OverviewHistoryStatsSecurity
npx skills add ...
Documentation
SKILL.md
npx skills add posthog/skills --skill posthog-onboarding-lead-research
Research and qualify onboarding team referral leads for PostHog. Use this skill when a TAE receives a lead from the onboarding team and needs a full research brief before deciding how to engage. Triggers on 'research this onboarding lead', 'onboarding team referred [company]', 'look into [company] from onboarding', 'qualify this onboarding referral', 'what do we know about [company] from onboarding', or any request to research a company that came through the onboarding pipeline. Also trigger when a TAE pastes a company name and mentions it's from the onboarding team, or says something like 'onboarding sent me [company]', 'got a handoff for [company]', or '[name] from onboarding sent me [company]'. This skill does deep research and qualification, then drafts outreach when the recommendation is to engage.
npx skills add posthog/skills --skill posthog-onboarding-lead-research
Deep-research a company referred from the onboarding team. Produces a complete brief: company overview, $2K+/month likelihood score, Vitally usage/billing analysis, onboarding team context, business growth trajectory assessment, and - when the recommendation is to engage - a draft outreach email with Slack channel guidance.
The PostHog onboarding team works with self-serve accounts that have a forecasted bill > $100 for an 8-week engagement. When accounts reach ~$1-1.5K/month and show growth potential, they hand them to the TAE via the "Sales Handoff" pipeline status in Vitally.
This skill does the research, scores the lead, and - when the recommendation is to engage - drafts the outreach. The TAE needs to understand the company, its trajectory, and the full context of the onboarding engagement before reaching out.
Ask the TAE for:
If the TAE only gives a company name, that's enough to start - pull the rest from Vitally and web research.
Run these five workstreams. Steps 1-2 use web search. Steps 3-4 use Vitally MCP. Step 5 combines web research with Vitally/SFDC/Harmonic data from traits. Interleave them efficiently — don't wait for one to finish before starting the next.
Goal: Understand who this company is and whether they're on a growth path that makes $2K+/month likely.
Web searches to run:
Gather:
Growth trajectory assessment: Based on the research, classify the company's growth trajectory:
| Trajectory | Description |
|---|---|
| Accelerating | Recent fundraise, hiring, product launches, or expansion. Usage will likely grow fast. |
| Steady growth | Established company, consistent hiring, stable product. Predictable growth. |
| Flat/uncertain | No recent funding, unclear growth signals, may be in maintenance mode. |
| Early stage | Seed/A, small team, product still finding market fit. High variance. |
Goal: Score the probability this account will cross $2K/month within 3-6 months.
Read references/scoring-model.md for the full scoring criteria.
Quick summary of the scoring:
Output a score from 0-100 with a tier:
Goal: Pull hard data on what this account is actually doing in PostHog and where the spend is going.
Search Vitally for the account. Try these in order:
search_accounts_advanced with the company nameOnce you have the account ID, pull:
get_account_full (with detailLevel: "full") — Gets everything: MRR, health scores, traits, product usage, billing data, team assignmentsget_account_health — Detailed health score breakdownAlways extract these three fields for the Quick Links section:
organization_id (in traits) → Vitally URL: https://posthog.vitally-eu.io/customers/{organization_id}sfdc.Id (in traits) → Salesforce URL: https://posthog.my.salesforce.com/{sfdc.Id}usage_dashboard_link (in traits) → direct Metabase usage dashboard linkExtract and analyze:
Current billing:
Usage trends (from traits):
Growth indicators:
Trajectory projection: Based on the trends, project where this account lands in 3 months:
Large companies sometimes have multiple PostHog orgs. Search for other accounts with similar names to flag consolidation opportunities.
Goal: Get the full history of what the onboarding team discussed, tried, and learned about this customer.
Use get_account_notes (with a high limit, e.g. 20) to pull all notes on the account. The onboarding team typically logs:
Summarize the onboarding narrative:
Use get_account_conversations to pull email threads and other conversations. These often contain:
Extract key details:
If the TAE or the onboarding team had calls logged in Granola, query for them:
This surfaces any call transcripts with additional detail beyond what's in Vitally.
Goal: Assess how fast the company's business is growing, because that directly predicts how fast their PostHog usage (and spend) will grow. This is the step that catches cases where Vitally billing data lags the real trajectory — a company doubling its customer base will eventually double its event volume, even if the billing hasn't caught up yet.
Web searches to run:
Also cross-reference signals already gathered from Vitally traits:
sfdc.harmonic_headcount__c vs sfdc.harmonic_headcount_180d__c — headcount deltasfdc.harmonic_headcountEngineering__c vs sfdc.harmonic_headcountEngineering_180d__c — eng headcount deltasfdc.harmonic_web_traffic__c vs sfdc.harmonic_web_traffic_90d__c / sfdc.harmonic_web_traffic_180d__c — traffic trendssfdc.webtraffic_growth__c — percentage growthsfdc.lnkd_follower_change__c — LinkedIn momentumsfdc.eng_headcount_change__c and sfdc.headcount_eng_growth__c — engineering team growthGather and assess these six growth vectors:
Map business growth to PostHog spend growth:
After gathering these signals, explicitly connect the company's business growth to expected PostHog usage growth. This is the key insight - it tells the TAE whether the billing trajectory in Vitally is understating the real growth potential.
The mapping logic:
Rate the business growth trajectory:
| Rating | Criteria |
|---|---|
| 🔥 Strong accelerating | 2+ of: recent fundraise, 50%+ revenue growth, 3+ product launches, rapid hiring, expanding into new markets. PostHog billing likely understates real trajectory. |
| 📈 Steady upward | Consistent hiring, stable funding, 1-2 new products, growing customer base. PostHog billing trajectory is probably accurate. |
| ➡️ Flat/stable | No recent funding, minimal hiring, no new products. PostHog billing is what you see. |
| 📉 Contracting | Layoffs, funding concerns, product sunsetting. PostHog spend may decline. |
Based on all five research workstreams, recommend one of these plays:
Recommend when:
Draft an email AND a Slack channel welcome message. See Step 7 for drafting guidance.
Recommend when:
Draft a short email only. No Slack channel.
Recommend when:
When recommending Skip, you MUST provide a DQ Reason - a plain-text summary of why this lead is not qualified. The DQ reason must be 250 characters or fewer so the TAE can copy-paste it directly into the Salesforce disqualification field. Be specific - name the concrete signals, not generic language.
Important: Non-response is NOT a disqualification signal. If the onboarding team reached out and the customer simply didn't reply, that is not evidence they don't want to hear from PostHog - they may have been busy, missed the email, or the timing wasn't right. Do not cite "no engagement" or "didn't respond to onboarding" as a DQ reason. The only engagement-related DQ signal is an explicit opt-out - the customer actively responded to say they're not interested. Even then, if their spend is meaningful and a TAE could offer genuine value (e.g. volume discounts, billing optimization), an explicit opt-out from onboarding doesn't necessarily mean they'd reject a TAE with a different angle.
Good DQ reasons:
Bad DQ reasons:
Recommend when:
When the recommended play is Outreach or Light Touch, draft the email using the guidance below. Read references/writing-style.md before drafting.
All draft emails must use this blockquoted format for easy copy-paste:
When the play is Outreach (Slack Channel + Email), also draft a Slack channel welcome message in the same blockquoted format after the email.
These are the same principles used across all PostHog lead skills. Follow them exactly.
Lead with the Slack channel. For multi-user accounts, the shared Slack channel is the most valuable thing you're offering - a direct line to someone technical at PostHog. Open with it. Explain why it's useful for their specific situation based on your research. Don't bury it.
Write like a fellow engineer, not a salesperson. These are competent developers. The TAEs are deeply technical. Write the way you'd message a colleague - direct, concise. No praise or cheerleading about their setup. "Your team's ramped up fast!" reads as patronizing to engineers who know exactly what they've done.
Start with an observation about them, not about you. The first real sentence should reference something specific from Vitally or your research - which products they're using, what their team is building, a specific usage pattern. Never open with "I'm [name], your PostHog contact."
Keep it short. The initial email should be 3-5 sentences. Slack channel + why it's useful + one specific observation or optimization + a question. Save the details for Slack.
One question per email. More than one question overwhelms and reduces response rate. Pick the most interesting one.
Be direct about billing when relevant. Unlike big fish outreach, onboarding handoff leads already have meaningful spend. It's OK to reference costs - pair every cost mention with an actionable way to reduce it.
Only reference what you can verify. Safe: Vitally usage data, public website, confirmed funding rounds. Never cite enrichment data (Clearbit/Harmonic tech stack fields) as fact. If you state something about their setup and it's wrong, you lose credibility.
Follow references/writing-style.md for the full guide. Key rules:
Multi-user account, growing spend: Lead with the Slack channel. Ground it in their usage: "Setting up a shared Slack channel for you - with [N] people using [products] across [N] projects, a direct line for technical questions will save your team time." End with a question about what they're building or trying to learn.
Multi-user account, flat or declining spend: Lead with a specific optimization. If you identified an issue (autocapture waste, stale flags, replay oversampling), that's the hook: "Noticed about [X]% of your events are autocapture without associated actions - [link to reducing event volume] covers how to trim that without losing anything useful." The Slack channel is where you help them act on it.
Single user, meaningful spend: Short and direct. No Slack channel. One specific optimization with a link. "Quick heads-up - at your current usage, [specific optimization + link]. Happy to help if questions come up."
Previously engaged with onboarding (had calls, completed onboarding): Reference the onboarding work. "Your team worked with the onboarding team on [what they covered] - wanted to make sure you have a direct line for questions now that you're past the initial setup." This creates continuity rather than feeling like a cold handoff.
Never engaged with onboarding (unresponsive, skipped): Don't reference the onboarding team at all - they clearly didn't want that engagement. Lead with something useful: a specific optimization, a resource relevant to what they're building, or the Slack channel as a low-friction support resource.
Example 1: Multi-user, high-growth, engaged with onboarding
To: team@acmecorp.com Subject: Slack channel for Acme <> PostHog
Hey Acme team,
Setting up a shared Slack channel for you - with 57 people in the account across analytics, replay, and error tracking, a direct line for questions will be useful as you scale.
One thing that might be worth exploring: Feature Flags for safe rollouts. With your new product line launching as a platform, gradual rollouts and instant kill switches are the kind of thing that keeps deploys boring. Happy to walk through the setup in Slack.
What's the main thing your team is trying to learn from the analytics right now?
[TAE name]
Example 2: Multi-user, flat spend, optimization hook
To: eng@example-payments.com Subject: Slack channel for Example Payments <> PostHog
Hey Example Payments team,
Sending over a Slack channel invite - with 42 people using analytics, flags, and replay, it'll be the fastest way to get answers on implementation or billing as things come up.
One quick win: if your team is running A/B tests on checkout flows, Experiments layers directly on top of your existing Feature Flags setup. For a payments product, measuring conversion impact per variant is where the real signal is.
What's the team trying to optimize right now?
[TAE name]
Example 3: Single user, growing spend, never engaged
To: dev@example-ai.com Subject: Your PostHog setup
Hey,
Noticed you're using LLM Analytics alongside product analytics - that's a sharp setup for monitoring AI-generated content quality alongside user behavior.
If you're tracking prompt-to-output quality, this guide on LLM observability covers how to connect traces to user outcomes so you can see which prompts actually help students learn vs. just generate text.
Happy to help if questions come up.
[TAE name]
Present the research brief as a structured report with clear sections. Use the Visualizer for the growth trajectory chart if the data supports it.
After completing all individual research briefs, output a Summary & Priority Stack Rank table at the top of the combined output (or as a final summary section). Rank accounts by score descending.
Column guidance:
references/scoring-model.md - Full $2K+ spend likelihood scoring model with weights and criteriareferences/writing-style.md - PostHog writing style guidelines for email draftingreferences/writing-style.md. Hyphens only (never em/en dashes). No bare URLs. No feature laundry lists. One question per email. Lead with the Slack channel for multi-user accounts.> **To:** [full email address - e.g. contact@acmecorp.com]
> **Subject:** [subject line]
>
> [email body]
>
> [TAE's name]## [Company Name] - Onboarding Lead Research Brief
### Quick Links
- **Vitally:** [https://posthog.vitally-eu.io/customers/{organization_id}](https://posthog.vitally-eu.io/customers/{organization_id})
- **Salesforce:** [https://posthog.my.salesforce.com/{sfdc.Id}](https://posthog.my.salesforce.com/{sfdc.Id})
- **Usage Dashboard:** [{usage_dashboard_link}]({usage_dashboard_link}) *(or "Not available" if trait is blank)*
### Company Overview
- **What they do:** [one sentence]
- **Industry:** [vertical]
- **Size:** [employees] ([engineering count] engineers)
- **Funding:** [stage], [amount] ([date])
- **HQ:** [location]
- **Growth trajectory:** [Accelerating / Steady / Flat / Early stage]
### $2K+ Spend Likelihood
**Score: [X]/100 - [High / Moderate / Low / Unlikely]**
Key factors:
- [Factor]: [+/- impact]
- [Factor]: [+/- impact]
- [Factor]: [+/- impact]
Projection: Currently at $[X]/month - projected $[X]/month in 3 months
### Vitally Usage & Billing
- **Current MRR:** $[X]
- **Products in use:** [list]
- **Products NOT in use:** [list - these are expansion opportunities]
- **Event volume:** [current] ([trend])
- **Active users:** [count] ([trend])
- **Health score:** [X]/10
**Billing trajectory:**
[Chart or description of the trend - is spend growing, flat, or declining?]
**Expansion opportunities:**
- [Product they're not using that fits their use case]
- [Volume tier they're approaching]
- [Config optimization that could increase value]
### Onboarding Team Context
**Engagement level:** [Engaged / Partially engaged / Unengaged]
**Onboarding rep:** [Name]
**What happened:**
[Narrative summary of the onboarding engagement - what was discussed, what was done, what's left open]
**Key quotes or insights from the customer:**
- [Any notable things the customer said or asked about]
**Open items / unfinished business:**
- [Anything the onboarding team started but didn't finish]
- [Any promises made to the customer]
### Business Growth Trajectory
**Rating:** [Strong accelerating / Steady upward / Flat/stable / Contracting]
**Key growth signals:**
- **Revenue:** [what's known about ARR growth, Inc. 5000 ranking, etc.]
- **Funding:** [recent round, amount, investors, what it's earmarked for]
- **Product velocity:** [new launches, features, platform plays in last 12 months]
- **Customer expansion:** [new segments, partnerships, geographic expansion]
- **Headcount:** [total delta, eng delta, GTM hiring signals]
- **Web/brand momentum:** [traffic trends, LinkedIn growth, awards/recognition]
**How this maps to PostHog spend:**
[Explicit analysis connecting each business growth vector to expected PostHog usage growth. e.g. "More customers on their platform - more end-user sessions - event volume will grow faster than current billing suggests." This is the most actionable part - it tells the TAE whether the Vitally billing trajectory understates or accurately reflects the real growth potential.]
**Score adjustment:** [If the business growth signals are strong enough to warrant adjusting the $2K+ likelihood score from Step 2, note the adjustment and reasoning here. e.g. "Business growth trajectory bumps the score from 72 to ~80 - billing data lags customer growth."]
### Recommended Play
**[Outreach / Light Touch / Skip / Wait and Watch]**
[Reasoning for the recommendation - 1-2 sentences.]
[If Skip: include DQ Reason below]
**DQ Reason (for Salesforce):**
`[250 characters or fewer - specific, concrete, copy-pasteable]`
### Draft Outreach
[If the play is Outreach or Light Touch, include the draft email in blockquoted format. If Outreach with Slack channel, also include the Slack welcome message.]
> **To:** [full email address]
> **Subject:** [subject line]
>
> [email body]
>
> [TAE name]
[If multi-user Outreach, also include:]
**Slack channel welcome message:**
> [Slack message body]
>
> [TAE name]
**Suggested angle:** [One sentence explaining why this hook was chosen and what it's designed to accomplish.]## Summary & Priority Stack Rank
| Rank | Account | Score | Tier | MRR | Trend | Action |
|------|---------|-------|------|-----|-------|--------|
| 1 | [Company] | [X]/100 | High confidence | $[X] | Growing | Outreach |
| 2 | [Company] | [X]/100 | Moderate confidence | $[X] | Flat | Light Touch |
| 3 | [Company] | [X]/100 | Unlikely | $[X] | Declining | Skip |