Agentic Marketing

Use Darwin to find Aaru, Seoul-based beauty creator Kyungjin Son, Jasper, and Smartly to test UGC concepts and launch a Purito Seoul skincare campaign in Seoul, South Korea.

Plan a Purito Seoul serum campaign in Seoul, South Korea with Kyungjin Son: three original UGC concepts, six ad variations, agreed usage rights, and a four-week media run within an $18,000 ceiling.

Campaign pilot brief for Purito Seoul, with Kyungjin Son as the proposed creator. Confirm brand authorization, creator availability, and usage rights before booking. Savings below are estimates for one campaign.

Intent

Launch a Seoul campaign

Produce six creator-led skincare ads

Three original skincare videos, each with two approved hooks, produce six ad variations for the Seoul campaign.

Time savings

Estimated hours saved per campaign

Estimate: 26 coordination hours reduced to 16 for one six-variation campaign.

Net value

$800

Estimated net time value per campaign

Estimate: 10 hours × $100/hour, less $200 in incremental coordination costs. Creator fees and media remain separate; recovered time is not a cash discount.

Find the creator and services for a Seoul skincare campaign

ChatGPT uses Darwin Search to compare Aaru concept research, Kyungjin Son’s published skincare work, Jasper creative support, and Smartly media execution. The proposed Purito Seoul brief keeps the product, creator, rights, and research question explicit before any work is commissioned.

A skincare team needs a creator whose work fits the product, a clear brief, usable advertising rights, and evidence that the final ads resonate. A simulated response alone cannot supply authentic creator footage or prove sales.

Start with a real creator and a specific product

The campaign product is Purito Seoul Wonder Releaf Centella Serum Unscented. The proposed creator is Kyungjin Son (@kitzi_cosmos), whose public profile lists Seoul, South Korea and a focus on skincare and makeup UGC. That establishes relevant public experience; it does not establish a new campaign agreement, a quoted rate, or permission to reuse existing footage.

The proposed campaign targets adult skincare shoppers in Seoul, South Korea with Korean-language videos and optional English subtitles. The brief requests three original concepts: a serum-texture close-up, a morning routine filmed in Seoul, and a product-use explanation. Language, location, and subtitling deliverables are confirmed with the creator before booking. Each concept becomes two ad variations. Product shipment, a mutually agreed testing period, production time, revisions, and approval all precede the four-week media run. The creator describes her actual experience and uses only claims the brand can support.

Sources: Kyungjin Son: skincare UGC portfolio · Purito Seoul: the campaign serum

Simulate UGC concepts, then commission a person

Aaru can explore how a defined audience of adult skincare shoppers in Seoul might respond to different creative concepts. The team can compare a texture-led opening with a routine-led opening while holding the product and offer constant. This is an audience simulation around UGC formats, not a synthetic version of Kyungjin Son or a substitute for her creative judgment.

Jasper turns the three selected concepts into a production brief and two hook variations per video. Kyungjin Son reviews what is feasible and records the original content after an agreement. Smartly is responsible for approved media execution and reporting. Giving each participant a distinct deliverable avoids treating four separate services as one already-connected agent.

Simulation informs the brief; original footage and observed ad results establish what happened.

Aaru: Simulate audience response to UGC concepts. Compare creative hypotheses for adult skincare shoppers in Seoul; retain the simulation assumptions.

Kyungjin Son: Film original skincare UGC in Seoul. Agree three original videos, two hooks per video, fees, revisions, and advertising rights before filming.

Jasper: Develop the brief and ad variations. Prepare the shot list, draft hooks, and captions for creator and fluent Korean-language review.

Smartly: Run approved ads and report results. Use the authorized ad account, approved assets, Seoul audience, and agreed spend cap.

Sources: Aaru: behavior simulation · Jasper: Ad Campaign Agent · Smartly: Meta campaign workflows

Agree the rights before pricing the media run

The US$18,000 campaign ceiling must cover research, creator production, editing, software, licensing, and media. Obtain separate terms for organic use, paid ads, territories, duration, revisions, and any account-level advertising permission. Specify whether creator-channel posting is included and limit usage to the contracted rights. Korean-won quotes are converted at the agreed exchange rate before checking the ceiling. Publication requires the actual rights and brand approval.

From the search brief to a reviewed selection

A campaign brief and provider proposals identify the creator, product, six variations, usage rights, and complete cost for approval.

Darwin Search API

via ChatGPT MCP

One brief becomes a capability selection

search request

Find Aaru concept simulation, skincare UGC creator Kyungjin Son, Jasper campaign development, and Smartly media execution for a Purito Seoul Wonder Releaf Centella Serum Unscented campaign in Seoul, South Korea. Scope six ad variations, creator rights, and a four-week media run within $18,000; confirm creator availability before booking.

query contains this brief; limit: 10 bounds the result set. $18,000 campaign ceiling; actual provider quotes establish cost, scope, and availability.

Request example

// Search is read-only. Preserve the selected capability identifiers and revision from the result.
// mcp is your connected, authorized MCP client.
await mcp.callTool({
  name: "search",
  arguments: {
    query: "Find Aaru concept simulation, skincare UGC creator Kyungjin Son, Jasper campaign development, and Smartly media execution for a Purito Seoul Wonder Releaf Centella Serum Unscented campaign in Seoul, South Korea. Scope six ad variations, creator rights, and a four-week media run within $18,000; confirm creator availability before booking.",
    limit: 10,
  },
});

Keep the exact selection
Preserve aiId, capabilityId, and capabilityRevision for each chosen owner. Keep a returned searchAttributionId so the assignment stays linked to its discovery.
Review the provider scopes before commissioning work
Campaign pilot brief for Purito Seoul, with Kyungjin Son as the proposed creator. Confirm brand authorization, creator availability, and usage rights before booking. Savings below are estimates for one campaign.
Carry the selection into Act
Darwin must verify an execution bridge and its input contract for the exact capability ID and revision. The current Search response does not include an Act-ready input contract. Until that bridge is verified, do not start the listing through Act. Search does not authorize work or spending.

Provider capabilities and sources

Aaru · Simulate audience response to UGC concepts: Simulate audience response to UGC concepts. Confirm the provider agreement and supported execution route.; Aaru: behavior simulation
Kyungjin Son · Film original skincare UGC in Seoul: Film original skincare UGC in Seoul. Confirm the provider agreement and supported execution route.; Kyungjin Son: skincare UGC portfolio
Jasper · Develop the brief and ad variations: Develop the brief and ad variations. Confirm the provider agreement and supported execution route.; Jasper: Ad Campaign Agent
Smartly · Run approved ads and report results: Run approved ads and report results. Confirm the provider agreement and supported execution route.; Smartly: creative and media

Act

Produce six approved ad variations and launch the campaign

Subject to approved agreements, Aaru tests the proposed creative directions, Jasper prepares the production brief, Kyungjin Son records original footage, and Smartly runs only the approved ads for the Seoul audience, using the agreed geography, language, budget, and four-week schedule. ChatGPT coordinates the review checkpoints through Darwin; the creator and brand retain control over their commitments and claims.

Use Aaru findings to sharpen the creative question

ChatGPT submits the approved product context and audience definition to the selected Aaru workflow. The output informs which three concepts to develop and how to refine their hooks, with its assumptions kept beside the recommendation. Jasper prepares a Seoul shoot brief, shot list, Korean-language draft hooks, and optional English subtitle copy for fluent human review; the team reviews the brief with Kyungjin Son before scheduling filming and reserves Smartly setup until the creative and rights are ready.

A favorable simulated response is a reason to test an idea, not evidence that an ad will convert. If concepts are close or findings are unstable, document that uncertainty instead of declaring a winner. The creator can suggest a more natural execution while retaining the agreed product message.

Film original content and review the six variations

Kyungjin Son would film the agreed product demonstration in Seoul after shipment and the agreed testing period. The proposed deliverables are three original videos with two approved hooks each, producing six ad variations for the Seoul audience. Jasper supports captions and hook variations, while the creator and brand check that the edit reflects what was actually said and shown. Skin imagery is not altered to invent results, and no unsupported before-and-after claim is introduced by an automated edit.

Each approved file gets a version, concept label, rights period, and review status. Feedback names the exact clip or line to change, so a new edit cannot silently replace the approved asset. Existing portfolio footage is a reference for fit, not free campaign material.

Sources: Kyungjin Son: skincare UGC portfolio · Jasper: Ad Campaign Agent

Launch approved ads and compare observed performance

Smartly receives the six approved variations, Korean-language product landing page, Seoul audience settings, rights limits, and spend cap through the brand’s authorized Meta ad account. The proposed placements are Instagram Reels and Stories; platform availability and targeting are checked before launch. ChatGPT tracks the launch record and review questions in Darwin. The marketer checks actual delivery, attribution settings, spend, and conversion data; a high click-through rate alone is not a claim of profitable acquisition.

Sources: Smartly: Meta campaign workflows

From selected capabilities to coordinated work

Original creator footage and approved copy become six launch-ready variations with documented usage rights and a media cap.

Darwin Act API

via ChatGPT MCP

1start_actionStart one scoped assignment
Start the authorized Aaru assignment using its verified capability ID, revision, agreed scope, and permitted inputs.
2get_actionRead progress and recover the same work
Read current progress, review questions, and returned artifacts for each assignment.
3continue_actionSupply a requested handoff
Return the specific approved answer or requested revision while preserving the agreed scope.
4end_actionClose the work and read its outcome
Close accepted work with the current revision after the buyer checks the deliverables.

Request fields and returned state

`start_action`

// Only after Darwin verifies an execution bridge and input contract for this capability revision. Persist startRequestId; reuse it for retries.
// mcp is your connected, authorized MCP client.
await mcp.callTool({
  name: "start_action",
  arguments: {
    capabilityId: selectedCapability.capabilityId,
    capabilityRevision: selectedCapability.capabilityRevision,
    inputs: capabilityInputs,
    requestId: startRequestId,
  },
});

capabilityId + capabilityRevision: Exact identifiers from the selected Search result, not names reconstructed from text.
inputs: Only the fields required by the verified capability input contract. The task brief below describes the work, not a universal JSON schema.
requestId: A new idempotency key for this assignment. Reuse it only when retrying this same start.

Retain actionId, revision, lifecycle, status, and availableActions. An accepted or running Action is not completed work.

`get_action`

// Use the actionId returned by start_action. Read state before deciding on the next operation.
// mcp is your connected, authorized MCP client.
await mcp.callTool({
  name: "get_action",
  arguments: {
    actionId,
  },
});

actionId: The identifier returned by start_action. Reuse it after an interruption; do not start a duplicate task.

Read status, result, actionRequired, availableActions, revision, lifecycle, and outcome. The result content depends on the selected capability.

`continue_action`

// Call only when availableActions includes update. Persist updateRequestId and retry only the identical update.
// mcp is your connected, authorized MCP client.
await mcp.callTool({
  name: "continue_action",
  arguments: {
    actionId,
    message: handoffMessage,
    requestId: updateRequestId,
  },
});

actionId + message: Send the requested clarification or safe output references to the existing Action, only when availableActions includes update.
requestId: A new key for this update; reuse it only to retry the identical update.

Reread the Action after the update. An ordinary message never approves an interaction or grants provider access.

`end_action`

// Finish only when permitted by availableActions and the work is checked. Persist endRequestId; read until lifecycle is ended.
// mcp is your connected, authorized MCP client.
await mcp.callTool({
  name: "end_action",
  arguments: {
    actionId,
    expectedRevision: latestAction.revision,
    intent: "finish",
    requestId: endRequestId,
  },
});

actionId + expectedRevision: Use the same Action and the exact revision from the latest get_action response.
intent + requestId: Use finish for completed work, with a new stable key for that closure request, when the current Action permits it.

An ending lifecycle has no final outcome. Read get_action until ended, then retain the returned succeeded, failed, canceled, or unknown outcome.

Action state, permissions, and recovery

State controls the next call: availableActions is the authority for mutations. Poll get_action with bounded backoff; pause while a person completes a hosted step, then reread the same Action.
Use the authorized AI: Actions use the caller's active authorized AI, or an explicitly authorized actingAiId. Carry a returned target AI only when the capability requires it; a public listing is not a permission grant.
Keep secure steps separate: Darwin OAuth grants the approved scopes. Provider authentication and any exact approval or payment request are separate interactions. Use the first-party webLink; never send credentials in task messages.
Close every assignment: Use end_action with the current revision and intent: finish when permitted. Follow ending to ended through get_action. A recorded outcome does not replace checking the delivered work.

Outcome

Review the creative, campaign results, and actual spend

The proposed handover connects Aaru’s concept findings, Kyungjin Son’s licensed footage, Jasper’s approved variations, and Smartly’s delivery report. ChatGPT distinguishes what was simulated, what was filmed, and what actually ran.

Connect the creator’s work to the campaign evidence

The report lets the marketer follow a concept from Aaru’s simulation through Jasper’s brief, Kyungjin Son’s source footage, and Smartly’s live ad record. It includes failed or inconclusive variations as well as promising ones. The practical outcome is a reusable account of what was approved, what ran, and what should be tested next.

Use the creator portfolio to assess fit before booking. After launch, use observed ad delivery, spend, and conversions to decide which variations to run again. Retain source footage and approvals so the team can produce follow-up edits within the agreed rights.

Accept the assets, rights, and delivery record together

All six variations must match an approved source file and the agreed usage rights. The report separates simulated audience response from real ad delivery and identifies actual spend. Rights expiration or an unresolved claim prevents continued use even when an asset performs well.

Calculate the net value of campaign coordination

For one six-variation campaign, the pilot estimate reduces team coordination from 26 to 16 hours. Ten hours at an assumed $100/hour is $1,000 of recovered capacity; subtract an estimated $200 in incremental coordination costs for $800 in net value for the campaign. Creator fees, media, and research remain in the campaign budget.

The estimate does not promise sales, a cheaper creator, or less time for product testing. Actual production schedules and licensing terms come from the engagement. Measure coordination hours and reconcile invoices before using the estimate in a business case.

The result and the effort behind it

The campaign record links each live ad to its source footage, approvals, rights, actual spend, and observed results.

1. Search
Find the right capabilities
A campaign brief and provider proposals identify the creator, product, six variations, usage rights, and complete cost for approval.
2. Act
Coordinate the handoffs
Original creator footage and approved copy become six launch-ready variations with documented usage rights and a media cap.
3. Outcome
Return the complete result
The campaign record links each live ad to its source footage, approvals, rights, actual spend, and observed results.

Pilot estimate

Less searching. Fewer manual handoffs.

Estimated buyer effort for the same scope and acceptance criteria. Track actual hours during the pilot to compare with these estimates.

Time to a reviewed selection

12 working hours without Darwin; 6 with Darwin in the pilot estimate.

6h lessestimated research effort

Without DarwinWith Darwin

Cumulative working hours

Research and comparison work accumulate until the buyer has reviewed the selection. Delivery time is separate.

Human effort across the same scope

26 staff-hours without Darwin; 16 with Darwin in the pilot estimate.

10h lessestimated coordination effort

Without DarwinWith Darwin

Staff-hours by work role

Work roles may belong to the same person or run in parallel. Specialist production, fulfillment, waiting and provider execution are excluded from both columns.

Calculation and chart data

A complete brief and the required access are available. The same buyer acceptance checks apply in both approaches.
ChatGPT organizes discovery and handoffs; people still approve scope and review results. Estimated savings come from research and coordination.
Discovery is included in total effort. Time value measures recovered capacity; basket savings measure a difference in purchase price.
Record actual hours and costs during the pilot to calculate the achieved savings.

Discovery: estimated cumulative working hours
Milestone	Without Darwin	With Darwin
Brief	2h	2h
Research	7h	4h
Compare	10h	5h
Select	12h	6h

Team effort: estimated staff-hours
Role	Without Darwin	With Darwin
Marketing	12h	7h
Creative review	6h	4h
Media review	5h	3h
Purchasing	3h	2h