Manage
Action Credit Costs
Understand which GrowthEngene actions spend credits, why some actions cost more, and how plan credit limits apply to automation workflows.
Quick Answer
How Action Credits Work
GrowthEngene credits are workspace-level usage units. A plan gives the workspace a monthly credit allowance; it does not give a fixed number of specific actions.
When a workflow queues work, credits can be reserved first, then consumed after successful completion or refunded when eligible. The final usage depends on the action mix the user chooses.
The action-credit model is designed so users can understand where usage goes: browser work and LinkedIn pacing are comparatively cheap, while Hunter/Apollo enrichment and GPT-5.4-class AI generation are real paid-provider cost centers.
- 1-credit actions are usually lightweight capture, search, or orchestration actions.
- 2-10 credit actions usually require deeper traversal, lightweight enrichment, verification, or public text handling.
- 20+ credit actions should cover expensive providers, especially GPT-5.4-class AI generation and Hunter/Apollo contact extraction.
- Email, AI, and URL finder allowances can be shown separately when the underlying provider cost is tracked separately, but they should still map back to the same credit economics.
- Plan limits decide how many credits are available; safety limits still decide whether an account or sender may act now.
Key Cost Drivers
The most expensive points in the platform are paid provider calls and high-end AI generation. Browser sessions, LinkedIn account pacing, verification, storage of evidence, and operational auditability still matter, but Hunter/Apollo and GPT-5.4-class calls drive the hard cost.
A good credit model should charge more when an action uses scarce account budget, creates live external impact, requires deeper enrichment, calls paid providers, or needs stronger safety checks.
- Browser-heavy LinkedIn work uses worker capacity, session health checks, and anti-overlap controls.
- Outreach actions consume LinkedIn or sender reputation budget and require pacing, review, and audit trails.
- Contact extraction can use email discovery, verification, confidence scoring, and evidence persistence.
- GPT-5.4-class drafting, classification, job matching, and resume analysis consume model budget and need quality controls.
- URL finding and email finding can carry third-party provider cost even when no outreach is sent.
- Company-to-people enrichment traverses more pages and creates more downstream candidate records.
Recommended Credit Cost Table
These values are the current recommended starting point for customer-facing pricing. They should be treated as action-cost rules, not fixed action allowances inside a plan.
Admins can use the internal pricing catalog to override global action costs when provider costs, runtime cost, or product economics change.
- Visit LinkedIn profile: 1 credit per action.
- Scrape profile details: 1 credit per action.
- Scrape people, job, company, or post search results: 1 credit per action.
- Run an Action Plan: 1 credit per batch before step-level work.
- Scrape company jobs: 2 credits per action.
- Enrich company profile: 2 credits per action.
- Scrape company people: 2 credits per action.
- Extract contact info: 2 credits per action.
- Find or verify an email: 10-25 credits per record depending on provider path and verification depth.
- Company/domain search through Hunter or Apollo: 50-60 credits per search.
- Full contact extraction: 60-100 credits per target when it includes company search plus email/provider work.
- Generate GPT-5.4-class AI outreach draft: 20 credits per action.
- Run GPT-5.4-class job or resume analysis: 30 credits per action.
- Post a LinkedIn comment: 2 credits per action.
- Send LinkedIn connection request: 3 credits per action.
- Send LinkedIn message: 3 credits per action.
- Send outbound email: 3 credits per action.
- Share a LinkedIn post: 3 credits per action.
How Plans Use Credit Limits
Plan credit limits should be easy to reason about. Free gives a small trial wallet, Starter supports lightweight workflows, Growth supports multiple concurrent workflows, and Agency supports always-on automation.
The public pricing page should show plan-level credit limits, while this page explains action-level usage. Together they help users estimate how many searches, enrichments, drafts, and outreach actions fit into a plan.
- Free: trial-oriented limits for validating setup and exports.
- Starter: limited concurrency and enough monthly credits for lightweight outbound workflows.
- Growth: larger monthly credit pool, more automation slots, and onboarding support.
- Agency: higher monthly credit pool, more execution time, and strategist support for client-scale operation.
When Credits Are Not Enough
Having credits does not guarantee an action will run immediately. GrowthEngene may still pause because of account safety, sender setup, review mode, recipient quality, domain pacing, schedule windows, or missing required data.
This separation protects users: credits answer whether the workspace has usage allowance; safety controls answer whether the action should happen right now.
- A workspace can have credits but be blocked by LinkedIn session health.
- A sender can be configured but paused by daily or domain pacing.
- An outreach action can be blocked by missing resume, low confidence, suppression, or review requirements.
- A queued job can reserve credits and later refund them if it fails before eligible completion.
Proposed Action Cost Model
Plans should grant credit pools. They should not promise a fixed number of specific actions. The actual number of actions a user can run depends on which actions they choose and the active credit cost catalog.
Low cost: capture and planning
Use 1 credit for lightweight browser checks, search capture, profile capture, draft-only work, and orchestration overhead.
Medium cost: enrichment and public text
Use 2-10 credits when an action does deeper traversal, lightweight provider-backed enrichment, verification, or public comment preparation/execution.
High cost: paid providers and AI
Use 20+ credits when an action calls expensive providers such as Hunter, Apollo, or GPT-5.4-class models.
| Category | Action | Credits | Unit | Why It Costs This |
|---|---|---|---|---|
| LinkedIn account work | Visit LinkedIn profile | 1 | per action | Browser session time and profile-open pacing. |
| Check connection status | 1 | per action | Light browser/account check used before deciding whether outreach is allowed. | |
| LinkedIn data capture | Scrape profile details | 1 | per action | Browser extraction, parsing, and profile evidence storage. |
| Search capture | Scrape people search results | 1 | per action | Search page capture, dedupe, and result normalization. |
| Job discovery | Scrape LinkedIn job search | 1 | per action | Search capture plus job-list normalization. |
| Company discovery | Scrape company search | 1 | per action | Search capture and company matching. |
| Post discovery | Scrape LinkedIn posts | 1 | per action | Feed capture, post parsing, and relevance evaluation. |
| Post engagement | Save a LinkedIn post | 1 | per action | Low-risk post interaction with account pacing and audit logging. |
| Like a LinkedIn post | 1 | per action | Low-risk post interaction with account pacing and audit logging. | |
| React to a LinkedIn post | 1 | per action | Reaction action with account pacing, selected reaction metadata, and audit logging. | |
| Post a LinkedIn comment | 2 | per action | Public text action with review, pacing, and account safety checks. | |
| Share a LinkedIn post | 3 | per action | Public distribution action with higher account and quality risk. | |
| AI drafting | Draft a LinkedIn post comment | 1 | per action | Draft generation and review storage without live public posting. |
| Draft a LinkedIn post share | 1 | per action | Draft generation and review storage without live public distribution. | |
| Generate AI outreach draft | 20 | per action | GPT-5.4-class model call, prompt context, personalization evidence, retries, and review-ready draft storage. | |
| Run AI job or resume analysis | 30 | per action | Longer GPT-5.4-class reasoning request over job, resume, scoring, and structured output. | |
| Company enrichment | Scrape company jobs | 2 | per action | Deeper company page traversal and job evidence capture. |
| Enrich company profile | 2 | per action | Company identity resolution, page parsing, and metadata storage. | |
| People discovery | Scrape company people | 2 | per action | Company-to-people traversal with role filtering and dedupe. |
| Contact enrichment | Extract contact info | 2 | per action | Email/domain resolution, confidence checks, and evidence storage. |
| Email discovery | Find or verify an email | 2 | per record | Provider lookup, verification, confidence scoring, and evidence persistence. |
| Outreach | Send LinkedIn connection request | 3 | per action | Highest-risk LinkedIn action with pacing, safety, and audit checks. |
| Send LinkedIn message | 3 | per action | Live outreach action with recipient checks and account pacing. | |
| Send outbound email | 3 | per action | Live sender reputation impact with suppression, unsubscribe, domain pacing, and audit checks. | |
| Workflow orchestration | Run an action plan | 1 | per batch | Scheduler, queue, and orchestration overhead before step costs. |
Frequently Asked Questions
Why do outreach actions cost more credits?
Outreach actions cost more because they create live external impact and require stronger pacing, review, account-safety, and audit controls.
Can action credit costs change later?
Yes. GrowthEngene stores action costs in a pricing catalog so admins can update the active cost model as provider costs, product behavior, or plan economics change.
Are credits the same as email credits or AI credits?
Plan pages may expose separate limits for platform, email, AI, or URL finder usage. The action-credit guide explains the platform action cost model, while plan limits decide the total allowance.
What happens if a reserved action fails?
Reserved credits can be refunded when a job fails or is canceled before eligible completion. Successful actions consume the reserved credits.