How do we build a community around our product's purpose?
Build a community around your product's purpose by finding a problem members already share, organising useful work between them and giving them genuine influence over the space. Start with a small recurring activity, name an accountable host and measure what members help one another achieve. Choose a platform after establishing that need.
Build a community around your product’s purpose by helping people work on a shared problem together. Identify a recurring difficulty they already care about, invite people with complementary experience and organise a useful exchange they would value without a product pitch. Give members influence over the work and measure the help they provide one another. The product can bring people together without becoming the centre of every conversation. Shared purpose should leave room for different methods, experiences and opinions.
For a post-seed startup building its go-to-market, start with one member need and one operating commitment. The hypothetical example throughout this page is a reporting-software company convening practitioners who want to make business reports understandable and traceable. Its proposed activities, wording and decision rules are examples, not customer cases or measured results.
What is the difference between a product audience and a community?
A product audience pays attention to a company; a product community also develops relationships, shared practices and responsibility between members. A newsletter, support queue or Slack workspace may serve people well without providing those relationships. Evaluate what happens between people before calling the space a community.
Muniz and O’Guinn’s 2001 study identifies three markers in the brand communities studied: shared consciousness, rituals and traditions, and moral responsibility. These describe social relationships, rather than the size of a contact list. The study uses ethnographic and online observations; it does not establish a growth effect for a new B2B group.
McMillan and Chavis’s sense-of-community theory offers a complementary account: membership, influence, fulfilment of needs and shared emotional connection. Use these as design questions about belonging, agency, useful exchange and accumulated experience. They are not a launch checklist that certifies a community exists.
Use this example distinction table to decide what you are actually proposing:
| Format | Main relationship | Useful purpose |
|---|---|---|
| Audience or newsletter | Reader to publisher | Learn from reliable material |
| Product support | User to accountable provider | Resolve a product problem |
| Customer user group | Customers to peers and provider | Exchange implementation experience |
| Brand community | Members to a brand and one another | Share an identity and associated practices |
| Community of practice | Practitioners to one another | Improve a shared craft or problem-solving capability |
| Open-source ecosystem | Users, contributors and maintainers | Develop and sustain shared software |
These forms can overlap. Preserve their different obligations. A customer reporting a defect should retain an accountable support route even when peers sometimes help. A practitioner using another product should know whether they are welcome before contributing.
Fournier and Lee, drawing on Jump Associates, distinguish pools, webs and hubs: affiliation through shared ideas, relationships between members, and attachment to a central person. Their management framework explains why a compelling founder or purpose statement needs accompanying peer connections. In the example group, introduce two reporting practitioners who can review each other’s work instead of requiring every question to pass through the founder.
If your company stopped posting for a week, who would still have a reason to speak to whom?

How do we turn a product purpose into something members can share?
A shared community purpose identifies a problem members recognise, an improvement they can pursue together and a benefit available through participation. Write the purpose from their working lives. Then check whether prospective members already spend time on the problem and want help from peers.
Interview prospective users, implementers and relevant noncustomers separately. Ask about the last difficult reporting handoff, what they tried, whose help mattered and what they could safely show another practitioner. Ask where they already seek advice. An existing professional association may meet the need better than another company-owned group.
Go beyond asking whether people like the mission. Ask what they would bring to a first session and what they hope another participant could contribute. Treat a specific commitment as evidence worth exploring; polite agreement remains agreement. If participants want incompatible things, narrow the activity or create distinct sessions rather than broadening the slogan.
Use this example purpose template: “We bring together [people facing a recurring situation] to [shared improvement] through [work they can do together]. Members gain [practical benefit]. The company contributes [resources] and seeks [declared business benefit]. Members can influence [decisions].”
Example completed statement: “We bring reporting practitioners together to make reports easier to understand and verify through peer reviews of sanitised examples. Members gain a reusable review method and access to other practitioners. The company hosts the sessions and maintains the agreed resources, while learning about implementation barriers. Members choose clinic topics and propose changes to the review method.”
This purpose is narrower than “transform reporting” and broader than “learn our dashboard”. It leaves room for different software, competing interpretations and useful disagreement. Connect it to the choices behind turning a founder’s vision into a recognisable brand, while keeping company positioning separate from the community’s participation agreement.
Write a short charter beside the invitation: who belongs, what happens, who funds it, what participation requires and what members can take away. Explain which resources remain accessible after someone stops buying. For shared code, define licensing and maintenance separately through building an open-source ecosystem around a useful foundation.
What would members still want to do together if your product disappeared for a quarter?
How do we find founding members and make the first activity worthwhile?
Recruit founding members around a specific shared activity and ask for a bounded contribution. Start small enough to host carefully. The first useful exchange should give someone help, introduce a relevant peer or improve a shared resource; an introduction thread alone need not accomplish that.
For the example reporting clinic, invite a mix of practitioners who prepare reports, review them and implement reporting processes. Ask each person privately about their needs, access requirements and permission to share material. Avoid filling the room entirely with the most enthusiastic buyers or senior voices. Include different experience levels without expecting newcomers to expose uncertainty in front of their own management.
An example invitation could say: “We are hosting a small reporting-review session. Bring a sanitised report or one question about making a handoff clearer. We will compare review methods and leave with an editable checklist. No product demonstration is planned. You can attend without presenting, and we will agree what may be shared outside the room.”
Prepare a first session around one report. Its author explains the intended decision and the confusing handoff. Peers ask clarifying questions, propose revisions and record remaining disagreement. Close by asking which change the author wants to try and who would like to compare notes later. Return to that question next time, with permission. Repeated work can then accumulate into a practice and a history together.
Offer several first steps: read an approved example, ask privately, comment on one checklist item or join a paired review. Make a public biography, testimonial or social post optional. Recognise the actual contribution with the person’s approval. A careful question or correction deserves acknowledgement alongside a polished presentation.
GitLab’s July 2025 account of contributor onboarding describes research with first-time contributors, confusing documentation and unclear help routes, followed by personalised onboarding connected to maintainers. Its later study involved different participants and reported better completion. This is the company’s own before-and-after account, not a controlled estimate. The useful design lesson is to inspect where a newcomer gets stuck and connect assistance to an actual first task.
For an early group, give the host responsibility for introductions, unanswered questions, resource upkeep and follow-through. Name a backup. Teach another willing person to facilitate before increasing invitations. A group that needs its founder to remember every promise has not yet established continuity.
What can a newcomer receive and contribute without first proving that they belong?
Should we use Slack, a forum or an existing community?
Choose a community platform after defining the work, access boundaries and knowledge that should survive each exchange. Begin in an existing space when it already serves the participants and welcomes your contribution. A new platform earns its place when it makes an established activity easier to sustain.
Use this example platform assessment rather than selecting by popularity:
| Shared work | Option to assess | Trade-off to examine |
|---|---|---|
| Occasional peer clinic | Scheduled meetings plus a maintained resource page | Deliberate follow-up and connections between sessions |
| Ongoing conversation in a familiar workspace | Slack | History access, retention, notifications and member access |
| Persistent discussions and reusable answers | Discourse | Configuration, moderation and continuing administration |
| Discussions attached to shared software | GitHub Discussions | Suitability for participants outside the development workflow |
| A need already served elsewhere | Existing association, forum or user group | Its rules, independence and permission for company involvement |
Slack’s documentation says its free version limits access to the latest 90 days of message and file history and deletes workspace data older than one year. Check the applicable configuration before relying on a conversation space as a permanent archive. Discourse documents persistent topics, email replies, private spaces and moderation features. GitHub Discussions documents project conversations, polls, categories and marked answers. These capabilities support different work; none supplies member relationships automatically.
Decide separately what is public, restricted or never collected. The reporting example could keep identifiable working material out of the group, discuss sanitised examples privately and publish an approved checklist openly. Do not move a private contribution into a public knowledge base merely because it is useful.
Before adoption, have prospective members try joining, finding an answer and changing notifications with their normal devices and accessibility tools. Check search, export, moderation access and account removal. Keep an authorised copy of shared resources with named owners. Document who can restore the space or arrange a move if access, terms or service availability change.
How much control should members have over the community?
Give community members influence over decisions that shape their participation, with clear limits on company authority. State who sets agendas, changes rules, controls shared resources and resolves disputes. Asking for opinions without explaining what can change creates consultation without a clear decision process.
Barbara Prainsack’s analysis of digital commons warns against simply transferring governance principles for physical resources to digital information. She argues that why and how digital commons exclude people, and the consequences of that exclusion, determine whether they challenge or perpetuate power asymmetries. For a company-hosted group, use that as a prompt to examine power and exclusion, rather than claiming that software ownership makes governance shared.
Publish an example decision-rights table before inviting broad participation:
| Decision | Proposed authority | Required follow-through |
|---|---|---|
| Clinic topics and resource improvements | Members propose and prioritise; host coordinates | Publish the next agenda and selection reasons |
| Conduct rules | Members review; accountable operator adopts changes | Explain accepted and rejected proposals |
| Product roadmap | Product owner decides after research | Return a reasoned response to member input |
| Moderation | Trained moderators apply published rules | Offer private reporting and an independent reviewer |
| Member data and publication | Responsible company owner assesses each use | Explain access, processing, retention and correction routes |
| Continuation or closure | Company sponsor decides with member input | Give notice and handle shared work responsibly |
For roadmap discussions, distinguish a proposal, a research commitment and a delivery commitment. Record what input changed, what remains unresolved and who decides. Use a customer-facing roadmap with honest participation boundaries when members want to influence future functionality.
Write conduct rules around observable behaviour: harassment, discriminatory treatment, unsolicited solicitation, disclosure of confidential material and repeated disruption. Permit criticism of the product and company. Have moderators explain decisions privately, document relevant evidence with restricted access and recuse themselves when personally involved. Match the response to the conduct and severity, with immediate protective action where needed. Publish an appeal route and support moderators through difficult incidents. Choose moderators from members who already help others, train them on the published rules using worked cases, rotate the role and review decisions at agreed intervals.
dbt’s community rules provide an inspectable B2B example: direct messages require express permission, vendors must identify themselves, and promotion belongs in specified spaces. The rules describe a professional-practice space rather than a prospecting list. This illustrates an explicit boundary; it is not evidence that copying the rules produces commercial success.
Which decision could a member genuinely change, including when the company would prefer another answer?
How do we encourage participation without creating another noisy feed?
Encourage community participation by making useful contribution easier, safer and worth the effort. Define what helps the shared purpose before rewarding activity. Reading, asking, answering, teaching, organising and maintaining resources are different ways to participate; message volume cannot represent all of them.
Do not impose the 90-9-1 participation split as a natural law or target. Antelmi, Malandrino and Scarano’s 2019 Twitter analysis identified four activity levels and reported roughly three quarters of users as lurkers, rather than nine tenths. That finding is specific to its platform and activity definitions. It supports questioning a universal ratio, not replacing it with a new B2B benchmark.
For the example reporting group, distinguish a reader who applies the checklist from someone who cannot find a relevant answer, someone who fears exposing a mistake and someone whose role has changed. Invite optional feedback on usefulness and barriers. Do not diagnose all silence as disengagement, or infer value from an account remaining registered.
Replace daily prompts with an unanswered-question review, a well-prepared clinic and a maintained resource. If recognition is helpful, acknowledge an answer that someone used, a careful correction or reliable organising. Base recognition on useful contributions rather than posting volume. The risk described in Goodhart’s law applied to marketing metrics becomes concrete when people optimise for the score while the shared work deteriorates.
Use this example activity review: “What useful action was this prompt meant to enable? Who benefited? What did it interrupt? Would fewer, better exchanges accomplish the purpose?” Retire a recurring activity when members cannot explain its value. Do not manufacture debate, tag absent members repeatedly or make withdrawal socially costly.
Treat member stories and shared content as contributions requiring an agreed use. Offer accurate credit, editorial review and an option to decline publication. Discuss compensation when requesting substantial recurring teaching or production. Do not relabel necessary company support work as a volunteer opportunity. Members should be able to contribute within their capacity without becoming a distribution workforce.
Would you still want this contribution if it increased understanding but produced no extra activity on your dashboard?
Which community metrics show that members are helping one another?
Measure community value through useful peer exchanges, relationships, member influence and progress on the shared task. Define the event, population and observation period for each measure. Keep company-to-member service separate from help between members so a busy support team cannot accidentally appear to be a thriving peer network.
Use this example measurement record, with a proposed monthly review and joining cohorts observed for the same period:
| Measure | Definition and denominator | What to investigate |
|---|---|---|
| First useful experience | New joiners reporting or demonstrating a defined benefit within 30 days, divided by all joiners in that cohort | Where people fail to receive value; report survey nonresponse |
| Peer-resolved questions | Questions the asker confirms a nonstaff member helped resolve, divided by eligible peer questions | Unanswered questions, resolution time and correctness |
| Repeat peer collaboration | Distinct member pairs completing useful shared work on separate occasions | Dependence on one connector or a closed inner circle |
| Continuing member value | Benefits reported at a later review by the original cohort | Quiet members, departures and missing responses |
| Exercised influence | Proposals receiving a decision and explanation, divided by proposals requiring one | Whether members can change more than superficial details |
| Shared resource usefulness | Confirmed applications and corrections to maintained resources | Obsolete guidance and tasks the resource cannot support |
| Operating burden and harm | Staff and volunteer hours, overdue work, complaints and unresolved incidents | Whether member benefit remains sustainable |
Keep raw counts alongside ratios. Record the denominator when nobody asks a question; a zero denominator is not a perfect resolution rate. Define what counts as staff, a member and a useful exchange before comparing periods. Separate invitations, accepted membership, attendance and contribution.
An example observation is a practitioner asking about a confusing handoff, another offering a review method, and the asker later confirming which change helped. Record the type of help and follow-through with minimal necessary detail. An emoji reaction alone does not establish resolution. Optional member feedback can reveal valuable offline contact without requiring surveillance of private conversations.
Review who benefits and who remains unheard. Compare newcomers and established members, relevant roles and participation formats where the data supports it. Protect confidentiality in small groups. Ask departing members what changed, without demanding an explanation. A tight group can be valuable, but check whether its convenience depends on newcomers repeatedly doing the work of finding their way in.
Does your measurement show what members do for one another, or mostly what they do for you?
How do we connect community value to retention or revenue honestly?
Connect a community to business outcomes through an explicit mechanism and a credible comparison. Member retention, product adoption and customer renewal are different outcomes. A commercial result occurring after community participation establishes a sequence; it does not establish that participation caused the result.
For the example reporting company, propose this mechanism: a peer review helps practitioners resolve an implementation obstacle; the team applies the method; product use becomes more useful; the account has a stronger reason to continue. Inspect each transition. If the method works equally well without the product, the community may still create member value while providing weaker support for the proposed renewal mechanism.
Compare participants with similar eligible nonparticipants cautiously. Existing commitment, account health, available time and implementation maturity may affect both joining and renewing. Matching on recorded characteristics cannot remove every difference. Avoid counting several practitioners from one company as independent renewal outcomes.
Where feasible, ask a competent analyst to design randomly assigned invitations across eligible accounts, preserving existing service obligations and member choice. Choose the outcome, observation period and useful difference beforehand. This estimates the effect of the invitation policy under its conditions; it does not automatically identify the effect of participation. Account for crossover, incomplete uptake and uncertainty. With limited volume, report structured learning rather than an unsupported uplift claim.
For routine reporting, label assisted opportunities as assisted: state the qualifying interaction, attribution window and account-level counting rule. Keep self-reported influence distinct from observed activity. Do not turn a public question into an undisclosed sales sequence. Offer a commercial conversation when relevant and respect the person’s response.
Plan the operating commitment alongside the business hypothesis. Include facilitation, moderation, onboarding, resource maintenance, platform administration, incident handling and backup coverage. Distinguish company work from members’ voluntary effort. Assign a sponsor who can protect that capacity through product launches, staff changes and the next funding round.
A community should not carry an urgent pipeline target that requires every participant to become a prospect. If outbound coverage is the actual gap, address when founders should hand off outbound work as a separate operating decision. Keep the community’s contribution and the sales team’s responsibilities visible in their own terms.
What can we automate while keeping the community personal?
Automate community administration when it preserves member choice and leaves people accountable for interpretation and decisions. Start with reminders people requested, access administration, approved-resource search and alerts for unanswered questions. Use the time released for listening, introductions and follow-through.
For the example clinic, a form can record a participation preference, send the agreed meeting details and place an unresolved question in the host’s queue. Keep community notifications separate from product marketing subscriptions. Check permission and audience before sending or republishing material. Route failures to a person rather than silently assuming an invitation or answer arrived.
AI may prepare a session summary from material approved for that use. Require the host to check attribution, source links, unresolved disagreement and action owners before publication. Give contributors a way to correct the record. Preserve private-space boundaries in search and generated answers. Do not expose restricted material through a public summary or an overly broad integration.
Keep humans responsible for sensitive disclosures, accusations, sanctions, appeals, recognition and promises about product direction. Use explicit access and publication controls around the model. Record who can pause the automation and how a mistaken disclosure will be handled. Small groups may be better served by a host maintaining one clear page than by several connected systems.
Build activity through real members and their own contributions, use authentic testimonials shared with permission and clearly label automated assistance. The point of community administration is to help real people find one another and retain useful knowledge. Artificial activity gives the host less reliable evidence about whether that is happening.
When should we avoid building a community or close the pilot?
Avoid building a new community when there is no recurring shared need, no capacity to host it or no willingness to let members shape the space. A purpose statement should describe work the company can support in practice. When company behaviour contradicts the claimed purpose, revise the behaviour and invitation before recruiting members around it.
Use this example decision table before committing:
| Situation | Better next step |
|---|---|
| People need an answer from your company | Improve documentation or accountable support |
| People want useful updates but little peer contact | Offer a relevant newsletter with clear subscription choices |
| An independent community already serves the need | Contribute under its rules or support its organisers |
| Members value peer review but cannot support ongoing chat | Run a focused recurring clinic |
| The business needs immediate selling capacity | Improve the sales process separately |
| The purpose requires changes leadership will not consider | Reconsider the promise and whether to host |
For an example six-week pilot, agree the charter and coverage first, recruit willing participants, run an initial clinic and maintain the resulting resource. Hold a second exchange around applying or revising the method. At the review, examine peer help, member feedback, access barriers and total effort. Six weeks is a proposed review point, not a universal deadline or proof of long-term value.
Continue when the shared work is useful and the next cycle has accountable capacity. Redesign when demand exists but the format, boundary or onboarding fails. Pause when confidentiality cannot be protected, incidents lack responsible handling or necessary work exceeds available capacity. Stop when members consistently prefer another route or the company cannot honour the participation agreement.
If closing, explain the decision, provide appropriate notice, resolve outstanding commitments and agree what happens to shared resources and personal data. Offer a suitable alternative where one exists. Do not silently turn the member list into a marketing audience. A small, well-served group can remain small; growth is a choice to justify against its purpose.
How can Poldermarketing help make a product community useful?
Poldermarketing can help a post-seed startup building its go-to-market turn a shared customer problem into a clear invitation, a useful recurring activity and an operating process. A concrete starting scope is one audience, one clinic, an approved resource and a review of member value and workload.
Piet Baudoin works remotely in Dutch and English. Alongside positioning and findability in Google and AI answers, he builds and runs AI automations with care: lead follow-up, enrichment and qualification, CRM setup, reporting, content distribution and tool integrations. For a community, that can mean reliable logistics and approved knowledge moving between systems, with moderation and product decisions assigned to responsible people. Explore how Piet scopes, builds and runs the work, or request a no-cost growth scan to identify whether a community addresses the problem your startup actually has.
How this connects to my larger vision
I see a product as a possible meeting place. Someone arrives because a task is difficult and discovers that another person has been struggling with the same thing. They exchange a method, notice something neither had seen alone and leave with more freedom to act. I want to build things that make those moments easier to find. The product has done useful work when people become more capable through it, including through the people it brings within reach.
That also asks something of me as a maker. I can give a place its first shape without deciding every conversation that belongs there. Other people bring knowledge and possibilities I could not have put into a plan. I want technology to hold the practical details carefully enough that there is room for that encounter: the right introduction, an answer someone can find again, a contribution remembered accurately. Knowledge becomes more valuable as people can understand it, change it and carry it into their own lives.
The kind of marketing I want to make helps that shared capability grow. Sometimes the most meaningful step is for the brand to move aside because people have found one another. I want commercial success to support more of those connections, with each useful exchange giving something back to the lives on both sides of it.
Questions people ask
How many members do we need to launch a product community?
There is no universal launch threshold. Recruit enough willing participants to carry out one useful shared activity, with complementary experience and someone available to host it. A small peer group can be worthwhile. Expand when newcomers can receive help, contribute and understand the rules without depending on a founder for every interaction.
Should our product community include people who are not customers?
Include noncustomers when they share the practice or problem and can participate on clear terms. Keep product support entitlements separate from community membership. Explain commercial interests and permitted promotion. A practitioner group can welcome people using other tools; a customer implementation group may reasonably have a narrower boundary. Choose deliberately and explain why.
Is Slack or a forum better for a B2B community?
Choose the format around the shared work and members' access needs. Check Slack's current history and retention limits before using it as a lasting knowledge archive. A forum such as Discourse is worth evaluating for persistent discussions. For an initial peer clinic, scheduled meetings and a maintained resource page may be sufficient.
How do we show that a community improves retention?
Define the mechanism and observation period before comparing outcomes. Members may already be more committed than nonmembers, so a renewal difference alone does not establish an effect. Where feasible, compare randomly assigned invitations among eligible accounts. Record participation, uncertainty, support effort and member outcomes alongside commercial results. Small cohorts may support learning without supporting a causal conclusion.
Can AI manage our community for us?
Use AI for bounded assistance such as drafting summaries from approved material or routing unanswered questions for review. Keep people responsible for moderation, sensitive disclosures, recognition and consequential decisions. Explain what is processed, preserve access boundaries and provide correction routes. Do not fabricate members, conversations or agreement to make a new space appear active.