AE Leadgen - Trigger Logic Framework

Trigger Logic Framework

Phase 3: Problem Type Detection

Status: Spec complete — ready for implementation
Owner: Jacob Stallings
Feeds into: LEAD_CLASSIFICATION_PLAN.md Phase 3
Implementation plan: PHASE3_IMPLEMENTATION_PLAN.md
Thresholds: leadgen/thresholds.py — all numeric constants live here; edit this file to tune any value
Last Updated: 2026-04-25

Strategic Framing

A trigger is not just an event. It is a mismatch between activity and momentum.

The goal of Phase 3 is not to detect that an artist released music. It is to detect the gap between what an artist is doing and what they need — and to respond with a message that is specific to that exact gap.

This is what makes Artist Evolve’s outreach feel like diagnosis instead of cold pitching.

Trigger Groups

All triggers fall into three top-level groups, matching the HubSpot trigger category field.

Group	What it detects
Release	Artist is releasing music but the rollout system or support content is weak
Post	Artist has content activity or social momentum but no release to capture it
Hybrid	Cross-signal mismatch — activity on one platform doesn’t match the other

A fourth group, Show, is defined in this document but is Phase 3B (requires a new data source).

Tier Definitions

Not all triggers have the same urgency or confidence level.

Tier	Name	Outreach approach
Tier 1	High-intent / high-opportunity	Fast automated outreach — gap is obvious and timely
Tier 2	Medium-opportunity	Automated outreach or soft nurture — message needs more nuance
Tier 3	Subjective / diagnostic	Manual review before outreach — requires human or AI interpretation

Phase Overview

Phase	Triggers	What’s required
Phase 3A	Objective, threshold-based	Add 3 new pipeline fields (release date, release count); everything else from existing data
Phase 3B	Subjective, AI-assisted, or new source	Viral post detection (Apify post-level engagement), show activity (Songkick/Bandsintown), brand consistency scoring (AI)

New Pipeline Fields Required for Phase 3A

Three new fields must be added to ArtistLead and the ChartMetric enrichment step. Without these, the release-based triggers cannot fire.

Field	Type	Source	Description
`days_since_last_release`	`int \| None`	ChartMetric — most recent track/album date	Days since most recent Spotify release. `None` if no release history found.
`release_count_90d`	`int \| None`	ChartMetric — filter releases to last 90 days	Number of tracks/albums released in the last 90 days.
`most_recent_release_title`	`str \| None`	ChartMetric	Title of most recent release (for personalized messaging).

These map to new HubSpot custom properties: days_since_last_release, release_count_90d, most_recent_release_title.

Phase 3A: Objective Triggers

T1-A: Release But No Content

Group: Release | Tier: 1 | Scenario ID: release_no_content

What it detects: Artist dropped a release in the last 30 days but has posted fewer than 5 times in the 8-week window. The music exists but the rollout support is essentially absent.

Condition logic:

days_since_last_release <= 30
AND recent_post_count < 5
AND days_since_last_release IS NOT NULL

Required data: days_since_last_release (new), recent_post_count (existing)

Priority over: None (Tier 1 fires first)

Message angle: Rollout gap. The music is moving but the surrounding content isn’t helping it travel.

Sample DM:

Caught your recent drop — really solid. One thing stood out when I looked at your Instagram: there’s barely any content around it. That’s usually what separates a release that builds momentum from one that just lands quietly. We work specifically on that rollout layer with artists at your stage. Would it be worth a quick conversation?

HubSpot recommended action: Schedule outreach — Tier 1 priority

T1-B: Active Poster, No Release

Group: Post | Tier: 1 | Scenario ID: active_poster_no_release

What it detects: Artist has been posting consistently and has built real Instagram presence, but has not released music in 60+ days. Content momentum exists but there is no release to direct it toward.

Condition logic:

recent_post_count >= 10
AND days_since_last_release > 60
AND instagram_followers >= 2000
AND days_since_last_release IS NOT NULL

Required data: days_since_last_release (new), recent_post_count (existing), instagram_followers (existing)

Priority over: T2-E (Momentum Lost) — if an artist is both active and has been posting recently, this fires instead of momentum lost.

Message angle: Content momentum without a release to convert it. The audience is engaged; there is no door to lead them through.

Sample DM:

Been watching your feed — you’re putting out a lot and people are clearly responding. What jumped out is there’s been no release in a while. Content like yours builds up a level of trust and attention that’s hard to get. The question is what you’re doing to convert that into something. Have you got music in the works or is it more of a timing/structure issue?

HubSpot recommended action: Schedule outreach — Tier 1 priority

T1-C: Release But Weak Rollout Content

Group: Release | Tier: 1 | Scenario ID: release_low_content

What it detects: Artist released recently and is posting, but volume around the release is low (5–14 posts in 8 weeks). There is some support but not enough for the release to compound. This is distinct from T1-A, which is near-zero content.

Condition logic:

days_since_last_release <= 30
AND recent_post_count >= 5
AND recent_post_count < 15
AND days_since_last_release IS NOT NULL

Required data: days_since_last_release (new), recent_post_count (existing)

Priority over: None — can co-fire with T2-C if release count is also high.

Message angle: Support exists but it is too thin to create cohesion or staying power around the release.

Sample DM:

Listened to your recent release — there’s clearly something there. Noticed the content around it is minimal. A few posts isn’t enough to build the kind of surrounding world that makes a release stick. The artists who get the most mileage out of their drops usually have a structured content layer that runs parallel to the music, not just a couple posts. Happy to walk you through how that looks in practice.

HubSpot recommended action: Schedule outreach — Tier 1 priority

T2-A: High Spotify, Low Instagram

Group: Hybrid | Tier: 2 | Scenario ID: high_spotify_low_instagram

What it detects: Artist has meaningful streaming traction but Instagram presence is small relative to their listener base. This disconnect typically signals that the artist is not building a brand or community around their music — they have an audience that doesn’t translate into a following.

Condition logic:

spotify_monthly_listeners >= 10000
AND instagram_followers < 3000
AND spotify_monthly_listeners IS NOT NULL
AND instagram_followers IS NOT NULL

OR (ratio-based fallback):

spotify_monthly_listeners >= 5000
AND instagram_followers IS NOT NULL
AND (spotify_monthly_listeners / instagram_followers) > 10

Required data: spotify_monthly_listeners (existing), instagram_followers (existing)

Message angle: Real music audience that has not been converted into community. Streaming is passive; Instagram is how artists build stickiness.

Sample DM:

Looked you up — the Spotify numbers are genuinely strong. What caught my attention is the Instagram is way behind it. That gap usually means there’s an audience out there who knows the music but doesn’t have a reason to follow you. Converting streaming attention into actual community is one of the specific things we work on. Worth a conversation?

HubSpot recommended action: Queue for outreach — Tier 2, softer angle

T2-B: High Instagram, Low Spotify

Group: Hybrid | Tier: 2 | Scenario ID: high_instagram_low_spotify

What it detects: Artist has a real Instagram presence but streaming numbers are significantly behind. This usually means the content is working but the music itself or the release strategy isn’t converting the audience into listeners.

Condition logic:

instagram_followers >= 5000
AND spotify_monthly_listeners < 5000
AND instagram_followers IS NOT NULL
AND spotify_monthly_listeners IS NOT NULL

Required data: instagram_followers (existing), spotify_monthly_listeners (existing)

Message angle: Social presence that hasn’t translated into music momentum. The content creates awareness but not music engagement.

Sample DM:

Your Instagram presence is legit — you’ve clearly built something there. What stood out looking at the full picture is the Spotify side hasn’t caught up yet. When social and streaming are this far apart, it usually comes down to how the music is being positioned in the content and whether the path from follower to listener is clear. That’s very fixable. Is that gap something you’ve been thinking about?

HubSpot recommended action: Queue for outreach — Tier 2

T2-C: Prolific Releases, Low Audience

Group: Release | Tier: 2 | Scenario ID: prolific_releases_low_audience

What it detects: Artist has released multiple times in the last 90 days but their audience (both Instagram and Spotify) remains small. This is often a sign that the release volume is there but the brand context, rollout structure, or artist identity is not helping any individual release gain traction.

Condition logic:

release_count_90d >= 3
AND instagram_followers < 3000
AND spotify_monthly_listeners < 5000
AND release_count_90d IS NOT NULL

Required data: release_count_90d (new), instagram_followers (existing), spotify_monthly_listeners (existing)

Message angle: High output, low signal. Releasing frequently without building context around who the artist is.

Sample DM:

You’ve been putting out a lot of music — more than most artists I look at. The thing that’s not moving in proportion is the audience. That pattern usually comes down to one thing: the releases aren’t connecting because the brand and world around the artist isn’t clear enough yet. The music is ready; the context around it isn’t built out. That’s exactly the gap we work on.

HubSpot recommended action: Queue for outreach — Tier 2

T2-D: Momentum Lost

Group: Post | Tier: 2 | Scenario ID: momentum_lost

What it detects: Artist has a meaningful history of posting (total post count is substantial) but recent posting has dropped significantly. This catches artists who were active and then went quiet — a pattern that is common before or after big life events, releases, or burnout.

Condition logic:

instagram_post_count >= 50
AND recent_post_count <= 2
AND instagram_post_count IS NOT NULL
AND recent_post_count IS NOT NULL

Note: Use classification_method == "recent_window" to confirm recent count is accurate (not a fallback). If fallback was used, do not fire this trigger.

Required data: instagram_post_count (existing), recent_post_count (existing), classification_method (existing)

Message angle: Reactivation. They built something before. Getting back after a break is harder than starting because the signal needs to be re-established, not just resumed.

Sample DM:

Noticed from your profile you were very active for a while and then went quiet. That pattern is more common than people think. The tricky part of coming back from a gap isn’t just posting again — it’s re-establishing the signal and making sure you’re not starting from zero in the algorithm. We have a specific framework for that reactivation phase. Would it be useful to talk through it?

HubSpot recommended action: Queue for outreach — Tier 2, reactivation angle

T2-E: Consistent Posting, Low Growth

Group: Post | Tier: 2 | Scenario ID: consistent_posting_low_growth

What it detects: Artist is posting consistently (Level 3 or Level 4 activity) but both Instagram followers and Spotify monthly listeners are relatively low for the posting volume. Consistent effort without proportional growth indicates a signal or positioning problem, not a consistency problem.

Condition logic:

recent_post_count >= 10
AND instagram_followers < 5000
AND spotify_monthly_listeners < 10000
AND instagram_followers IS NOT NULL
AND spotify_monthly_listeners IS NOT NULL

Required data: recent_post_count (existing), instagram_followers (existing), spotify_monthly_listeners (existing)

Priority note: Do not fire if T1-B (Active Poster No Release) already fires — they would overlap.

Message angle: Disciplined effort without proportional return. The work is happening but the system isn’t converting it.

Sample DM:

You’re clearly putting in the work — the consistency is obvious. What jumped out is the audience growth hasn’t kept pace with the output. That usually means there’s a positioning or signal issue, not a work ethic issue. Artists who post consistently but aren’t growing are almost always missing one specific thing in how the brand is being communicated. Happy to point to what it probably is.

HubSpot recommended action: Queue for outreach — Tier 2

Phase 3B: Subjective / AI-Dependent Triggers

These triggers are defined but not automatable without additional data sources or AI assistance. They are included here so the system can be extended in a later sprint.

T3-A: Viral Post, No Recent Release

Group: Post | Tier: 1 | Scenario ID: viral_post_no_release

What it detects: A specific post has significantly outperformed the artist’s baseline engagement (target: 5x or more above average), and there is no recent release to capture that attention.

Why it’s Phase 3B (currently): The computation — comparing peak vs. average post engagement — is not yet wired into the pipeline, even though the data itself is already available.

Cost to add: Near zero. The Apify instagram-scraper actor already returns likesCount and commentsCount per post. This data is already normalized in _normalize_post() in apify_client.py and is present in every post scrape response today — it is just not being used beyond timestamp counting. The only required changes are:

Extend the post window from 56 → 90 days so the baseline is more reliable. This fetches ~60% more posts per artist in the same actor run, which adds marginal compute — roughly $0.50–1.00 extra per 90-artist batch at current Apify usage.
Compute engagement baseline in scenario_detector.py: sum likes_count + comments_count per post, compute average over 90-day window, flag if any single post in the last 30 days exceeds 5x that average.

No new actor runs. No new Apify actors. No additional API calls. The data is already coming back — this is purely a pipeline computation change.

Estimated incremental cost at scale:

Volume	Additional Apify cost	Reason
90 artists/run	~$0.50–1.00 extra	60% more post data per run from window extension
1,000 leads/month	~$5–10 extra	Same ratio
5,000 leads/month	~$25–50 extra	Same ratio

Note: These are rough estimates based on the current ~$1–2 per 90-artist run for the post scrape step. The actual number depends on average post volume per artist.

Condition logic (when implemented):

max(likes_count + comments_count for post in last_30d_posts) >= (avg_post_engagement_90d * 5)
AND days_since_last_release > 60
AND len(last_90d_posts) >= 3  # require enough history to establish a baseline

Required changes: scrape_recent_posts() window parameter 56 → 90 days; scenario_detector.py engagement baseline computation; add avg_post_engagement_90d and peak_post_engagement_30d to ArtistLead model.

Proxy (Phase 3A stopgap): This scenario partially overlaps with T1-B (Active Poster No Release) when instagram_followers >= 5000. Use T1-B as a proxy until this is implemented.

Sample DM:

That post you put out [approximate timeframe] clearly hit different — the engagement on it was well above what you normally get. The timing of it struck me because there’s no release out right now. You basically got a room full of people’s attention and then had nothing to direct them toward. That’s exactly the window we help artists close. Is there anything you’re sitting on right now?

T3-B: Release, Scattered Brand Content

Group: Release | Tier: 3 | Scenario ID: release_scattered_brand

What it detects: Artist released recently and has supporting content, but the visual or thematic identity across those posts is inconsistent — different aesthetics, mixed tones, no cohesive thread.

Why it’s Phase 3B: Requires visual AI classification across an artist’s recent post grid. Instagram grid coherence is not something that can be measured from a simple scrape.

Required new data: Grid coherence score (AI, or curated human review)

Sample DM:

Listened to your release and then went through your Instagram. The music has a clear direction but the visual world around it is all over the place. That fragmentation usually means people who find the song can’t quickly form a picture of who the artist is. Cohesion between the music and the content is one of the most underrated levers in an early rollout. Worth talking through what that could look like for you.

T3-C: Consistent Shows, No Content

Group: Show | Tier: 1 | Scenario ID: consistent_shows_no_content

What it detects: Artist is performing regularly but not turning those shows into online content or audience growth.

Why it’s Phase 3B: Requires a show/gig tracking data source. Options include Songkick API, Bandsintown API, or scraping artist pages that list performances.

Required new data: show_count_90d (from Songkick/Bandsintown)

Sample DM:

You’ve been on stage a lot lately — I’ve been tracking your shows. The wild thing is it’s barely showing up online. Real-world momentum that doesn’t compound digitally is one of the most common traps for artists at your stage. You’re already doing the hard part. The system around it just isn’t documenting it. Would be worth a quick conversation.

T3-D: Consistent Shows, Content But No Growth

Group: Show | Tier: 2 | Scenario ID: consistent_shows_content_no_growth

What it detects: Artist is performing regularly and posting about those shows, but the follower and listener numbers aren’t reflecting the real-world activity.

Why it’s Phase 3B: Same dependency as T3-C — requires show tracking data.

Required new data: show_count_90d (from Songkick/Bandsintown), alongside existing instagram_followers and spotify_monthly_listeners

Sample DM:

The shows are happening and you’re documenting them online — that’s more than most artists do. What’s not moving is the actual audience numbers. That disconnect usually comes down to a positioning issue: the content is capturing proof of activity, but it’s not being used in a way that converts viewers into followers or listeners. That’s a specific and fixable problem.

Scenario Priority and Conflict Resolution

An artist can match multiple scenarios. The system should:

Store all matched scenarios in lead_scenarios (comma-separated scenario IDs)
Select a primary_scenario using this priority order:
- Tier 1 scenarios always win over Tier 2 or 3
- Within Tier 1: release_no_content > active_poster_no_release > release_low_content
- Within Tier 2: most recent signal wins (if temporal data is available), otherwise lowest audience growth
- Phase 3B scenarios are stored but never set as primary_scenario until Phase 3B is implemented
Do not fire multiple Tier 1 scenarios simultaneously in the same outreach sequence. The primary scenario determines the message; secondary scenarios inform context.

Example:

Artist: 8-week post count = 4, last release = 15 days ago, IG followers = 3,200
→ Matches: release_no_content (T1-A), release_low_content (T1-C)
→ primary_scenario = "release_no_content" (more urgent: near-zero support)
→ lead_scenarios = "release_no_content,release_low_content"

HubSpot Properties Added in Phase 3

These are additions to the existing 18 custom properties from Phase 2.

New Data Fields (from pipeline)

HubSpot Property	Type	Description
`days_since_last_release`	Number	Days since most recent Spotify release
`release_count_90d`	Number	Releases in the last 90 days
`most_recent_release_title`	String	Title of most recent release

New Classification Fields

HubSpot Property	Type	Description
`lead_scenarios`	String	Comma-separated list of matched scenario IDs
`primary_scenario`	String	Single scenario ID for primary outreach angle
`scenario_tier`	Number	Tier of primary scenario (1, 2, or 3)
`recommended_action`	String	Human-readable next step (e.g., “Schedule outreach — Tier 1 priority”)
`trigger_detected_at`	DateTime	When primary scenario was first detected
`outreach_message_type`	String	Message angle label (e.g., “Rollout gap”, “Content momentum without release”)

Trigger Detection Summary Table

ID	Name	Group	Tier	Phase	Key Condition
`release_no_content`	Release But No Content	Release	1	3A	release ≤30d AND posts/8wk < 5
`active_poster_no_release`	Active Poster, No Release	Post	1	3A	posts/8wk ≥10 AND release >60d
`release_low_content`	Release, Weak Rollout Content	Release	1	3A	release ≤30d AND 5 ≤ posts/8wk < 15
`high_spotify_low_instagram`	High Spotify, Low Instagram	Hybrid	2	3A	Spotify listeners ≥10K AND IG followers <3K
`high_instagram_low_spotify`	High Instagram, Low Spotify	Hybrid	2	3A	IG followers ≥5K AND Spotify listeners <5K
`prolific_releases_low_audience`	Prolific Releases, Low Audience	Release	2	3A	releases/90d ≥3 AND IG followers <3K
`momentum_lost`	Momentum Lost	Post	2	3A	total posts ≥50 AND recent posts/8wk ≤2
`consistent_posting_low_growth`	Consistent Posting, Low Growth	Post	2	3A	posts/8wk ≥10 AND IG followers <5K AND Spotify <10K
`viral_post_no_release`	Viral Post, No Release	Post	1	3B	max_engagement ≥5x avg AND release >60d
`release_scattered_brand`	Release, Scattered Brand	Release	3	3B	release ≤30d AND grid coherence score low
`consistent_shows_no_content`	Consistent Shows, No Content	Show	1	3B	shows/90d ≥3 AND posts/8wk <5
`consistent_shows_content_no_growth`	Consistent Shows, Content, No Growth	Show	2	3B	shows/90d ≥3 AND posts ≥10 AND audience flat

Data Pipeline Additions — Implementation Checklist

Phase 3A additions (ChartMetric)

[ ] Add days_since_last_release to pipeline
- ChartMetric endpoint: GET /api/artist/{id}/albums?sortColumn=release_date&sortOrderDesc=true&limit=1
- Compute (today - most_recent_release_date).days
- Handle case: no releases found → None
[ ] Add release_count_90d to pipeline
- ChartMetric endpoint: same album endpoint, paginate and count releases with release_date >= today - 90 days
- Or: filter the album list by date in Python after fetching
[ ] Add most_recent_release_title to pipeline
- From same ChartMetric album call — obj["name"] of the first result
[ ] Add 3 new fields to ArtistLead model (models.py)
[ ] Update chartmetric_client.py with new method get_recent_releases(artist_id)
[ ] Update pipeline.py to call new method during enrichment step
[ ] Create scenario_detector.py with rules engine
- JSON config for thresholds (easy to tune without code changes)
- detect_scenarios(artist: ArtistLead) -> ScenarioResult
- Returns scenarios: list[str], primary_scenario: str, tier: int, recommended_action: str
[ ] Add scenario fields to ArtistLead model
[ ] Update hubspot_client.py to include new properties in upsert
[ ] Wire scenario detection into pipeline (between classification and HubSpot steps)
[ ] Update HubSpot ensure_properties() to auto-create new custom properties
[ ] Add tests for scenario detection — unit tests per scenario with edge cases

Threshold Reference

All numeric thresholds are based on current Toolroom label validation results and are expected to be tuned after Phase 3 runs on a broader label set.

Signal	Threshold	Rationale
Recent release window	30 days	Captures active rollout period
No release threshold	60 days	Meaningful gap but not inactive career
High post cadence (active poster)	≥10 posts/8wk	Level 4-equivalent, clearly intentional content strategy
Low post support for release	<5 posts/8wk	Near-zero rollout support
Weak but present rollout	5–14 posts/8wk	Some support, but thin
High Spotify listeners	≥10,000	Below this, low absolute numbers less meaningful
Low Instagram (vs Spotify)	<3,000 followers	Clearly below what Spotify numbers would suggest
High Instagram	≥5,000 followers	Real presence worth noticing
Low Spotify (vs Instagram)	<5,000 monthly listeners	Clearly underperforming the social following
Prolific release count	≥3 releases/90d	More than one release per month
Small audience ceiling	<3,000 IG followers AND <5,000 Spotify	For prolific/consistent-posting scenarios
Momentum lost (active history)	≥50 total posts	Establishes prior activity
Momentum lost (recent drop)	≤2 posts/8wk	Near-silence from previously active account
Consistent posting (low growth)	≥10 posts/8wk	Intentional content schedule clearly in place

Future Enhancements

Sublabel-Based Discovery (Phase 3B+)

Problem: Major labels like Toolroom have sublabels for emerging artists (e.g., “Toolroom Trax”, “Leaders of the New School” compilations) that are more likely to feature young, hungry talent. Current discovery uses the main label name and captures all artists, including established names.

Proposed Solution:

Extend Spotify discovery to track sublabel names from album metadata
Create a configuration file mapping main labels to their development-focused sublabels
Add --sublabel CLI flag to filter discovery to specific sublabels
Add released_on_label_within_days filter to focus on artists with recent label activity (e.g., released on Toolroom Trax in the last 12-24 months)

Example Sublabels:

Toolroom → Toolroom Trax, Leaders of the New School (compilation series)
Anjunabeats → Anjunadeep, Anjunabeats Rising
Defected → DFTD, Classic Music Company

Implementation Priority: Medium (Phase 4 candidate)

Document History

Date	Version	Changes
2026-04-25	1.0	Initial framework — 8 Phase 3A scenarios, 4 Phase 3B scenarios, full threshold spec, HubSpot property mapping, implementation checklist