Street Traffic | Nutra Case Study: 103% ROI on Prosteron in Algeria from a Single Ad Account
If your nutra creative lasts a week, that's normal. Two weeks — already good. A month — you're doing great.
We had one video for Prosteron in Algeria running on Google Ads for 2.5 months. It closed at 103% ROI from a single ad account.
For 5 years we've been running nutra quietly. No chat appearances, no stage time, no published cases. We've decided to change that — and start showing, through our own examples, what results are achievable in this market when a team has built solid internal infrastructure, people trust each other, and buyers are given the freedom to test: want to try a new bundle, angle, GEO, or format — go ahead, no one sets limits. This case is the first.
Numbers
| Metric | Value |
| GEO | Algeria |
| Offer | Prosteron (prostatitis) |
| Source | Google Ads |
| Network | Public CPA |
| Payout | $16–17 |
| ROI | 103.10% |
| Leads | 14,274 |
| Sales | 3,666 |
| Approval rate | 25.68% |
| Trash | 7% |
| CPA fact | $2.14 |
| Landing CR | 0.85% |
Tracker screenshot for the entire campaign period
Essentially, we doubled the budget. Not x5, not x10, the way people love to draw it up in chats. Just a steady x2 that ran for a long time, predictably.
Google Ads account screenshot (account identifiers blurred)
Under the Hood
Breaking it down honestly, because there's zero magic here:
- Accounts: in-house farm
- Proxies: mobile
- Cloaking: proprietary
- Tracker: Keitaro
- Landing pages: took a reference from spy tools and reworked it for the GEO
A separate note on landing pages, because this is where people often screw up. We didn't just upload them as-is. We took a structure that was already working for competitors, then rewrote the header, changed the visuals, and added a localized expert angle. That's the baseline — without it, Algeria simply doesn't convert.
Landing page header: styled as a local health portal, with an "expert" and native presentation
Price block: classic "old price / new price" mechanic in local currency DZD
The Creative That Ran for 2.5 Months
This one surprised us too. Usually a video on Google lasts a week or two, a month at best. Ours ran for 2.5 months without dropping.
What was in it? Everything at once:
- Before/after
- Expert presentation (Doctor-style format)
- UGC segment / social proof (persona-based segment)
- Attention-grabbing intro
All of it stitched into one video. Sounds like a mess, but that exact combination is what gave it longevity. The viewer is walked through a chain: sees the problem → sees the expert → sees a regular person with results → sees the reaction. Each stage catches a different type of user, and CTR doesn't die off.
A separate point on language — this matters. A lot of people think "Algeria is Arabic-speaking, so you run in Arabic." That's a mistake most beginners make.
In Algeria, French is what actually works. Yes, the official language is Arabic (more precisely — the Algerian dialect "Darija," which locals use in everyday life). But French has remained the language of education, business, press, and medicine since the colonial era. When an Algerian reads serious health content, they expect to see it in French — it automatically builds trust and reads as authoritative.
That's why our landing pages and creatives were in French. Translations were done via AI with thorough human review. No cultural or religious missteps — because French is neutral by nature, without the religious undertones that need to be carefully navigated.
If you want to enter Algeria — drop the "I'll run in Arabic like every other Arabic GEO" mindset. This market is different.
Why a 0.85% Landing CR Is Fine Here
Someone will look at this and say "your CR is pennies." Fair enough — let's look at the full picture:
- Algeria runs on DZD, low purchasing power
- Prostatitis isn't an impulse buy — men take their time
- But CPA fact is $2.14, approval rate 25.68%, trash 7%
At a payout of $16–17 with that math, you get +103% ROI. The low CR is offset by a cheap lead and a solid approval rate. In nutra, looking at a single metric in isolation is pointless — calculate the full unit economics.
Seasonality in Arabic GEOs
Ramadan and summer dipped. That's a fact you need to accept upfront when working Arabic GEOs. Conversion drops, approval rate fluctuates, the audience shifts into a different mode. But once the seasonal factors pass, everything recovers quickly.
The main advice here: don't kill a bundle just because it dipped for a week or two. If it was working before Ramadan, it'll work after. Just switch into test mode — try new angles, formats, and offers, find what works specifically during that period. Stopping entirely is never the right call.
What We Learned About the Algerian Audience
A few things that only became obvious to us in the process:
Polished creatives don't work. The more beautiful and "professional" the video looks, the more it looks like an ad. The less trust it gets. A cheap camera, a home setting, an "ordinary person" — that lands far better.
The expert character in the creative needs to look like someone people trust. In most Arabic GEOs, a local archetype outperforms a European one. But Algeria is an exception. Thanks to the French legacy, a European expert-style image is perceived here as a mark of expertise, not as "foreign." In our creative, that approach delivered better conversion. The same angle wouldn't work in another Arabic country — Saudi Arabia, UAE, Egypt.
Trust isn't a "bonus" — it's the foundation. Comments, reviews, native mentions in the landing page copy. Without this, Algerians don't buy. Full stop.
Social proof needs to be adapted, not copied. If you pulled comments from a spy tool and ran them through Google Translate, the landing is dead. You need to rewrite them manually (or with a solid prompt) to match the local language and tone.
Problems? Honestly, There Were None
No mass bans, no approval rate drops, no sudden conversion dips. And not because we got lucky. It's because problems don't arise when the bundle is built correctly from the start.
In-house farm + mobile proxies + proprietary cloaking = one account handled the entire volume without a ban. Localized landing pages + proper translation = clean approval rate. It all follows logically.
And it's not a coincidence. Over 5 years, we've built internal infrastructure where every process is dialed in: from account farming to cloaking, from creative to analytics. When everything inside the team runs like clockwork and people trust each other, you're not spending resources on fires — you're spending them on running traffic.
What We Would Have Done Differently
Honest answer: we would have pushed a bigger budget. When we saw the bundle running steadily into its second and third week at 100%+ ROI, we should have scaled more aggressively. We played it safe, and it cost us additional volume.
It's the classic cautious buyer mistake: you wait for one more confirmation, then one more, and while you're waiting, the offer starts to fade.
Tips for Anyone Looking to Enter Arabic GEOs with Nutra
No fluff, straight to the point:
- Don't cut corners on localization. This is not the line item to save money on. Localization is half the success in Arabic GEOs
- Test multiple creatives in parallel. Even if one hits, don't get comfortable. A creative that ran for 2.5 months is an exception, not the rule. Always have 3–5 fresh videos in testing
- Don't chase the cheapest lead. A $0.50 lead with a 10% approval rate = minus. A $2 lead with a 25% approval rate = plus. Look at CPA fact, not CPL
- Plan for seasonality. Ramadan, summer, religious holidays — these aren't reasons to kill a bundle. They're reasons to test new approaches, formats, and angles, not to stop
Summary
- ROI 103.10%, approval rate 25.68%, trash 7%
- 14,274 leads, 3,666 sales
- Algeria + Prosteron + Google Ads + one properly built video + one ad account
In Arabic GEOs, the winner isn't the one who ran the most traffic — it's the one who didn't cut corners on the details. Localization, translation, landing pages, the face in the creative — that's what separates +100% ROI from a blown budget.
About Us
Street Traffic — a team that's been running nutra for 5 years. All that time we worked quietly, without public cases or self-promotional posts. Now we've decided to step out of the shadows.
Not because we want hype. But because we see how the market is changing: it's getting harder for newcomers to get in, experienced buyers are moving into team lead roles, yet examples of systematic work in the public space are almost nonexistent. Everyone is either burning bundles six months too early, or not showing anything real at all.
We're going to show it. Not life hacks and not "ran traffic in 24 hours." But what the work of a team actually looks like — one that has invested years into internal infrastructure: farming, cloaking, anti-detect, landing pages, translation, analytics — and that is built on internal trust. When everyone is in their place and no one is pulling the blanket their way, the numbers grow on their own.
This Prosteron case is the first. More are coming.
🔥 We're Hiring Buyers
Stepping out of the shadows, we're opening recruitment. We're scaling and looking for the people we'll build the next cases with.
Who we're looking for:
- TL Media Buying Facebook Nutra
- Media Buyer Native Nutra
- Media Buyer Google Nutra
Driven and hardworking people — men and women — with non-standard thinking and an analytical mindset. Those who don't stop at the first creative that works, but constantly look for new approaches. People who can tell the difference between luck and a system, and build the latter rather than rely on the former.
What we offer:
- Ready-made infrastructure: in-house account farm, anti-detect, mobile proxies, proprietary cloaking, Keitaro
- In-house landing pages and pre-landers tailored to GEO, localized into the required language
- Working offers from top networks with no payout shaving
- Freedom to test — want to test a new bundle, angle, GEO, or format, no limits set
- Fair split, transparent numbers, no "surprises" in the stats
- A team where issues get resolved, not left hanging
