How Smilodox went from tracking to the Adverfly AI... 🧠
Smilodox2026- Customer
- Industry
- Fitness apparel · D2C
- Powered by
We have been working together since 2023, since day one. And every day another step forward.
Uwe Schüder(aka Flying Uwe) · Founder, Smilodox & NeosuppsThis is the shop. 👕
Hundreds of styles, a sale running somewhere at all times. Every collection launch, every discount code and every creator posting about it lands in the same shop, and has to be told apart afterwards.
And if the next session deserves something better than a washed-out shirt, this way shop Smilodox
Many channels. Even more confusion. 🤯
At this size every channel has its own report, and every report says it won. What was missing was a system connecting all of them, although they all work together.
- Platform reports contradicted each otherEvery channel had its own numbers. They just refused to add up to the same story.
- New and returning customers in one potThe platforms optimised on people who buy anyway.
- Creators counted next to the numbersThe creators ran in a tool of their own, so they could never be compared with paid.
- Reporting was handworkEvery week the same numbers pulled together by hand.
Help is on the way 🚒
First a system was set up that connects everything. Then the creators came in with their valuable data, and on top of that came the Adverfly AI 🧠
Every app runs on the same numbers, so each new one is worth more than the last.
Tracking
Measured in the shop
- Adverfly pixel: server-side in the shop itself.
- New, returning and creator kept apart.
- Health Center watches the data quality.
Content
Creators in the same model
- Link and code per creator, checked against each other, so a sale never counts twice.
- Creator revenue competes for budget on the same numbers as paid.
- What works as content becomes the next ad. Nobody has to book a shoot for it.
And now it runs by itself. 🧠
The result: the Smilodox team has far more time for other things, because Adverfly takes tons of work off their hands and serves it up on a silver platter 🍽️ The competition is still doing all of it by hand 🔨
Not a snapshot of one good month, but the setup that keeps running. Every figure comes from the Smilodox workspace, September 2026. Neosupps, the same company's supplements brand, has a study of its own.
One set of numbers instead of five reports
Meta, TikTok, Google, creators and organic are measured by one yardstick. Every second purchase runs over more than one of those channels. That is exactly what the last click throws away.
One in five tested ads moves up a level
Around seventy new ads start each week. About half get a real test budget, and one in five of those passes €1,000 in spend. The top five percent by budget carry almost half the spend. Now the team can look at what is inside those creatives, too.
Creators and ads in one model
Creators and ads were put in one model because they feed each other: two out of three purchases on a creator code had also clicked a paid ad in the month before. Counted separately, both channels would have claimed the same sale; in one model it counts once.
The AI reads the material
Every video and every image gets read (hook, format, message) before anyone rates it, so performance can be read by content and not by ad ID. That is where the AI part starts.
The AI proposes, they approve. 🤖
Adverfly's AI infrastructure is ready 🧠 Is your brand? 🚀
Next up: budget proposals out of their own numbers. Check, approve, let the platform do it, the decision stays with them and their agencies.
That is what Adverfly's AI infrastructure is for: all the apps on one data foundation.
72 apps on one data foundation · 57 integrations
Capability by capability — what's integrated here, and what you'd otherwise buy separately.
Capability
Adverfly
All in one platform
3,600 €/ month · 1 invoice
All apps work togetherOther tools
Separate & additional cost
~20,900 €/ month · 18 tools + 1 role
Each tool works on its ownBuilt-in
Mixpanel, Hotjar · separate
Built-in
Google Analytics, Plausible
Built-in
Funnel.io, Supermetrics
Built-in
Native per tool · separate
Built-in
Sprout Social, Buffer
Built-in
Ahrefs, Semrush
Built-in
Typeform, SurveyMonkey
Built-in
Trustpilot, Yotpo Reviews · separate
Built-in
Brand24, Mentionlytics
Built-in
Manual integration into BI
Built-in
Notion + manual sync
Built-in
Klaviyo Lists, RudderStack
Built-in
Manual tagging in DAM
Built-in
Manual analysis in spreadsheets
Built-in
Manual creative reviews
Built-in
Looker, Tableau · per seat
Built-in
Motion, Atria · separate
Built-in
Heap, Amplitude
Built-in
Manual cohort analysis in BI
Built-in
Custom dashboards
Built-in
Looker, Tableau · per seat
Built-in
Hotjar, Microsoft Clarity
Built-in
FullStory, LogRocket
Built-in
Manual P&L in Excel + BI
Built-in
Triple Whale, Northbeam
Built-in
Consulting
Built-in
Custom statistician
Built-in
Manual stats in R/Python
Built-in
ChatGPT, Claude · manual prompting
Built-in
Custom-built chatbot per workspace
Built-in
Excel + consultant
Built-in
Yotpo, Smile.io
Built-in
Privy, Justuno
Built-in
Midjourney, Runway
Built-in
Stape, Funnel
Built-in
Zapier, Make
Built-in
Build it yourself · eng team needed
Built-in
Modash, Aspire
Built-in
Aspire, GRIN
Built-in
WordPress, Webflow + manual tracking integration
Built-in
Frontify, Brandfolder
Built-in
Custom alerting + event tracking
Built-in
Funnel, Supermetrics
Built-in
Manual in spreadsheets
Total / month
3,600 €saves 17,300 €
~20,900 €
Illustrative for one brand at 10 Mio. € annual revenue. Adverfly: Managed Infrastructure, billed yearly — flat price, no revenue share. The comparison uses list prices of the tiers such a brand lands on, or typical retainers, rounded, plus half an engineering role for pipelines and upkeep — that work is priced in on the Adverfly side. Vendors named in the rows illustrate the capability; the total is not a price quote for any single product. As of September 2026.
What actually changed. ✨
| Area | Before | After | With |
|---|---|---|---|
| Measurement | Every platform reported its own sales. | Measured server-side in the shop itself. | |
| Attribution | The ad platforms claimed 93% of every order in the shop (2024). | One model every channel is measured by. | |
| New customers | New and returning ran in one number. | Measured separately and fed back server-side. | |
| Creators | Tracked in a separate tool, next to the real numbers. | A link and a code each, measured like paid. | |
| Decided on | Revenue, which says nothing about what was left over. | Profit, after returns, cancellations and cost of goods. | |
| Creatives | An ad was judged by its name and its spend. | The AI reads what is in it, so content becomes comparable. |
Become the market leader in your niche, on Adverfly's AI infrastructure. 🚀
In 30 minutes we show you what is hiding in your data: on your channels, not on a demo account.
Read next
AYLASHESOne of the first brands to bet on real omni-channel tracking instead of platform self-reporting, with creators measured the same way as every paid channel.
Blume IdealIn the first month, Blume Ideal found €40,000 in Meta spend with no revenue showing up in its shop measurement. That look behind the platform reports is why Adverfly stayed.
Medizen DigitalMedizen sells course places, not products, on a platform it built itself. An off-the-shelf shop plugin was not going to do the job. So the tracking was tailored to their platform, rather than working around it.
NeosuppsSupplements are bought again. Neosupps counts a customer over twelve months instead of an order at checkout, and steers the ads on that.
glouragloura tracked its creators with discount codes alone. The numbers were wrong in both directions until link and code were counted together.
Written by Ezhan Atak · Founder, Adverfly
Published: Updated: