PRD v1 (internal)

The Motherhood: Employee Generated Content (EGC) Platform

Version: 0.1 (Draft for internal review) Date: 2 June 2026 Prepared by: SingleBit Prepared for: The Motherhood Primary client contact: Deanna Tomaselli (EGC Lead)

Source of truth. This PRD is built primarily from Deanna's questionnaire responses (submitted 29 May 2026), which are the latest and most authoritative client input. Where those responses conflict with the earlier proposal or the project handoff, the questionnaire wins. Conflicts and the items still open are called out explicitly. Anything not yet decided is marked [YET TO BE DISCUSSED] and gathered in Section 13. Cost figures are left blank on purpose, since pricing has not been set; SingleBit owes those estimates before the PRD locks.


1. Purpose and Background

The Motherhood runs a structured Employee Generated Content training program that turns a client organization's own staff, volunteers, and members into authentic voices on LinkedIn. The program is delivered offline in three phases: Discovery, live Training, and Follow-up and Ongoing support.

This platform digitizes the last part of that engagement: the Follow-up and Ongoing phase, plus the SMM Strategy Session. Its single job is to keep employees posting and engaging after the trainers leave the room. It does this by giving each employee personalized monthly story prompts tied to their organization's brand and their own voice, plus light reminders to stay active. Employees write their own posts and publish them on their own LinkedIn profiles.

The platform never posts to an employee's personal account. The only place the platform publishes directly is the organization's official LinkedIn Page, and only when the SMM chooses to.


2. Source of Truth and Decision Hierarchy

When documents disagree, the order of authority is: questionnaire responses (29 May 2026), then the project handoff, then the proposal. The questionnaire revised three things the earlier documents treated as settled. Those revisions are now the requirement:

TopicEarlier position (proposal / handoff)Confirmed position (questionnaire)Governs
What AI produces for an employeeReady-to-use post drafts the employee editsIdeas, prompts, and fill-in-the-blank prompts only; the employee writes the postQ1, reinforced by Q18 and Q23
Approval of employee postsManager approval flows are inNo approval gate; the SMM does not even see employee draftsQ2, Q3, Q18
Planning modelCalendar-centric, with weekly calendar trackersFlat monthly list of prompts; calendar is optional and may not be builtQ5, Q6, Q19, Q20

These three points shape the rest of this document.


3. Goals and Non-Goals

3.1 Goals

3.2 Non-Goals (v1)


4. Users and Roles

Three roles ship at launch. Deanna confirmed these are enough to start, while noting future clients may need more (Q25). The role model is built to extend, and the Enterprise tier supports custom roles.

Volunteer is handled as an Employee-type account at launch (same core experience, organization may scope what they see). It is listed as its own role for future flexibility.

4.1 Permissions (high level)

CapabilitySuper AdminSMMEmployee
Create orgs, onboard SMMsYesNoNo
Template library and cloningYesNoNo
Cross-org analyticsYesNoNo
Client-facing monthly reportsYesView own orgNo
Manage employeesWithin any orgOwn orgNo
Brand setup: first draftGenerated by platformConfirms after MotherhoodNo
Brand setup: edit and approveYes (first)Confirm stepNo
Prompt library: addYesYes, no approval (Q21)No
Prompt library: edit Motherhood promptsYes[YET TO BE DISCUSSED] (Q22)No
Monthly plan and themesOversightOwns (Q4)No
View employee moments, drafts, voice profileNoNo (Q3, Q10)Own only
Company LinkedIn Page postingNoYesNo
Personal post writingNoNoYes
Log momentsNoNoYes
Voice profileNoNoYes (private)
Metrics entry and growth journalNoView org summaryOwn

5. Core Principles

These are non-negotiable and apply across every feature.

  1. AI accelerates configuration and prompting. Humans own decisions, voice, and posting.
  2. AI generates ideas and prompts only. It never writes a complete post for anyone (Q1, Q18, Q23).
  3. No auto-posting to personal accounts. Employees always copy and paste to their own LinkedIn. The single exception is the SMM publishing to the company Page through the workspace.
  4. No approval gate in the employee flow. Employees post on their own. The SMM does not see drafts or moments (Q2, Q3).
  5. Employee privacy and autonomy. Moments, drafts, and the voice profile are private to the employee. The SMM cannot view a voice profile or generate posts on an employee's behalf (Q3, Q10).
  6. Authenticity bar everywhere. Every AI-surfaced prompt reflects the client's brand voice and avoids the overused-AI-phrase blacklist. The lens is "does it sound human, does it sound like the client."
  7. The hand-written prompt library is never overwritten by AI. AI may suggest new prompts to add; it never rewrites or replaces existing ones (Q23).
  8. Configured per client, not built per client. The frameworks are constant; the inputs vary. The platform's value is making that configuration easy.

6. Scope (v1)

6.1 In scope

6.2 Out of scope or later


7. Functional Requirements

7.1 Onboarding and Access

7.2 Brand and Content Configuration (per org)

The platform stores and manages each client's configured frameworks. These are the durable artifacts that drive every prompt:

Creation and approval workflow:

Ownership of content buckets (Q9):

7.3 Prompt Library

7.4 Employee Voice Profile

7.5 Moment Logging

7.6 Monthly Prompts (core employee flow)

7.7 Personal Idea Bank

7.8 Weekly Engagement Tracker (optional)

7.9 Calendar

7.10 SMM Posting Workspace and LinkedIn Page Integration

7.11 Metrics, Growth, and Reporting

7.12 Asset Library

7.13 Transcripts and Recordings

7.14 Template Management (Super Admin)

7.15 Notifications and Email

7.16 LinkedIn Profile Analyzer (planned, confirm scope)

7.17 Authenticity Self-Check (SingleBit proposal, confirm)


8. Non-Functional Requirements


9. Data Model (high level)

Core entities and their tenancy:


10. Integrations and Dependencies


11. Build Phases and Timeline

From the approved proposal. Total estimate 8 to 12 weeks. Timelines depend on timely inputs and approvals from The Motherhood.

PhaseDescriptionDuration
01Discovery sessions, finalize this PRD, complete UI and UX design2 to 3 weeks
02Build the backend, the prompt-generation engine, and LinkedIn integration3 to 4 weeks
03Build the web platform with all three dashboards and ship the first release2 to 3 weeks
04Add the PWA layer, end-to-end testing, and launch readiness1 to 2 weeks

12. Acceptance and Support


13. Open Items: Yet to Be Discussed

Every item below needs a decision, a cost estimate, or a feasibility check from SingleBit before this PRD locks. Cost figures are intentionally blank here.

#ItemQuestion(s)TypeWhat is needed
1Retention of old ideas, drafts, and momentsQ14CostSingleBit cost estimate of keeping vs auto-archiving, then a policy decision
2Weekly engagement tracker as an optional featureQ15Cost and scopeCost estimate; confirm the optional on/off behavior
3Asset library: whether it exists and its storage policyQ24Cost and scopeDecide if needed at all; storage policy (auto-delete?); cost estimate; option to keep assets on the client's domain
4Transcripts and recordings on-platform vs Google DriveQ28CostCost estimate; default to Google Drive if there is any cost
5Calendar featureQ19, Q20DecisionConfirm the calendar is cut or optional before wireframing
6SMM editing or overriding Motherhood-written promptsQ22DecisionConfirm; SingleBit recommends duplicate-then-edit to protect brand voice
7Automatic brand-mention trackingQ26FeasibilityValidate what is possible under LinkedIn API limits, then confirm
8Authenticity self-check on employee-written textSingleBit proposalDecisionConfirm whether to build, since AI no longer drafts
9LinkedIn Profile AnalyzerHandoff (planned)DecisionConfirm in scope for v1 or a later phase
10Approval as an optional Enterprise capabilityReconciliationDecisionApproval is cut from the employee flow; confirm whether it survives anywhere (for example, Enterprise) or is fully removed

14. Appendix A: Per-Client Frameworks and Live Examples (reference)

The frameworks in 7.2 are configured per client. Two current examples show how different the output gets from the same structure, which is exactly why the configuration tooling matters.

Catalyst Connection (manufacturing-growth nonprofit, southwestern Pennsylvania, around 30 staff, SMM "Ali"):

PA Women Work (women's career nonprofit across Pennsylvania, mixed staff, volunteers, and members):

These confirm the model: the platform should make configuring a new client fast and faithful, not generate the judgment from scratch.


End of v0.1 draft. Next step: resolve the items in Section 13 with Deanna, attach cost estimates, then move to UI and UX design.

●  Have your own?

Paste your markdown, get a link like this.

Write your own
PRD v1 (internal) · DocHouse