Teams often compress stickiness into retention. If users return, the product is sticky. If they leave, it isn’t.
Retention counts who came back, not why they stayed. Did trying feel low-risk? Did the right person recognize the situation? Has the product earned a place in the recurring job, or a recommendation? Each answer belongs to a different stage.
Stickiness has a sequence
Knurture uses six stages: Resistance, Recognition, Relief, Ritual, Retention, and Reputation. Each stage carries a different user question and a different product responsibility. The first three are fast sticky. The last three are slow sticky.
Skip one and growth gets expensive. Marketing has to replace users the product never truly kept.
Sequence map
The Six Rs
Map what each stage has to do and where users stall.
- Resistance: reduce the risk of movement with a narrow first step, familiar objects, visible proof, migration help, and reversibility / check whether it’s being treated as objection handling instead of design material.
- Recognition: name the trigger, struggle, and desired progress / check whether generic category language is forcing translation.
- Relief: remove a burden before asking for belief / check whether a fast output is creating heavy verification.
- Ritual: attach to a recurring trigger, remember useful context, and make re-entry easier each time / check whether returns come from the job or from notifications.
- Retention: become more useful through decisions, preferences, objects, history, and trust / check whether the product is trapping data or punishing departure.
- Reputation: help users do work they’re proud of / check whether the recommendation feeds recognition for the next user.
Resistance: even obvious pain can feel safer than change
The starting point isn’t an empty market. It’s the status quo.
Users have habits, workarounds, anxiety, sunk cost, and social arrangements.
The product has to understand why staying makes sense. Reduce the risk of movement with a narrow first step, familiar objects, visible proof, migration help, and reversibility.
Resistance isn’t objection handling. It’s the design material for the switch.
Recognition: the right user should recognize a lived situation
Generic category language forces translation. Precise framing names the trigger, struggle, and desired progress: “That’s me. That’s what’s happening. That’s the better state I want.”
Recognition is a filter. It attracts people with the active job and lets everyone else move on.
Relief: the product has to remove a burden before asking for belief
Relief might be clarity, saved effort, reduced risk, or a finished task the user no longer has to carry. It should happen quickly enough to make the promise credible and transparently enough to feel safe.
A fast output that creates heavy verification isn’t relief. It’s shifted labor. If AI is involved, the output may feel impressive just because it compressed something that used to take time—and the brain mistakes compression for competence.
Example
The Six Rs: relief that never became a ritual
A user liked the product and still didn’t come back. The output was good and the value was real, but the next hiring moment wasn’t clear.
- Resistance: Not known yet: What made trying the product feel risky?
- Recognition: Not known yet: What hiring situation made the user look for help?
- Relief: real. The value was real.
- Ritual: likely broken. The output was impressive, but the next hiring moment wasn’t clear. It didn’t attach to a situation, so when the situation returns, the product doesn’t come to mind.
- Retention: never reached. The user didn’t come back.
- Reputation: Not known yet: Did the user recommend the product to someone else?
Resist the easy explanations—weak onboarding, a missing reminder. Often the product created a satisfying moment without owning a recurring job.
The last three Rs turn a first win into a default
A first successful job earns another chance. Continued use still depends on repeated value, reliable recovery, and a place in the user’s actual workflow.
Ritual
A good first experience becomes sticky when it attaches to a recurring trigger.
The user doesn’t return because the interface sent a notification. They return because the job came back: the weekly review, the new project, the customer call, the publishing cycle, the month-end close. A nudge says, “Don’t forget this exists.” A re-hire says, “I need this again.”
The product should remember useful context, reduce re-briefing, and make re-entry easier each time.
Retention
Retention deepens when the product becomes harder to replace for good reasons.
Context accumulates. Workflows fit. Collaboration grows. The system becomes more useful through decisions, preferences, objects, history, and trust—not by trapping data or punishing departure.
The user stays because leaving would mean giving up earned progress.
Reputation
The final stage is identity and advocacy.
The product helps users do work they’re proud of. They recommend it because the recommendation reflects well on their judgment. The product’s behavior becomes its brand.
Reputation feeds recognition for the next user. The loop begins again with borrowed trust.
Diagnose the broken R
Ask where users stall:
0 of 6 done
Don’t add habit mechanics to a product that hasn’t created relief. Don’t buy more traffic for a recognition problem. Don’t call lock-in retention.
