⚠️ Channel adjustments for 2C products
This skill defaults to dev / B2B channels.2C consumer goods / education / apps: The main battleground for customer acquisition is vertical communities + short videos + vertical KOLs, choose the first platform based on publicly available regional data (e.g., short video in Indonesia/Thailand has already overtaken Facebook). For KOLs, prioritize nano / micro verticals—the more followers, the lower the engagement rate; micro > mega offers better value for money. For the complete 2C channel database + public sources, see → gingiris-seo-geo/references/2c-adaptation.md。
KOL Outreach and Influencer Marketing Practical Handbook
🌍 Language / 语言: Chinese | Full SOP
Core Principles
Only real people, no zombies. Concentrate budget on a very small number of valuable key nodes, rather than spreading it thinly.
The cold start of a successful KOL collaboration chain depends on key nodes with high-quality "real-person followers": entrepreneurs, big-tech executives, technical experts, and investors.
Key Insight: A Newsletter founder's personal account is an extremely valuable ignition node (Manus: Rowan Cheung → 2.8M views; Devin: Zain Kahn → 1.4K reposts)
Full KOL Collaboration SOP (6 steps)
Step 1: Screening and List Building (L-4 weeks)
Tiering Strategy：
- Tier 1(ignition nodes): Newsletter founders, top technical experts, top VCs → free outreach via connections
- Tier 2(amplification): mid-tier KOLs (10K-200K followers) → paid collaboration
- Tier 3(long-tail coverage): micro creators → templated / value exchange
Screening Criteria：
| Dimension | Standard |
|---|---|
| Follower count | 10K-100K offers the best value for money |
| Engagement rate | Average views / followers > 10% |
| Content verticality | Highly relevant to the product field |
| Audience profile | Highly aligned with the product ICP |
Step 2: Price Inquiry and Outreach (L-4 weeks)
Pricing References by Platform：
| Platform | Collaboration Format | Reference Price |
|---|---|---|
| Twitter/X | 1 Thread | $50-200 |
| YouTube | Long video (feature) | $600-1,500 |
| TikTok | Single collaboration | $200-500 |
| Single post | $2,000-8,000 |
Negotiation Strategy: After the first quote, reduce it by 30%-50% on the grounds of "limited budget," and negotiate using added value such as "long-term collaboration + exclusive early access."
Step 3: Send Content Package (L-3 weeks)
The Content Package must include: Product Overview, Brand Assets, Key Talking Points, Publishing Requirements (including specific date and time), UTM Links (dedicated for each KOL).
See details → references/full-sop.md
Step 4: Review KOL Drafts (L-2 weeks)
Review Checklist: UTM link correct → @official account → includes Hashtag → brand video attached → core link position correct → publishing time consistent → content authentic and not ad-like
Step 5: Payment (L-1 day to L-Day)
Pay uniformly 1-2 days before content publication. Collect: Account name / Account number / SWIFT code / Bank name / Routing number (for US accounts).
Step 6: Publication Monitoring and Data Review (L-Day to L+1 week)
- Confirm publication on time → interact and activate immediately → request 24h/48h/72h data screenshots
- UTM tracking conversion → Calculating ROI:
(KOL带来收入 - 合作成本) / 合作成本
Real project record in Apple Notes: AFFiNE Product Hunt
Project note record from 2024-03-04: a 3-person operations team coordinated 30 Twitter/LinkedIn KOLs and continuous multi-channel publishing within 24 hours, and the report received 700,000+ total-network impressions and 1,000+ Product Hunt upvotes, and entered the 8th place of the month within 96 hours. This record verifies that "even a small team needs clear division of labor, asset packs, publishing cadence, and centralized monitoring"; it cannot prove each KOL's fixed impressions or vote output, nor can it replace UTM, signup, activation, and paid attribution.
Data evaluation system
| Metric | Evaluation standard |
|---|---|
| ROI | (Number of signups brought × LTV) / KOL cost |
| Channel CPM | Calculate using the median of the most recent 10 periods to avoid extreme values |
| Engagement rate | Randomly take 5 posts, add them up, and calculate the average |
| Conversion tracking | Dedicated UTM for each KOL → at least track to PV |
Publishing scale reference
- Goal: 200 influencers do Quote/Repost (100 AI General + 100 vertical industries)
- Roll out 30-40 Threads + 15 in minor languages
- Estimated total spend: ~$10,000
Toolchain
| Tool | Purpose |
|---|---|
| EezyCollab | KOL matching and emailing |
| Nanoinfluencer | KOL matching/competitor research |
| Growi | Influencer/UGC creator management |
| XAutoDM | Twitter DM automation |
| Phantombuster | LinkedIn DM tool |
| Bitly/Juicebox | Short links + UTM tracking |
Key algorithm adaptation points for each platform
| Platform | Most core metric | Key points of KOL Brief |
|---|---|---|
| TikTok | Completion rate/watch time | Strong Hook in the first 3 seconds, 15-60 seconds, Loop design |
| YouTube Shorts | Satisfaction Signals | The first 3 seconds decide life or death; content has informational value |
| Instagram Reels | DM share rate (3-5x the weight of likes) | Content that makes people "want to send it to friends" |
| X (Twitter) | Author reply (150x the weight of likes) | After posting a Thread, reply under popular comments |
| Comments (15x the weight of likes) | Controversial viewpoints that spark discussion | |
| Voting speed (first 1 hour) | Fit the community tone; soft promotion > hard sell |
Scenario binding and AI-powered outreach efficiency (external real-world evidence)
Citation: Superlinear Seven Wang Yue interview (Monica Going Global Talk, 2026-07). Superlinear = Lessie (~100k users' AI Agent for finding people) + DeepLink (influencer marketing service), serving 100+ going-global brands, $7-8 million USD ARR, paid influencer fulfillment rate 90%+.
Outreach and beta handoff
When a campaign also invites creators to test the product, use outreach-and-beta-handoff.md. Separate connection acceptance, replies, invitations, actual use and completed tests; compare channel performance from observed cohorts rather than universal reply-rate claims.
1. Content strategy anti-pattern: don't cram 10 selling points into a 15-second video
A short video of 15-60 seconds with 6-8 selling points crammed in = each selling point gets less than 4 seconds; users scrolling through videos in fragments remember nothing—this is the most common and most expensive mistake made by Chinese brands going global. The root cause is cognitive misalignment: caring only about "what I want to tell users," not "what users want to get." Today, with products proliferating and becoming homogeneous, straightforward feature promotion moves no one.
The right approach—scenario binding hook(A video should thoroughly explain only 1-2 selling points, customized by influencer vertical):
| Influencer vertical | Hook example |
|---|---|
| Workplace influencer | "The thing my boss asked me to do overtime, I finished in 5 minutes" |
| Photography creator | "One second fixes my bad shots" |
| Entertainment creator | "I can't edit video, AI edited great transitions for me" |
Choosing creators should work backward from the scenario: office products look for workplace creators, image/video generation products look for photography or entertainment creators. This mutually corroborates the Plaud case in this handbook (vertical scenario match > follower count)—one reaches the same conclusion from the hardware side, the other from the content side.
II. AI outreach pipeline: assistant, not substitute
- Efficiency numbers: one operator's monthly outreach volume goes from 20 → 80-100 creators (3-4x). In the past, finding one person meant paging through 20 pages of LinkedIn and verifying one by one; now one prompt gets a list in 1-2 minutes, and the human only does verification.
- Division-of-labor red line: sourcing, profile judgment, and writing the first personalized email are handed to AI;sending email is an irreversible action, so the decision must remain with a human—"there always has to be someone to take the blame"
- Feedback flywheel: when a list is rejected by the client, feed the reasons for rejection back into the prompt (what to exclude next time). Every adjustment feeds data, making it more accurate the more it is used.
III. How to achieve a 90%+ fulfillment rate (no mysticism)
- Paid + signed contract: creators earn money + contract constraints = fulfillment incentive; pure commission/barter has insufficient incentive. Overseas creators' schedule delays are common, but under paid contracts, unexplained drops are rare.
- Prepayment risk control: if you can avoid prepayment, pay after delivery; if prepayment is required, use PayPal (platform risk control; if not fulfilled, it can be recovered within the time limit).
- KPI boundaries: only commit to content delivery quality (creator's average level + brand approval), not to conversion/registration/payment rates—those are affected by landing pages, payment flows, and many other factors.
- Fronting payment model(agency perspective): front creator payments for clients and settle monthly/per campaign, provided your own cash flow is healthy + you screen client size.
IV. Market priority: Latin America is underestimated
North America and Japan/South Korea are the base → Latin America, Europe, Middle East → Southeast Asia/South Asia (poor paid conversion).Latin America, especially Brazil's sole-proprietor-style SMB merchants have "come out with pretty good data" on paid conversion for AI efficiency tools. This aligns with Gingiris's own testing: with a $5,000 KOL budget, Latin America got 12 conversions vs. the US's 3. For communication in minority languages, use AI translation as a base, and have someone with local cultural background check taboos.
Complete reference document
| Document | Content |
|---|---|
| references/full-sop.md | Complete KOL Operations Guide (18 chapters, including email templates, Content Package templates, top KOL resource library, case reviews, and in-depth analysis of platform algorithms) |
Hardware Product KOL Strategy
Source: compiled from Plaud official blog / Kickstarter page / Insta360 prospectus / media reports
I. Plaud's "mid-tier vertical matrix" strategy
Plaud's early KOL growth path overturned the "throw money at big influencers" approach commonly used by hardware brands:
Core principle: vertical scenario match > follower count
| Dimension | Traditional hardware KOL approach | Plaud's approach |
|---|---|---|
| KOL selection criteria | Follower count first (digital big influencers like MKBHD) | Vertical scenario match first (lawyers/doctors/students/podcasters) |
| Collaboration Method | Paid advertising + designated script | Free product samples in exchange for genuine review content |
| Early-stage Scale | Spend budget to cover 20-50 big influencers | Start with about 10-20 small and mid-sized creators |
| Content Nature | Product feature demonstrations | Workflow integration experience (real usage scenarios) |
Why bypass digital big influencers like MKBHD?
- The audience of digital big influencers is "tech enthusiasts," not Plaud's core users (knowledge workers)
- Lawyers filming Plaud recording court meetings, doctors reviewing AI medical record-taking, reached niche scenarios with real willingness to pay
- Vertical creators have stronger product perception, content is more authentic, and conversion rates are significantly better than general tech reviews
- Free samples in exchange for real content, extremely low cost, and word-of-mouth content remains effective long term (vs. paid ads disappearing once taken down)
Visual Hook Design Principles (Plaud Case):
| Hook Element | Specific Expression | Sharing Driver |
|---|---|---|
| Credit card thickness | Comparison image placed side by side with a credit card | Counterintuitive visual impact, "This thin?" |
| MagSafe attachment | Snap-and-use attached to the back of an iPhone | Natural scenario for seamless-use lifestyle UGC |
| AI output comparison | Before/After of 1-hour meeting → structured summary | Instant sense of functional value, strong conversion |
🇺🇸 Plaud Mid-Tier Vertical Matrix: Skip the tech mega-influencers. Find creators whose audience IS your ICP (lawyers, doctors, consultants, students). Send free devices, ask for honest reviews. 10-20 micro-creators with scene-matching audiences outperforms 1 MKBHD-type placement for hardware with a narrow professional use case.
II. Insta360 2000+ Ambassador System
Insta360 (Arashi Vision) has built a multi-tier creator collaboration system. Official public information shows a team of about 2,000 people globally supporting the creator ecosystem:
- Four-tier structure: Affiliate (5% commission, fully open) → Rising Talents → Campus Creator → Official Ambassador
- KOL selection logic: Prioritize "visual content creators" + "extreme sports/travel/outdoors" verticals, naturally matching strong hardware visual capabilities
- Lowering barrier strategy: Open Affiliate lets any creator join with zero friction, recruiting a large-scale KOL matrix of "income-driven" creators
- Top resources concentrated: Official Ambassadors enjoy co-branded releases, creator events, and custom equipment, which is the top-tier incentive
🇺🇸 Insta360 Ambassador Scale: Open bottom (Affiliate, any creator) drives volume; selective top (Official Ambassador, co-branded content) creates prestige. The visible career ladder keeps mid-tier creators invested in climbing, not just one-shot deals.
III. Natural Visual Hook Design Principles for Hardware KOLs
The effectiveness of KOL content for hardware products is determined by the product's own "visual hooks." The core of the KOL brief is to activate these innate advantages:
| Principle | Description | Case |
|---|---|---|
| Counterintuitive form | The product's physical form exceeds user expectations, creating "How did it do that?" | Plaud credit card thickness, DJI Mavic palm-sized after folding |
| Usage scenario visualization | How the product looks in real use is more shareable than static images | Insta360 NoseMode extreme perspective, Plaud MagSafe attached to iPhone |
| Quantified Before/After effect | Clearly show the difference before and after use, especially the output results of AI features | Plaud 1-hour meeting → AI summary screenshot, Insta360 normal perspective vs 360° playback |
| Imitability | After creators see the content, they want to try it themselves | NoseMode (just pick it up and shoot), Plaud meeting recording (everyday scenarios) |
Brief guiding principles:
- The Brief given to KOLs must include: 3 specific visual hook scene demonstrations
- Do not specify a script, but provide 'recommended shooting comparison combinations' (e.g., credit card comparison / AI output screenshots)
- Encourage KOLs to embed their own real workflows, rather than simulated use
- Best content lengths for hardware KOLs: YouTube (5-8 minute in-depth review) / TikTok (30-60 second visual impact) / Instagram (9:16 Reel, product reveal in the first 3 seconds)
🇺🇸 Hardware Visual Hook Framework: (1) Form-factor surprise (physical impossibility), (2) in-use scene visualization, (3) Before/After AI output. The best hardware KOL content doesn't need a script — it needs a creator who genuinely uses the product in their real workflow and a camera to capture what the hardware makes possible.
Hardware case data sources: Kickstarter official (Plaud KS: $1,108,181 / 7,564 backers; iZYREC: $208,713 / 2,391 backers) / Insta360 IPO prospectus / Forbes / media reports compiled
Frontline oral practical experience: price floor table / negotiation SOP / long-term outreach to top-tier creators
Source: Cyber Island podcast Vol 27 conversation (2026-04 oral account), self-negotiation pricing experience during the AFFiNE period. The following numbers are all self-reported, not platform rate cards.
1. Influencer price floor table (self-negotiated pricing, 2026-04 oral account, approximate values)
| Platform / Format | Scale | Floor price reference |
|---|---|---|
| Twitter/X thread | Under 100k followers, single post reach 10k+ | ≈$50–200/post |
| YouTube long-form video | Under 100k followers, about 10k views | ≈$800–1,200 (deals also closed at $300–500) |
| YouTube mid-roll | Same as above | ≈$500–800 |
| KOC | 3,000 followers | Lowest collaboration was $80 |
| US market YouTube | Same range | At most up to ≈$1,200–2,000; any higher = you didn't negotiate well enough |
2. Negotiation SOP: start at 30% of the quote + create a sense of internal competition
- Anchor: If an influencer quotes $1,000, start negotiating at 30% ($300).
- Internal competition script: Leverage influencers' herd mentality—"This collaboration has 20+ influencers, your friends have all posted."
- Real case(2024-02 first PH campaign): An influencer quoted too high and was left on read; after seeing peers all post, afraid of being left behind, they came back on their own and posted for free.
- Look at competitors when choosing people: First check which influencers competitors have worked with, and go directly for the same group.
- Verification: Cross-compare views + engagement + likes together, scan comment sections one by one to judge real fans—fake data is common, don't just look at follower count.
- Bypass agencies: Overseas influencers are mostly individuals / studios; direct outreach is faster, and the agency layer has pitfalls.
3. Outreach channel upgrade path (to prevent no-shows)
- Use email for first contact; after 2–3 back-and-forth emails, get WhatsApp / Telegram / LinkedIn to prevent delays and no-shows.
- Twitter allows direct DMs; for TikTok / YouTube, send an email and directly attach an IM QR code.
- Cold start scenario: on launch day, directly DM the Twitter followers of benchmark competitors (e.g., Notion mirror-type products)—one of the combo moves for AFFiNE's open-source day.
- Content package additions: Record a product onboarding video to teach creators how to use it + summarize the 1/2/3 selling points for them, but don't lock them into a rigid template—let creators produce content in the format they're best at (review / unboxing / tutorial).
- Mindset and volume benchmark (analogy from academic applications, not first-hand KOL data): 300 cold emails yield 20 offers (≈6.7%)—"being rejected is completely normal; getting a response is a stroke of luck."
4. Long-term outreach to top-tier KOLs (who don't take money)
- Top industry influencers make money through content / affiliate—paying them doesn't work. First understand each creator's business model (taking ads is just one of them).
- Approach:One email a month, "like writing a little essay to your homeroom teacher", reporting on new feature progress—they will read the emails, they just won't reply.
- Real result: after persisting for about six months, a top KOL in a productivity niche proactively notified us that "I made a video about you," entirely unpaid.
5. Goodwill flow screening: how to allocate follow-up investment
- First proactively extend goodwill (only give what you "can afford to give"), and observe whether the other party catches it and returns it.
- Relationships are divided into "flowing vs. frozen": frozen (only absorbing, never giving back) means stop investing.
- Manage in tiers by far / medium / close distance, and only increase budget and resources for KOLs who reciprocate.
6. Collaboration safety and real ROI
- In the contract, clearly specify ad disclosure, scope of material licensing, secondary editing, whitelisting placement duration, and takedown mechanism; do not clone voices, face-swap, or impersonate creators without written permission.
- Acceptance is not just about screenshots: record the publish link, publish time, UTM/promo code, and 24h/72h/7d impressions, clicks, activations, payments, and refunds.
- For quotes, view counts, and case conclusions, note the market, date, sample size, and whether they are self-reported. A single viral hit (there was once a quote post that approached 700,000 impressions) only shows the creative is reusable; it does not represent a stable average.
- Do not use the fabricated pressure tactic that "competitors have all already published"; you can truthfully explain the campaign scale, schedule, and budget boundaries.
🇺🇸 Field notes (AFFiNE-era, self-reported in a 2026-04 podcast): Price floors — Twitter thread (<100k followers, 10k+ impressions) ≈$50–200; YouTube dedicated video (<100k subs, ~10k views) ≈$800–1,200 with deals as low as $300–500; mid-roll ≈$500–800; cheapest KOC deal that shipped was $80 at 3k followers; US-market YouTube tops out around ≈$1,200–2,000 — anything above that usually means you didn't negotiate. Negotiation: counter at ~30% of the first quote and create peer momentum ("20+ creators are already in this campaign") — one over-priced creator eventually came back and posted for free. Escalate from email to WhatsApp/Telegram/LinkedIn after 2–3 replies to prevent no-shows. For top creators who monetize via content/affiliate rather than ads, money doesn't work: send a short monthly what's-new email — they read even if they never reply; a free dedicated video landed after ~6 months. Allocate follow-up budget only to creators who reciprocate goodwill; stop investing in one-way relationships.