Struggling to verify ROI from influencer partnerships? here's how we built a simple tracking system

Кейсы и истории успеха
1

I’ve been frustrated with a problem for the longest time, and I suspect some of you deal with this too: we run influencer campaigns, we see engagement metrics that look great, but then when it comes time to actually prove ROI to leadership, everything gets murky.

We’d have conversations like:

  • “Did that influencer partnership actually drive sales, or just awareness?”
  • “How do we compare a $5k influencer campaign to a $5k paid ad campaign?”
  • “Which influencers actually work for our brand long-term?”

And honestly, we had no good answers. We were relying on influencers to self-report their traffic numbers, or we’d look at swipe-up rates (if we had them), but it always felt incomplete.

Here’s what we did:

We started with a simple principle: every influencer partnership needs its own tracking mechanism. We created UTM parameters for each influencer (utm_source=influencer, utm_medium=partnership, utm_campaign=influencer_name). Made it mandatory.

Then we built a spreadsheet—yeah, just a spreadsheet—where we tracked:

  1. Influencer name
  2. Engagement rate on their post
  3. Influencer-specific UTM clicks
  4. Actual conversions from those clicks
  5. Customer acquisition cost (CAC) from that specific partnership
  6. Customer lifetime value (LTV) of customers acquired from that influencer
  7. Notes on the collaboration (timeline, content type, audience overlap)

Then we sat down with people who actually knew the influencer space—not just our analytics team, but also people who manage relationships with creators—and we asked: “Looking at this data, what patterns do you see? What would you do differently next time?”

The breakthrough: We discovered that some of our highest-engagement influencers weren’t driving conversions because their audiences didn’t overlap with our target market. Meanwhile, some mid-tier creators we almost overlooked had audiences perfectly aligned, and their CAC was lower than our paid ads.

We also realized that follower count was almost useless as a predictor. A 50k-follower creator with audience alignment outperformed a 500k follower influencer with lower relevance every single time.

The awkward part: Having this data meant we had to have harder conversations. Some influencer relationships we thought were working… weren’t. We had to make changes based on data instead of gut feel.

But here’s where I’m still stuck: this system works, but it’s manual and time-consuming. We’re manually pulling UTM data every week, updating the spreadsheet, trying to connect backend sales data to specific influencers. And communicating these results to influencers themselves is tricky—how do you tell someone “your engagement looked great, but your audience didn’t actually buy” in a way that’s constructive and keeps the relationship good?

How are you all handling attribution and ROI verification for influencer campaigns? Are there tools or methods you’ve found that actually work without driving you crazy? And more importantly, how do you communicate performance data back to influencers in a way that strengthens partnerships instead of damaging them?

2

Отлично, что вы тянете за ниточку атрибуции. Это действительно одна из самых запущенных областей в маркетинге инфлюенсеров. Большинство компаний лежат на диване, погружаясь в тщетность, и просто верят «так же, как раньше».

Пару комментариев на основе моего опыта:

  1. UTM параметры—хорошо, но неполно. UTM не ловит людей, которые видят пост в инстаграме, но потом заходят на сайт через поиск или прямую ссылку через неделю. Вам нужна более тонкая граница. Я рекомендую добавить код скидки, уникальный для каждого инфлюенсера. Так вы поймете не только кто кликнул, но кто реально купил.

  2. LTV вычисляете как? Это критична метрика. Если вы берете просто первый покупку, а потом не отслеживаете repeat purchase, вы упускаете половину пазла. Клиенты, привлеченные микро-инфлюенсером с высокой доверотворностью, часто возвращаются чаще.

  3. Про САЦ и сравнение с пейдом: есть еще скрытая переменная—бренд-лифт. Даже если инфлюенсер не дал конверсию, он может дать осведомленность, которая потом конвертится через пейд канал. Это сложно измерить, но стоит помнить.

По инструментам: ми используем комбинацию Shopify Analytics + Google Analytics 4 с наложением данных о публикациях инфлюенсеров. GA4 лучше отслеживает customer journey, чем универсальная Google Analytics.

3

Я вижу вашу точку зрения по атрибуции, но я хочу добавить еще одну сторону медали—отношения с инфлюенсерами.

Мои лучшие долгосрочные партнерства с создателями построены именно на открытости и взаимном понимании. Когда я делюсь данными с инфлюенсером и говорю «вот что получилось, давайте вместе поймем почему», это укрепляет доверие, а не разрушает его.

Проблема в том, как вы framing это. Если вы просто показываете «твоя конверсия ниже чем нужна», это звучит как критика. Но если вы говорите «твоя аудитория супер engaged, но они немного отличаются от нашей целевой demographic, давайте вместе поиграем с контентом чтобы лучше подойти»—вот это совсем другая история.

Секрет: пригласите топ-инфлюенсеров (даже просто на кофе виртуальный) и покажите им общую картину. Не секреты всей программы, но общие инсайты. Спросите их мнение. Лучшие инфлюенсеры сами заинтересованы в том чтобы их контент конвертился, потому что это значит они могут повышать ставки в будущих сделках.

4

Okay, real talk from a creator side: please share your ROI data with influencers. But do it right.

I’ve worked with brands that literally never told me what happened after I posted. Like, I’d create content, they’d go radio silent, and I’d assume it flopped. Then I moved on. But later I found out the content actually did drive sales—they just didn’t communicate it back to me.

When a brand shows me: “Your post brought 200 clicks, 15 conversions, and these customers have 35% repeat purchase rate,” suddenly I understand why you want to work with me again, and I’m way more motivated to improve next time. Plus I can use that as social proof when I pitch to other brands.

The tricky part you mentioned—telling someone their engagement didn’t convert—that’s less painful if you frame it as learning together. Like: “Your audience loved the content (high engagement!), but they’re maybe not ready to buy yet. Let’s try a different product angle next time.”

Also: code-based discounts are your friend. Every influencer I work with gets a unique code, and I can see exactly what my audience bought. It takes 2 seconds for me to share that back with the brand. Way easier than trying to decode UTM parameters.

5

Спасибо, что делитесь этим. У нас была похожая проблема когда мы запустили первую кампанию с инфлюенсерами. Мы потратили немалые деньги, видели лайки, но не было ясности по продажам.

Вопрос: когда вы поняли, что някоторые инфлюенсеры не работают, как вы сделали переход? Просто прекратили сотрудничество? Или вы предложили им изменить подход? Интересует твой experience с трудными разговорами.

6

Solid methodology. UTM + discount codes + LTV tracking—that’s the right foundational stack. Here’s what I’d add to make it scalable:

First, automate the UTM naming convention. Make it so systematic that even a junior team member can’t mess it up. We use: utm_source=influencer | utm_medium=partnership | utm_campaign=[influencer_handle] | utm_content=[post_type]. Every time.

Second, stop updating a spreadsheet every week. Use a BI tool (Tableau, Looker, even Google Data Studio). Pull from Shopify and GA directly, set up a dashboard. Takes one afternoon to build, saves you hours every month.

Third, and this is important: differentiate between micro and macro influencers in your reporting. Micro-influencers (10k-100k) almost always outperform on CAC and repeat purchase because their audiences are tighter. Macro influencers drive awareness. Different KPIs, different expectations. If you’re judging both on conversion, you’re comparing apples to oranges.

On the relationship side—yes, share data with influencers. But frame it right: “Here’s what worked, let’s do more of it. Here’s what didn’t, let’s adjust.” The ones who respond well to data are the partners you want. The ones who get defensive? They’re probably not scalable anyway.

How are you segmenting by audience overlap? That seems to be your biggest insight—and it’s the thing most brands miss.

7

You’ve identified a real pain point, and your tracking framework is a good start, but I want to push on the methodology itself.

First, on attribution: single-touch attribution (crediting one influencer for a conversion) is increasingly flawed in multi-channel marketing. A customer sees an influencer post, clicks, leaves, comes back via Google search, converts. Who gets credit? If you attribute 100% to the influencer, you’re overstating their impact. You need multi-touch attribution or at minimum, last-click + assisted conversions.

Second, you’re measuring at the transaction level, which is good, but you’re missing the business architecture layer. Two questions:

  1. What’s your customer acquisition cost across all channels? Where does influencer CAC fall relative to paid social, organic, email, referral?
  2. What’s the incremental LTV lift? Like, if a customer came from an influencer, do they have genuinely higher LTV, or is that attributable to selection bias (maybe your best customers just happen to follow influencers)?

Third, on communication: influencers are not the same as paid media channels. They’re partners. But they also have their own economics. If you tell a 100k-follower creator “your CAC is too high,” they may simply not work with you again. Smart influencers know their per-post economics and will self-select out if ROI doesn’t work. So focus on: which tier of influencers (by follower size, engagement rate, niche) actually makes unit economics work before you onboard them.

One more strategic question: are you using influencers for bottom-funnel conversion, or top-funnel awareness? Because the ROI model is completely different. If they’re conversion-focused, CAC matters. If they’re awareness-focused, brand lift and downstream effects matter more. Are you clear on what you’re optimizing for per influencer, or are you treating them all the same?