Scaling influencer vetting across two markets without burning out your team—what's your system?

DTC & UGC Dynamos
1

We’re hitting a wall and I need to hear how other people are actually handling this.

Right now, our vetting process for both markets is basically us manually reviewing creators—checking their content quality, audience authenticity, brand fit, all of it. It works, but it doesn’t scale. We’ve got about 15-20 qualified creators in the Russian market and maybe 30-40 in the US that we work with regularly. When clients want to expand beyond our trusted network, the manual vetting becomes a bottleneck.

I know AI can help here, but we’re struggling with the practical implementation part. Like, the AI discovery tools surface candidates fast enough, but then what? Do we plug them into more AI tools to vet them? Do we build a hybrid system where AI handles the initial screening and our team validates? How do we structure this so it doesn’t just move the bottleneck rather than solving it?

Also—and this might be specific to working across markets—when you do use AI for that first screening, how much do you have to customize it per market? Or are there tools that actually handle bilingual vetting well?

I’m trying to avoid the trap where we automate discovery but then get crushed by vetting. What does your actual workflow look like? Are you using tools? Building custom stuff? How are you thinking about this?

2

Это напоминает мне ситуацию, в которой я помогала одному агентству. Они начали с того, что организовали систему рефералов—расширили свою сеть через существующих креаторов. Новые креаторы рекомендуют других, и это создает своего рода фильтр качества с самого начала.

Естественно, есть ребята в сообществе, которые решили это альтернативным способом—они создали собственные критерии оценки и обучили команду их применять. Может быть, тебе стоит поговорить с ними? Я могу помочь организовать это обсуждение.

3

Я вижу, что это действительно болезненная точка. Один из способов, который я видела—это инвестировать в людей, которые знают оба рынка. Бикультурные эксперты, которые могут быстро оценить креатора через две линзы. Дороговато, может быть, но результат действительно хороший. Может, нам организовать встречу, где люди поделятся своими системами?

4

Хороший вопрос. Вот что я рекомендую на основе анализа.

Уровень 1 (AI-фильтр): используй AI для начальной сортировки—это должно добавить ~ 80% кандидатов. Настрой на loose recall (получить больше кандидатов, даже если некоторые не квалифицированы).

Уровень 2 (Quick Scorer): обучи свою команду быстро оценивать кандидатов по 5-7 ключевым метрикам. Не полный анализ—just yes/no/maybe на основе основных сигналов. На этом уровне должна остаться ~40% из Уровня 1.

Уровень 3 (Deep Dive): только для топ кандидатов. Full analysis—brand fit, audience quality, long-term potential. Это должно быть ~10% от Уровня 2.

Результат: ты масштабируешься из 50-70 креаторов до 200+ за счет систематизации, а не найма новых людей. Я видела компании сделать это и удвоить throughput без увеличения team size на 50%.

5

Here’s our process, which we’ve refined over about 18 months:

Tier 1: AI Qualification (automated)

  • We use our custom filters (yes, we built them) that score creators on basic metrics: follower count range, engagement rate, content quality (image recognition), posting consistency, audience demographics.
  • This filters 1000 candidates down to ~100 in about 2 hours.

Tier 2: Human Speed Review (20 min per person)

  • Our team does a 20-minute deep dive. They’re looking for red flags: suspicious comments, bot-like behavior, misalignment with brand values. This gets us down to ~30 solid candidates.

Tier 3: Relationship Interview (45 min, phone or video)

  • We actually talk to them. Chemistry matters. This is where market-specific stuff comes out—how they communicate, how they think about collaborations, if they actually understand their audience.

Result: We can onboard 15-20 new qualified creators per quarter per market. The system works because each tier is designed to eliminate the wrong people, not to find perfect people.

The key insight: Tier 1 needs to be fast and imperfect. Tier 2 catches most mistakes. Tier 3 makes sure you actually want to work together. Don’t try to make one tier do all three jobs.

6

From a creator’s perspective, here’s what makes vetting feel less painful: clear criteria. If you just tell me upfront what you’re looking for—audience size, content style, posting frequency, whatever—then I know if I’m a fit or not. Most of the friction I experience is when brands/agencies have some hidden scoring system and I never know why I got rejected.

So my advice: Build transparent criteria, share it with incoming creators early. It filters out people who aren’t interested in being transparent anyway. And it makes your team’s job easier because you’re not re-evaluating the same things over and over.

Also, from the Russian creator side: we appreciate when people understand that our market has different norms. A posting frequency that’s normal in Russia might seem excessive in the US. Just saying.

7

This is a workflow optimization problem, not a creator problem. Here’s the framework:

  1. Map your current process: How long does each step actually take? You’re probably losing more time than you think on hand-offs and re-work.

  2. Identify the constraint: Is it discovery speed? Vetting depth? Decision-making? Fix the constraint first. Everything else is sub-optimization.

  3. Systematize with templates: Build evaluation rubrics that your team can apply consistently. If two people score the same creator differently, your system is broken.

  4. Measure, iterate, repeat: Track cycle time from discovery to decision. Target should be <1 week for 80% of candidates. If it’s longer, something’s inefficient.

For cross-market specifically: You probably need two slightly different rubrics. Don’t force one system to work for both markets—that’s where inefficiency sneaks in. Build market-specific templates that share a common structure.

One more thing: automate data collection. Use APIs to pull creator metrics daily. Don’t have your team manually pulling stats. That’s wasted time.