Pauline Narvas wrote a great community piece asking a deceptively simple question:
What job is community being hired to do?
Her answer maps community across the GTM funnel: awareness, acquisition, activation, retention, advocacy, and expansion.
That framing clicked for a lot of people. Community is not one job anymore. It shows up as marketing, DevRel, customer education, support, advocacy, events, growth, feedback, and customer success depending on what the company needs from it.
I think there is a CX version of the same question:
What job is community being hired to do inside the customer experience?
My answer is this:
Community is the routing layer between customer intent and the systems meant to serve it.
Community is where customers say what they are trying to do before the company has translated it into a ticket category, a docs page, a roadmap item, a support macro, or an AI answer.
That makes community much more than a place where people gather.
It is an intent surface.
Community reveals intent in the customer’s own words
Support tickets are already translated. By the time a customer files one, they have usually accepted the company’s structure: severity, product area, category, environment, reproduction steps.
Community is different.
In community, customers ask messy questions:
- Is anyone else seeing this?
- Am I using this wrong?
- How are people handling this workflow?
- Did something change?
- Is there a better way to do this?
- Why does the product behave like this?
- What should I do next?
That language is gold.
It shows the distance between what the customer is trying to do and what the system currently gives them.
A mature CX system should not treat those posts as isolated conversations. It should treat them as signal.
The repeat question is a knowledge gap.
The workaround is a product signal.
The confused launch thread is a communication gap.
The peer answer is a candidate for self-service.
The unanswered post is a trust leak.
The heated thread is a state change: customers moving from curious to frustrated, from blocked to doubtful, from hopeful to at-risk.
Community shows you that movement in public.
Community should feed self-service
The biggest mistake companies make with community is letting useful answers die in the scrollback.
Someone asks a question. A power user answers. A support engineer adds context. The customer gets unstuck. Everyone moves on.
Then next week, another customer asks the same thing.
That is not a community problem. That is a systems problem.
If a community answer helped one customer, the system should ask:
- Should this become a knowledge article?
- Should this be linked from onboarding?
- Should this be surfaced in search?
- Should this train the support bot?
- Should this become a product tooltip?
- Should this tell product something is unclear?
- Should this be part of launch communication next time?
Community is not just where answers happen.
Community is where reusable answers are discovered.
The job is to turn that discovery into infrastructure.
Community, self-service, and knowledge are one system
This is why I keep coming back to the same triad:
community, self-service, and knowledge
Community reveals what customers are actually trying to do.
Self-service gives them the path to do it without waiting.
Knowledge gives the system a durable source of truth.
When those three are disconnected, customers pay the tax.
They find the answer in Discord but not in search. They read a help article that does not reflect the workaround everyone uses. They ask support a question that was already answered by a peer. They get an AI response that sounds confident but cannot see the community context where the real answer lives.
The company thinks it has multiple help surfaces.
The customer experiences one broken path.
Community is also a launch surface
This is where Pauline’s funnel framing meets the CX Gate.
A launch is not just a product event. It is a community event.
When something changes, customers do not only need a changelog. They need to understand:
- what changed
- why it changed
- what it means for them
- what they should do next
- where to ask questions
- where feedback goes
- whether the company is listening
Community is one of the best places for that.
A good launch loop should connect:
- changelog
- docs
- in-app communication
- email or outbound messaging
- community discussion
- support routing
- feedback capture
- roadmap updates
Without that loop, the launch becomes a broadcast.
With that loop, the launch becomes a learning system.
Community should make AI better, not less human
AI makes community more important, not less.
If an AI support experience is disconnected from community, it misses the customer’s real language. It misses the workaround culture. It misses the way people actually describe problems before the company has formalized them.
But I do not think the answer is to dump AI bots into every community thread.
Community should stay human.
The better pattern is AI behind the scenes:
- summarize repeat questions
- suggest likely answers to moderators
- identify knowledge gaps
- detect emerging product friction
- route solved threads into a knowledge queue
- cluster launch confusion
- help community teams see what is becoming systemic
Let AI help the humans see the pattern.
Do not let it flatten the space where trust is built.
The CX job of community
So what job is community being hired to do in CX?
Not just engagement.
Not just support.
Not just advocacy.
The CX job of community is to connect customer intent to answers, access, and action.
- Answers: peer replies, expert responses, knowledge articles, AI-ready source material.
- Access: discovery paths, onboarding context, launch communication, “where do I go?” guidance.
- Action: escalation, feedback, product fixes, roadmap signal, customer recovery.
That is why community touches everything.
It is not a department.
It is a sensing layer.
It is a routing layer.
It is a trust layer.
And when it works, community becomes part of the product behind the product: the system that helps customers get unstuck, adopt faster, and stay.
The important question is not only:
What job is community being hired to do?
It is:
What does your customer experience become capable of when community is connected to the rest of the system?