Training Manual

Learn how to use Pinkk.

Everything from influencer marketing fundamentals to running AI-powered intelligence reports. Seven modules, practical examples, and interactive demos — at your own pace.

1

Module 1 of 7

Influencer Marketing Fundamentals

What you'll learn

  • • Why the influencer marketing industry has a trust problem
  • • The key metrics marketers use — and why they fall short
  • • Creator tiers and what they mean for campaign decisions
  • • Industry standard terminology every marketer should know
  • • How Pinkk fits into the modern vetting workflow

The industry today

$24B

Projected global influencer marketing spend in 2026

$4.8B

Estimated annual spend reaching audiences that don't exist — roughly 1 in every 5 dollars

47%

Of brands using influencer marketing have encountered some form of fraud

The industry has scaled faster than the tools used to vet it. Follower counts are easy to inflate. Engagement rates can be manufactured. Traffic sources are opaque. And the platforms themselves don't have a financial incentive to surface these problems for you.

Key metrics — and what they actually tell you

Creator tiers

Tier classification is used across the industry to benchmark expectations for reach, engagement, and cost. These are approximate — platform algorithms and niche communities can shift typical behaviour significantly.

Common fraud signals

These are the patterns that surface most often when audience quality is questionable. None of them alone is conclusive — Pinkk is conservative by design and flags signals as indicators, not verdicts.

Where Pinkk fits in your workflow

Step 1

Shortlist candidates

Use discovery tools, referrals, or your own research to identify creators worth investigating.

Step 2 — Pinkk

Run intelligence reports

Add each profile to Pinkk, run the relevant gene type, and get structured, confidence-rated findings.

Step 3

Make the decision

Use findings to inform your brief, negotiate rates, or rule out poor-quality traffic before committing budget.

Knowledge check — Module 1

2

Module 2 of 7

Getting Started with Pinkk

What you'll learn

  • • The three-axis mental model for navigating Pinkk
  • • How credits work and what each action costs
  • • Free tier vs subscriber — what changes and what doesn't
  • • The onboarding flow from sign-up to first report
  • • How to orient yourself inside the dashboard

The three-axis mental model

Pinkk is organised around three distinct questions. Keep them separate and navigation becomes intuitive.

Axis 1 — Account

"Whose account is this?"

Team, Wallet, Settings, API tokens. Accessed via the account dropdown and the team context switcher in the header.

Axis 2 — Subjects

"Who am I studying?"

Profiles and Workspaces. These are the creators, brands, and affiliates you're researching.

Axis 3 — Reports

"What did the analysis find?"

Reports, Comparisons, Changes, Workspace intelligence, and Checks. All output lives here.

Primary navigation

The same four destinations appear on desktop, mobile, and the PWA — in the same order.

Navigation bar
Dashboard Profiles Workspaces Reports

Dashboard

Credits, recent reports, stats, notifications

Profiles

All the creators, brands, and affiliates you're researching

Workspaces

Groupings of profiles by campaign, client, or shortlist

Reports

All output: reports, comparisons, changes, checks

How credits work

Credits are the currency for running analyses. Every action that triggers AI research costs credits. Actions that retrieve already-recorded results are free.

Sign-up bonus

5 credits

Free on registration, no card required

Per analysis

1 credit

Deducted when you run any gene type

Checks (live)

1 credit

Free if answered from recorded evidence

On failure

Refunded

Credits are returned if a job fails

Credit lifecycle

Register → +5 credits Run analysis → −1 credit Buy pack → +10/25/100 Subscribe → +monthly credits

Free vs subscriber

Feature Free Subscriber

Your first session — step by step

Knowledge check — Module 2

3

Module 3 of 7

Profiles & Workspaces

What you'll learn

  • • The three profile types and when to use each
  • • Intelligence lenses and vertical packs
  • • What platform handles to add and why
  • • Sable Autofill and how it speeds up profile creation
  • • Organising profiles into Workspaces

Profile types

Every profile you create is one of three types. The type determines which genes are available and what the report research focuses on. Once a profile has been analysed, the type is locked and cannot be changed.

Intelligence lenses

A lens (also called a vertical pack) tells Pinkk the industry context for a profile. It changes the custom fields shown, the gene options available, and the language used in reports. Like profile type, the lens is locked after the first analysis.

Supported platforms

Add as many platform handles as you have. More data points = better cross-platform context in the report. At minimum, add the platform where the creator is most active.

Sable Autofill

Sable is the AI intelligence layer powering Pinkk. On the profile creation and edit forms, each field has an autofill button. Click it and Sable researches the answer — you review and keep or discard. You stay in control.

Profile edit — field autofill demo

Click "Autofill" to see Sable research and fill the field automatically.

Single-field autofill

Available to all users. One click per field.

Metrics autofill

Fills follower count, engagement rate, avg views, verified status in one click.

Batch autofill

Subscriber only — fills all fields in one action. Huge time saver at scale.

Workspaces

Workspaces group profiles. A profile can belong to multiple workspaces. Create a workspace per campaign, per client, or per shortlist — whatever matches how your team works.

Example workspaces

Knowledge check — Module 3

4

Module 4 of 7

Genes & Reports

What you'll learn

  • • What a gene is and the five built-in types
  • • When to use each gene type
  • • How to read a Pinkk report — sections, confidence levels, conservative language
  • • Full interactive example reports for all five gene types
  • • What a report reference number (PKK-XXXX-XXXX) means

What is a gene?

A gene is an analysis type — a named, configurable AI analysis that produces a structured report. Think of genes as different lenses through which Pinkk examines the same profile. Each one asks a specific question and returns a report focused on that question.

Profile + Gene + AI = Structured Report

You bring the profile. You choose the gene. Sable does the research and returns a conservative, confidence-rated report within minutes.

The five built-in genes

Which gene should I run? — quick selector

What is your primary goal?

Recommended:

Example reports — all five gene types

These are realistic mock reports showing exactly how Pinkk structures its findings. The profile is fictional. Click each tab to switch gene type.

Reading a report — key concepts

Conservative language

Pinkk always uses cautious phrasing: "may indicate", "suggests", "could be consistent with". This is deliberate — the platform never makes fraud judgments or identity claims. If a finding is strong, it says so. If it's uncertain, it says that too.

Confidence levels

Every report carries a confidence level: High, Medium, or Low. This reflects the quantity and independence of the data sources reviewed, not the severity of any findings. A Low confidence report means less data was available — not that something is wrong.

Reference numbers

Every report gets a unique reference: PKK-XXXX-XXXX for analyses, PKK-CMP-XXXX-XXXX for comparisons. Use /verify to prove authenticity publicly without exposing the findings. The badge can be embedded on your website.

Recommended checks

Most reports end with a "Recommended checks" section — specific follow-up questions or actions the AI suggests based on the findings. These are starting points for your own due diligence, not conclusions.

Knowledge check — Module 4

5

Module 5 of 7

Advanced Features

What you'll learn

  • • Profile Comparisons — how to run a side-by-side verdict
  • • Change tracking / Diffs — how to track what changed between reports
  • • Checks — evidence research for anything without a profile
  • • Private share links — sharing reports with clients without a login
  • • Public report verification and the embeddable badge

Profile Comparisons

A Comparison takes two completed analyses of the same gene type and produces a direct, structured verdict. Use it when shortlisting two creators for the same campaign slot.

Step 1

Run a Traffic Gene on both profiles

Comparisons work between two completed analyses of the same gene type. Both reports must exist first.

Step 2

Go to Reports → Comparisons → Create

Select both analyses. Optionally add a focus question (e.g. "Which is better for a fitness campaign targeting 25–35 UK women?").

Step 3

Read the comparison report

You get a direct verdict, section-by-section signal breakdown, an analyst question response (if provided), and links back to both source reports.

PKK-CMP-4821-9037 — Example Comparison Mock

Profile A

@fitnesswithkara (Instagram)

Traffic Gene · 2.3M followers

Profile B

@liveliftlara (Instagram)

Traffic Gene · 1.1M followers

Direct verdict

Profile B (@liveliftlara) shows stronger traffic quality signals despite a lower follower count. Source attribution is more diversified, engagement patterns are more consistent across content types, and audience anomaly indicators are notably lower. For a campaign prioritising authentic reach over raw scale, Profile B appears the stronger candidate based on the available signals. Profile A's larger following may still justify consideration for pure impressions-based objectives, but the quality differential is meaningful.

Section-by-section breakdown

Section
Profile A
Profile B
Source attribution
Concentrated
Diversified
Quality signals
Mixed
Strong
Risk signals
Elevated
Low

Change tracking (Diffs)

Run the same gene on a profile at two different points in time and Pinkk can produce a structured diff — showing exactly what changed between reports. Essential for ongoing creator relationships and periodic account reviews.

When to use diffs

  • • Mid-campaign check — has anything changed since you vetted them?
  • • Renewal review — how has the profile evolved over 6 months?
  • • Post-controversy review — did audience patterns change after a public incident?

How to create a diff

  • 1. Go to Reports → Changes → Create
  • 2. Select the same profile and gene type for both analyses
  • 3. The older report is "before", the newer is "after"
  • 4. Pinkk returns a structured comparison of what changed

Checks — research without a profile

Every other Pinkk feature needs a Profile first. A Check does not. Paste any free text — a domain, social handle, brand name, token, contract address, or suspicious offer — and get back cited, evidence-based findings.

Any subject

No profile setup needed. Just paste what you want to investigate.

Free recall

If Pinkk has already researched this subject, the answer is free — no credit charged.

Cited evidence

Every finding comes with source URLs and an observation date — not just AI opinion.

Example Check subjects

Private share links

Subscribers can generate a token-gated share link for any report. The recipient sees the full report — no login required. The token is the credential. Rate-limited to prevent abuse.

Example share link

https://trypinkk.com/share/eyJhbGciOiJIUzI1NiJ9.abc123...

Share links are subscriber-only. Manage them from the report view. Revoke at any time.

Report verification & embeddable badge

Anyone can verify a Pinkk report is genuine at /verify using the reference number. The verification page shows authenticity details only — profile identity, reference number, date — never the findings themselves.

Public verification (/verify)

Enter a PKK reference number. Confirms the report is genuine and when it was issued. Useful for partners and clients who receive a shared report and want to confirm it wasn't fabricated.

Embeddable SVG badge

Every verified report generates a "Pinkk Verified" SVG badge. Embed it on a media kit, proposal, or website to signal that your claims are backed by a structured intelligence report.

Knowledge check — Module 5

6

Module 6 of 7

Teams & Credits

What you'll learn

  • • How team accounts work — owner, members, shared resources
  • • Per-member monthly credit budgets
  • • How the credit wallet and purchase flow works
  • • Subscription plans and what's included
  • • Bonus codes and one-time credit packs

How teams work

A Team is a multi-user collaboration boundary. It's not a content concept — it's an account/org concept. One person owns the team; other members operate against the owner's data and credits.

Owner

Holds the subscription and credits. Sets member budgets. Sees all reports and profiles. Manages team settings.

Members

See a banner noting they're viewing the team's shared resources. Spend from the owner's credit pool (up to their monthly budget if set).

Shared data

Profiles, workspaces, and reports all resolve through the owner's data. Members see the owner's resources — not their own isolated set.

Per-member monthly budget

Owners can cap how many credits each member can spend from the shared pool per calendar month. Null = unlimited. The budget resets on the 1st and refunds from failed analyses free budget back automatically.

Example: 3-member team

Sarah (owner)No budget cap
Marcus (member)20 credits / month
Priya (member)10 credits / month

The credit wallet

Your wallet shows your current balance, full transaction history (grants, deductions, refunds, adjustments), and your purchase history. Credits never expire.

Wallet — example

Credit balance

47

Buy 10 credits — $10
Buy 25 credits — $20
Buy 100 credits — $50

Recent transactions

Subscription plans

Subscription plans are configured by the Pinkk team and shown on the pricing page. All plans include monthly credits, unlimited analyses (plan-dependent), and subscriber-only features. Check /pricing for current options.

What subscription unlocks

  • • Monthly credits included with each billing period
  • • Comparisons (multi-profile analysis)
  • • Private share links
  • • Batch autofill (fill all fields in one action)
  • • Custom analyst note on reports
  • • Priority queue for faster analysis
  • • Suggest-a-question on analysis form

Bonus codes

Pinkk occasionally releases bonus codes for promotional campaigns, partner deals, or events. A valid code grants credits on redemption. Codes may be limited by date, redemption cap, or user eligibility — check the terms on any code you receive.

Knowledge check — Module 6

7

Module 7 of 7

Quick Reference

Glossary

Platform coverage

The following platforms are supported for profile handles and cross-platform analysis.

Key URLs

REST API quick start

Pinkk exposes a REST API at https://trypinkk.com/api/v1. Authentication uses Bearer tokens (Laravel Sanctum). Manage tokens at /api-access.

# Get your analyses

GET /api/v1/analyses

Authorization: Bearer <your-token>

# Get a specific report's content

GET /api/v1/analyses/{id} /report

# Explore the full spec

GET /api/v1/openapi.json

What Pinkk is — and is not

Pinkk IS

  • ✓ A structured traffic intelligence platform
  • ✓ Conservative and evidence-based by design
  • ✓ A tool for better-informed campaign decisions
  • ✓ A platform that flags uncertainty honestly

Pinkk is NOT

  • ✗ A fraud detection tool — no fraud judgments ever
  • ✗ An identity verification service
  • ✗ An SEO audit tool
  • ✗ A vanity metrics dashboard

Training complete.

You've covered everything from influencer marketing fundamentals to running intelligence reports, advanced features, and team management. Time to put it into practice.

Ready to prove it? Sit the 3-question certification exam with a shareable certificate of completion.

Create free account See pricing →

Final assessment

Certification Exam