Skip to content

Go-to-Market Strategy

Ibby Expansion Loop

Summary

Ibby launches with shareable role/candidate agents to prove distribution + build match-ready inventory, then repeatedly turns on matching per archetype once (BYR/MAR)[#byr-mar] and coverage thresholds show reliable conversation-ready shortlists.

Objective

Solve the cold-start / flywheel problem of a dual-sided marketplace by launching single-party "agents" that deliver value independently, generate real-world usage data, and accumulate durable marketplace inventory—then enable profile matching once candidate supply reaches critical mass.

Critically, Ibby treats market expansion as a repeatable Phase 1 → Phase 2 loop per role archetype (not a one-time platform-wide conversion). Each archetype becomes its own liquidity "unit" that can be proven, activated, and scaled.

Core primitives being validated

Claim-Based Fit Modeling

  • Both sides submit basic information, then answer clarifying questions driven by previous submission and learned behavior.
  • The system produces structured claims about a person/role and embeds them to enable precise matching, weighting, and filtering across semantic signals.
  • See entry in (Core Mechanics)[/core-mechanics.md#claim-based-fit-modeling] for further details.

Agent Interrogation

  • Agent Interrogation: Each agent can be "interrogated" by the public: people ask questions, the agent answers from claims + authored context.
  • When the agent can't answer confidently, it generates follow-up questions for the author.
  • Authors answer follow-ups asynchronously; the agent improves continuously.
  • See entry in (Core Mechanics)[/core-mechanics.md#agent-interrogation] for further details.

Archetypes: how Ibby defines liquidity and expansion

The Archetype is Ibby's standardized role category defined by shared evaluation dimensions and a shared candidate supply pool.

Archetypes are the unit of measurement for:

  • "critical mass" inventory thresholds
  • matching viability
  • Phase 1 → Phase 2 activation gates

Initial Target Archetype

Software engineer

  • backend/full-stack
  • mid to senior level
  • B2B SaaS
  • US/Canada time zones
  • remote-friendly

This archetype is intentionally chosen because:

  • it has large candidate supply
  • evaluation criteria are relatively standardized
  • remote-friendly roles are common and searchable
  • willingness to adopt new workflows is high (post-seed to growth-stage startups)

Phase 1: Single-party launch

Phase 1 is designed to:

  1. generate durable inventory (agents + claims)
  2. validate distribution via existing behaviors (job posts, applications, outreach)
  3. validate interrogation + follow-up loops as repeatable usage

Service distribution channels

  • Landing pages + SEO
  • LinkedIn networking + targeted ads (including niche communities)
  • Reddit/community ads
  • Job boards and other placements that can host/point to agent URLs

Agent URLs

  • Unique URL per agent

    • Company agent may cover one role
    • Candidate agent covers one individual

  • Agent links are designed to be forwarded in Slack/email and embedded in job postings/applications.

Shareable agent URLs (with surface attribution)

  • Role agents embedded in job posts and shared in applications
  • Optional link tags ("Copy link for job post/application/outreach/referral") used to attribute which surfaces generate reach and engagement

Companies: Company/Role Agent

  • Company creates an agent, optionally with multiple role "threads."
  • Candidates can ask anything about the company/role/process.
  • Defaults: open answers for non-blacklisted topics; optional user-controlled gating.
  • PII is excluded by default; salary sharing is configurable.

Candidates: Professional Profile Agent

  • A candidate creates a shareable profile agent that can discuss background, skills, history, and role-relevant questions (including common interview-style prompts).
  • Companies can interrogate the candidate agent before scheduling interviews, improving signal and reducing wasted calls.
  • Not "spammable" — agent is not an outbound blast mechanism; it's a high-signal artifact shared in-context (e.g., with an application or recruiter outreach) to enable deeper evaluation asynchronously.

Virality loop

  • If a question triggers a clarifying follow-up and the company later answers it, the system can optionally collect the questioner's contact info and notify them when the answer is available.
  • Agent landing page contains Ibby branding and a direct link to "create your own agent."

Feedback loops

why Phase 1 matters

Phase 1 produces real-world signals that improve both primitives:

  • Claim quality improves because clarifying questions are driven by what real people ask.
  • Coverage improves because authors fill gaps over time, asynchronously.
  • Matching readiness improves because structured claims become more complete and comparable before matchmaking is turned on.

P1 Metrics

Agents -> inventory + distribution

Distribution + interrogation funnel

Primary metrics that prove agents are getting shared through real workflows and generating real interrogation.

Metric What it measures Window
External reach Unique non-owner visitors per agent 7-day
External engagement % of non-owner visitors who submit >= 1 Q&A prompt 7-day
Engagement depth Non-owner Q&A prompts per agent per week

Loop health guardrail

This ensures the interrogation loop actually closes and the agent improves over time (vs stagnating).

  • Follow-up SLA
  • % of follow-up questions resolved within 72 hours (via response or claim update)
  • tracked continuously
  • Goal: 80%

Phase 1 success criteria

These thresholds define "the Phase 1 loop is working" for Archetype 1.

Criterion Description Threshold
External reach Average number of unique non-owner visitors per agent (7-day) 3
External engagement % of agents receiving >= 1 question (7-day) 25%
Follow-up SLA % resolved within 72 hours 80%

Critical mass

Phase 1 -> Phase 2 bridge

We enable Phase 2 matching for an archetype only when Ibby can reliably produce Qualified Match Briefs for that archetype while meeting SLA stipulations.

Critical mass metric

  • Match-ready inventory
  • # of match-ready profiles, per archetype
  • rolling 30 days
  • Goal: 2,000-3,000 (viable minimum ~1,000)

Match-ready profiles have sufficient coverage to evaluate fit across the archetype's core dimensions (not just a resume upload).

Activation criteria

  • Shortlist reliability: Ibby can generate 4 Qualified Match Briefs (QMBs) within 7 business days of a company agent for that archetype reaching match-ready status (per archetype).
  • Supply quality, not just volume: "Match-ready" profiles meet a minimum structured-claim completeness threshold of >= 80% against the archetype's required evaluation dimensions (e.g., coverage in >= 8 of 10 required dimensions).
  • Repeatability requirement: All activation criteria must hold week over week for 4 consecutive weeks (per archetype), measured on a minimum weekly sample size of:
  • >= 10 new or updated company agents declaring intent for that archetype (or equivalent "standing role intent" objects), and
  • >= 50 QMBs delivered generated by matching those company agents against candidate agents, and
  • across >= 5 distinct companies.

We Enable matching for an archetype once a role can consistently receive 4 qualified Match Briefs within 7 business days, supported by stable BYR and MAR, with sufficient claim completeness to sustain credible mutual affirm.

BYR + MAR

core conversion rates that determine liquidity

Phase 2 viability depends on two rates that connect inventory size to reliable outcomes.

Together, BYR and MAR allow Ibby to forecast whether current inventory can produce 3–5 conversation-ready briefs per role without manual sourcing.

BYR (Brief Yield Rate)

The % of match-ready profiles in an archetype that can become a Qualified Match Brief for a specific role once constraints and weighting are applied.

  • Early target assumption (Archetype 1): BYR ≈ 5–10%

MAR (Mutual Affirm Rate)

The % of qualified Match Briefs that reach mutual "affirm interest" within a defined time window.

  • Early target assumption: MAR ≈ 20–30%

Phase 2: Two-party matching

Phase 2 turns Phase 1 inventory into a structured marketplace: Ibby begins producing and operating real matches between company agents and candidate agents.

Phase 2 is designed to:

  1. Activate matching only where Phase 1 has proven sufficient critical mass (per archetype)
  2. Convert match-ready inventory into Qualified Match Briefs (QMBs) with measurable engagement (interrogation, affirm interest)
  3. Operationalize the Handshake as a reliability layer that drives fast, human-to-human first conversations
  4. Validate repeatable outcomes loops (which matches progress, where they fail, and why) to improve modeling and guardrails

What gets added in Phase 2

  • Matching engine + scoring: Ibby compares claim-based profiles across the archetype's core dimensions and produces ranked candidates.
  • Qualified Match Briefs (QMBs): the unit of delivery that summarizes fit, gaps, and next-best questions to ask.
  • Context interrogation loop: either side can ask targeted follow-ups to resolve uncertainty before committing.
  • Handshake workflow: mutual affirm interest triggers scheduling + interview windows with guardrails to prevent ghosting and abuse.

How Phase 2 is distributed

Phase 2 expands distribution by attaching value to the same surfaces Phase 1 already validated:

  • Company-side: job posts, outbound outreach, referrals, and internal hiring workflows that point to company agent URLs
  • Candidate-side: applications, inbound interest, recruiter intros, and referrals that point to candidate agent URLs

Match surfaces

  • Match-ready company agents are eligible to receive QMBs for the archetype.
  • Match-ready candidate agents are eligible to appear in QMBs for that archetype.
  • Both sides still share the same agent URLs, but Phase 2 adds a new outcome: a brief becomes a handshake when both parties affirm interest.

P2 Metrics

Matching → outcomes + liquidity

Success Criteria

Criterion Description Threshold
Shortlist reliability # of qualified Match Briefs delivered within 7 days of role intake 3-5
Agent engagement % of delivered QMBs to receive 1 or more context interrogation questions within 3 business days of delivery 60%
Handshake conversion % of Match Briefs that reach mutual "affirm interest" within 7 days 20%
Intro completion % of mutually affirmed matches that complete a first conversation within 10 days 75%

Trust guardrail

  • Employer follow-through rate (no-ghost)
  • % of mutual affirms where the employer completes the promised first conversation
  • Goal: 85%

Liquidity / role coverage

  • Coverage ratio
  • Median match-ready candidates available per role intake, per archetype), plus p25
  • Track p25 to avoid "average-only" success
  • Goal: 60-100 (~30 minimum; > 150 comfortable)

Iterative Phase Transitions

repeatable loop by archetype

Ibby does not rely on a single platform-wide "flip" from Phase 1 to Phase 2. Instead, Ibby repeats the same loop as it expands:

1) Phase 1 (for Archetype X): build match-ready inventory + prove interrogation and follow-up behavior 2) Activation gate (for Archetype X): confirm critical mass + coverage ratio + stable loop health 3) Phase 2 (for Archetype X): enable matching and measure shortlist reliability, BYR, MAR, and intro completion 4) Scale: expand acquisition and repeat for the next archetype

This produces a controlled expansion model where Ibby can grow liquidity systematically without sacrificing match quality.

Instrumentation notes

how metrics are measured

Ibby's Phase 1/2 metrics are designed to be auditable from first-party logs without requiring visibility into external ATS systems or private email threads. Where "sharing" cannot be observed directly (copy/paste into an application form), Ibby measures downstream outcomes (external landings and engagement) and attributes them to distribution surfaces via optional link tagging.

Identity: owner vs non-owner

Each agent has a canonical owner (creator) and all sessions/events are classified as:

  • Owner: authenticated creator session (or access token associated with the agent owner)
  • Non-owner: any session not associated with the owner (including anonymous visitors)

All "external reach / engagement / depth" metrics are computed using non-owner sessions/events only.

Uniqueness and session counting

To avoid "sessions ≠ people" problems, we track two related measures:

  • Unique visitor: de-duplicated by a first-party cookie + user agent fingerprint (and optionally hashed IP as a secondary signal)
  • Session: a time-bounded interaction window (e.g., 30 minutes of inactivity ends a session)

Primary reporting uses unique visitors where possible; sessions are used for secondary diagnostics.

Ibby provides optional "Copy link for …" buttons that append a lightweight query param or short-code:

  • ?src=job_post
  • ?src=application
  • ?src=outreach
  • ?src=referral

Notes:

  • These tags do not need to identify the recipient or platform.
  • If a link is shared without tags (manual copy from address bar), it still counts toward reach/engagement, but is recorded as src=unknown.

Primary channel mix metric:

  • Channel mix: % of non-owner landings by src over a rolling 7/30 day window

Event taxonomy (what we log)

At minimum, Ibby logs the following events with agent_id, timestamp, owner/non-owner, and src tag:

Traffic + reach

  • agent_view (agent page loaded)
  • agent_session_start / agent_session_end

Engagement

  • prompt_submitted (non-owner submits a Q&A prompt)
  • response_generated (agent produces an answer)
  • followup_generated (agent generates a clarifying question for the owner)

Loop closure

  • followup_resolved (owner answers follow-up OR updates structured claims to address it)
  • claims_updated (structured claim set materially changed; see definition below)

Matching (Phase 2)

  • role_intake_complete
  • brief_created (Match Brief generated)
  • brief_qualified (passes minimum quality threshold)
  • affirm_interest_employer
  • affirm_interest_candidate
  • mutual_affirm
  • intro_initiated / intro_completed (first conversation confirmed)
  • employer_no_ghost (employer meets handshake expectation within time window)

Definitions for "meaningful" and "match-ready"

To prevent vanity metrics and spam inflation, Ibby uses operational definitions:

Meaningful engagement

  • A non-owner session counts as "engaged" if it includes ≥ 1 prompt_submitted
  • Optionally, require ≥ N seconds on page OR ≥ 2 interactions to exclude accidental bounces

Prompt quality guardrails (anti-spam)

  • Prompts from the same visitor can be rate-limited
  • Prompts below a minimum length or obviously spam can be excluded from "depth" reporting
  • Repeated identical prompts from the same visitor can be de-duped

Match-ready profile

A profile is "match-ready" for an archetype when:

  • Required claim dimensions for that archetype meet a minimum completeness threshold
  • Constraints (timezone/work mode/seniority band) are present
  • The profile is active in the last 30 days (viewed, updated, or engaged)

Ibby reports inventory by archetype and by completeness tier (e.g., 60% / 80% / 95% coverage).

BYR and MAR measurement (Phase 2)

BYR (Brief Yield Rate)

  • Computed per role intake: qualified_briefs / match_ready_candidates_in_scope
  • "In scope" is defined by hard constraints + archetype membership at time of intake

MAR (Mutual Affirm Rate)

  • Computed per role intake (or cohort): mutual_affirms / qualified_briefs
  • Time-windowed (e.g., within 7 days) to reflect real marketplace cadence

Privacy and compliance notes

  • Ibby can operate with anonymous non-owner sessions (no login required to interrogate).
  • PII is excluded by default from public agent outputs; sensitive fields are controlled by policy and author settings.
  • Link tags (src) do not contain recipient identity—only coarse distribution surface classification.

Reporting cadence and stability requirement

Because early usage can be noisy, "working" thresholds are evaluated as:

  • rolling 7-day metrics for reach/engagement
  • rolling 30-day metrics for inventory
  • "stable week-over-week" defined as meeting success criteria for 4 consecutive weeks before activating Phase 2 for a new archetype