The customer can tell it's automated, that nobody noticed why they left, that no specific change is being communicated. Win-back, done badly, confirms the decision to leave.

How we use this at Product Pieces.

When a partner team's metrics tell a story the in-product analytics can't explain (churn that doesn't match feature usage, renewal slippage that doesn't match engagement, NPS that drops between cohorts for no in-product reason), the issue is usually CX, not UX. The fix is to walk the seven stages with the team, name the owner at each step, and most importantly name the seam-owner between each pair. Almost always, three of the six seams have no owner. The work is to put one there.

That is rarely a product engineering project. It is a small operating-model intervention: a half-day mapping session, a written seam contract, a recurring forty-minute meeting between three people who hadn't been meeting before. Cheap to do, and surprisingly hard to start, because no individual function naturally raises the question. CX, as a piece of practice, is the question being raised on purpose.

Next issue: N°11, Personas. The artefact most teams write once and then ignore.