Luxury lingerie leads at €4.26 across Europe
Project overview
Luxury lingerie under the MissTease brand.
Average order value: €108 (≈ RUB 10,762.85 at the FX rate at the time of the case — included here as a reference for a Russian-speaking team and reporting).
Starting point
They have an active Instagram* account with 200K followers, but it’s run with a focus on Russian-speaking audiences. They also have several websites — one for Russian-speaking customers and one for Europe.
*Let’s agree (here and throughout the case) that Meta is classified as an “extremist organisation” in the Russian Federation (this is a legal status specific to Russia). The client operates through a European legal entity, so they are not violating their local laws.
Goal
Establish a position in the European market. The goal: customers from Europe.
The unit economics model looked like this:
Cost per click (CPC): €0.14 ($0.15 / RUB 13.97)
Cost per lead (CPL): €3.50 ($3.78 / RUB 349.31)
Customer acquisition cost (CAC): around €50 ($53.98 / RUB 4,990.45)
Timeline and budget
Budget: ~€500/month (~$540.80 or ~RUB 50,014.07 at the FX rate at the time). At the start, we didn’t set a fixed end date — we’ve been working with the client for six months (since June 21).
About us
We are The Quite Orbit. A full-stack digital marketing agency.
We work on both performance marketing and marketing strategy.
At this point, our experience includes:
Total ad spend managed so far: RUB 3,681,853.58 (we keep RUB here because it’s how the team historically tracks total spend; for non-RU readers, treat it as “overall spend to date,” currency is secondary).
Prep / Kick-off
Initially, we decided to run Meta ads and drive traffic to the website. A nuance appeared immediately: the client wanted ads to feature models — not flat product shots of lingerie.
That’s a problem. The models in this brand’s lingerie look extremely sexy, and “puritan” moderation rules in Meta (and even VK, a Russian social network) can be hard to pass with this kind of creative.
We entered Meta Ads expecting a moderation fight — but the first problems came from a different direction:
- First: we had to re-build the entire advertising setup — Business Manager + Business Page + Ad Account — to operate in euros, with local settings outside Russia, under a European legal entity.
Spoiler: you don’t need anti-detect browsers. You just need to follow the rules: Ad Account location not set to Russia, a payment method that is not issued in Russia, billing currency not RUB (here: EUR). We also had to deactivate all Ad Accounts with Russia as the location and RUB billing. This can be done via Business Manager or Meta Ads Manager (we recommend Meta Ads Manager for this).
- Second: a card issued by a European payment provider did not support recurring charges. The bank treated Meta charges as a subscription and didn’t allow them.
We spent about two months fighting this with support:
By the end of August, we decided to move to Google Ads until the core Meta issues were resolved.
From August 25, we tested Google Ads and Google Analytics. We built transparent funnel analytics from visit to form-fill on the payment flow (not fully end-to-end at the final point at that time — we addressed that later), so we could see movement and conversion rates between funnel stages.
At first, we tested creatives across two countries together, then split them into separate campaigns to compare. Most purchases came from Germany.
When we moved to test other countries that were convenient for delivery, we returned to Meta.
We re-entered Meta in early November 2023 and still use it: we fight moderation and search for creatives that convert like model-based content but don’t get rejected as adult content.
Audiences we tested
We started from the audience description the client gave us:
By GEO:
Germany and France, as the wealthiest.
But in terms of demand, Eastern Europe could be stronger — and there is a shared customs zone inside the EU.
Who can generate revenue and pay back faster? We expanded on the client’s ideas.
1. Mostly “sugar babies,” “escorts,” “sex workers,” “girls with an OnlyFans account” — people who shoot photos and videos. In some European countries, parts of this are legal (local laws vary).
- Women with a strong aesthetic/fetish for beautiful lingerie — they invest in their appearance.
- In the client’s view, people who enjoy sex (broadly speaking).
- Women with larger sizes are highly likely to buy.
Europe can behave differently.
Tastes differ: a German woman aged 20–30 is not the same as a French woman aged 20–30 — and that differs again from Dutch, Estonian, or Lithuanian audiences.
Mostly around 30, starting from 25.
But possibly from 19–20 as well.
Psychographics: financially comfortable men buying lingerie as gifts for their partners — that’s the main male segment we discussed.
From Russian-market experience, these men may change partners and buy sets worth RUB 30K for each (RUB is used here because the example comes from Russian purchasing behaviour, not a claim about Europe).
We tested Germany + France, and also ran them separately with different keyword sets. Keywords were in English, German, and French.
Then we switched the targeting to match delivery countries: Belgium, Germany, Luxembourg, Netherlands, Spain, Italy, Austria, Portugal, United Kingdom, Ireland, Greece, Sweden, Finland, Denmark, Bulgaria, Latvia, Lithuania, Malta, Poland, Romania, Slovakia, Slovenia, Croatia, Czech Republic, Estonia, Hungary, Cyprus.
Luxembourg was the most expensive for traffic (unsurprising — population in 2021 was 640,064).
Women aged 25–35 in France, separately in Germany.
We tested new creatives on France + Germany.
Then we tested interests per country: cosmetology, fashion & style, beauty & care, love/relationships.
Then we tested other countries — starting with Poland, because Instagram follower analytics showed a lot of audience from there.
We also tested interest combinations using AND logic (not OR): travel + cosmetology.
We collected 25 geo-points of the most expensive SPAs and hotels and intersected that audience with demographics (women 25–35) and the cosmetology interest (which was performing best at that time). We tested a hypothesis about women for whom lingerie is a “work tool.”
We tested several language approaches:
- English creative + “audience interface language: English”
- Russian creative + “interface language: Russian”
- German creative + German interface
- same approach in French.
In the end, France was the most active audience with these settings.
All this time we also kept pushing with retargeting: people who visited the site, interacted with Instagram or ad content, followers, and other segments.
This retargeting setup gives us almost a 1:1 ratio between profile visits and retargeted traffic — plus a cheap “link click” metric.
For Christmas, the client prepared a 20% promo code, and we show it to people who engaged with the account.
Creative ideas we tested
- Model videos — if the preview frame is “safe,” the video can run for a while.
- When an image with a model got restricted, we replaced it with an unboxing video or a flat-lay style photo. Because the client didn’t have many of these assets, customer reviews helped a lot.
- We also tested model photos with text overlay. The system was accepting these at that time.
- We debated the copy for a long time — what it should contain. After iterations and arguments, we landed on a simple format.
MissTease lingerie — the luxury you
deserve!
We are giving you a promotional code! Enter
LoveAtFirstSight in the cart on our official website and
get a 20% discount!
The customer journey we managed to build
- When we ran Google Ads, we drove directly to the website, so it was clear which audience came from Google.
- When we moved into Instagram in autumn, driving only to the website didn’t work. Tests showed that managers were actively chatting in DMs, and customers found it easier to start there — ask questions and begin the order. Driving to the website from Meta became much more expensive than driving directly to the Instagram account.
- So at the time of writing the case, the funnel looked like this: we send traffic to the Instagram account, and then people decide whether to buy and go to DMs.
In ads, we say the promo code is a 20% discount for ordering via the website. But we don’t include the promo code in every ad.
The website does generate purchases, but too weakly to treat it as the primary conversion point with confidence. We used it more as a repeat touchpoint.
What else supports this funnel choice?
We very often saw traffic to the website coming from Taplink (a “link-in-bio” landing page tool, similar to Linktree), meaning:
- They visit the Instagram account.
- They like what they see.
- They click the Taplink.
- So we started reproducing this path in paid traffic as well.
Here’s what it looks like over 90 days in Analytics
And here we see Taplink as a source:
Do we really have leads from ads?
We use a few criteria to count a lead in our analysis table:
- the person must be in Europe
- it must be a new conversation
- when we see a person came from the site and sends screenshots to the manager in chat: we had cases where they couldn’t find the right size on the site and went to Instagram
- in Google Analytics we track add-to-cart and started checkout events tagged as Taplink
- during purchase, managers ask where the person came from; if they say “from a blogger/influencer,” we don’t count that as an ad lead
Results
Best link click: €0.03; average link click: €0.20–0.25; best overall click: €0.01–0.02; average overall click: €0.02–0.04.
CPF (cost per follower) in December: €0.81.
CPL in December: €3.54. Average leads per day: 6 people. We count leads as “DM conversations + add-to-cart events on the site.”
Website purchases in December: €816.
Most recent snapshot: 26 February – 2 March
Clicks: 118 link clicks, 949 total clicks.
CPC:
best link click: €0.05;
average link click: €0.08;
best overall click: €0.01;
average overall click: €0.01.
Leads: 3 people added to cart.
CPF:
best: €0.25;
average: €0.89.
Misses and f***-ups
What we fight on a daily basis:
- Restrictions on model creatives. Even if we can work around the system, if a complaint is submitted, we still have to replace the creative. We try to catch these moments fast.
- Scepticism — and the client often disappears for long periods, leaving us without feedback.
This is what causes the biggest pain in the project.
We can implement many ideas, but one thing the client doesn’t approve is emphasising lingerie quality.
The client has their own experience talking to women who buy lingerie from them often. They are convinced women don’t understand quality. This is not a message we can safely push, even though part of the reviews we analysed suggested the opposite. Yes, many won’t tell “our lace” from French lace, or Chinese fabric from Italian fabric. But they still have a strong idea that they want to buy quality lingerie.
When we saw French audiences respond well to French creatives, we suggested adding French captions in posts (the way they currently duplicate in English). We also created highlights with sizes, FAQs, and an up-to-date catalogue for them. The client refused.
A lot got stuck at “we need it right now.” We didn’t always have the full picture, or we needed new data from the client.
Example: we needed to prep for Black Friday, but we received the information only after it ended. The website was also updated after Black Friday. For Christmas, we managed to create a promo code, but the website update happened much later.
On Black Friday and Christmas, the website homepage had a discount banner, but the discounts were not reflected inside the site itself. For a woman buying lingerie for herself, it would be much clearer if, during checkout, she could see the full price and the discounted price. This increases transparency and visually proves the discount is real.
Another nuance: managers often tell European customers in Instagram DMs to find the website link in the bio header and purchase there.
- When we ran Google Ads by individual countries, some countries barely spent. One of the worst cases was Luxembourg: extremely expensive leads.
- Language targeting using Russian and German performed the worst. It was painful to watch next to retargeting at €0.01.
- In the EU, some Meta message objectives are restricted by policy and regulation (rules differ by objective and placement). DM-focused work also produced tiny results. The ad account reported zeros. We found a few people in DMs who entered from an ad, but it was very little compared to other setups and the overall DM activity when those setups were active.
- At first, we drove to the site. Google brought purchases, but Meta driving to the site brought none. After switching tactics — driving hard to Instagram — purchases returned.
Summary
Meta
Where we landed by the time of this case:
Total spend: €1,765.12 ($1,905.71 / RUB 176,183.06)
Cost per click: €0.09 ($0.097 / RUB 8.99)
Cost per lead: €4.19 ($4.52 / RUB 418.30)
Customer acquisition cost (CAC): €250.87 ($270.95 / RUB 25,086.10)
Total spend: €896.95 ($968.44 / RUB 89,532.41)
Cost per click: €0.06 ($0.065 / RUB 5.99)
Cost per lead: €4.33 ($4.67 / RUB 432.18)
Customer acquisition cost (CAC): €224.24 ($242.12 / RUB 22,383.56)
Average across both platforms
Total spend: €2,605.07 ($2,874.15 / RUB 265,715.47)
Cost per click: €0.075 ($0.081 / RUB 7.49)
Cost per lead: €4.26 ($4.60 / RUB 425.24)
Customer acquisition cost (CAC): €237.56 ($256.54 / RUB 23,870.26)
€4.19 per lead (CPL), with lingerie priced at €100–199 per full set.
With an average order value of €108 and 11 sales, that’s €1,188 in revenue.
This is not profitability yet. Clicks are cheaper than planned, leads are slightly more expensive than planned, and CAC is far from the target.
Reminder: the initial model was:
Cost per click (CPC): €0.14 ($0.15 / RUB 13.97)
Cost per lead (CPL): €3.50 ($3.78 / RUB 349.31)
Customer acquisition cost (CAC): around €50 ($53.98 / RUB 4,990.45)
Roughly speaking, our CPL is about €1 higher than the planned level. At the same time, we regularly see leads below the planned CPL. For example, in the recent snapshot, the lead cost was €3.09.
Google traffic is slightly cheaper than Meta traffic.
But sales… In euros it looks less dramatic; in rubles you immediately feel how expensive it is (that’s why we keep both currencies here — it changes perception for a Russian-speaking team).
Sometimes, of course, we got “golden customers,” like this one.
She came in and bought for RUB 65,000 (included as-is because the client tracked it in RUB; for EU/US readers, treat it as “a high-ticket order”).
This case clearly shows that traffic is not the end of the story.
This is why we grew into a marketing agency after starting as a traffic team. Strategy matters. Digital infrastructure matters.
Remembering our mission — “to seek truth in the love between the giver and the seeker” — and leaning on our homage slogan to a famous formula, SALES = LOVE * TRUTH², we prepared an honest report for the client and proposed the following solutions.
Solutions and what we are doing now
We are talking to integration specialists who can set up full end-to-end attribution and a complete touchpoint chain, so we can see more transparently where orders come from. This will help us understand which traffic sources are most effective and which need optimisation or should be switched off. In other words, it should raise ad efficiency.
With end-to-end analytics, we’ll also understand the actual path: did the person go straight to Instagram, or first land on the website? Or the other way around? That helps us design the best sequence of touchpoints.
For this, we fine-tune events and trigger logic in Google Analytics and Meta Pixel via Google Tag Manager.
We also connected Yandex Metrica (a Russian analytics tool similar to heatmaps/session recordings; used here as an additional diagnostic layer) to test hypotheses:
- People add products to cart as if it were a wishlist.
- At checkout start, they drop off when they discover the full price including shipping.
- Sometimes people accidentally tap a button on mobile and it fires as a “goal”; Google Analytics doesn’t make this obvious.
- People don’t complete the form. We also can’t see whether they paid, because recording stops when they jump to a bank payment link.
- The “cart as wishlist” hypothesis was confirmed.
- The “full price discovery at checkout” hypothesis was confirmed.
The website itself currently has low conversion — this is a one-month snapshot:
So the website is a beautiful storefront — but it doesn’t sell yet.
In Google Analytics, we recently finished adding tracking marks for lingerie selection, so over time we can run ads to audiences who showed interest in specific products.
We first tracked size selection globally, but that doesn’t show which sets people choose. So we set up separate tags for each part of a set. Later, we can transfer these audiences from Google Analytics into Meta and target them precisely. Example: advertise the Sins set to people who selected a size and added Sins to cart on the site.
They are not отображаются in Google Analytics yet. We’re working on it. There are inconsistencies in JavaScript event IDs in the site code.
We also configured conversion tracking in Google Analytics:
There is a visual funnel report. Some of these events are built-in and standard. Others we created: cart, delivery, set-size. Right now, it’s still not fully clear, so we are considering tagging collection views by name as well — so we can track interest in specific sets from here.
What we did to improve website conversion
- We added a second Google Analytics stream into Taplink to make the touchpoint chain more transparent.
- We installed Meta Pixel into Taplink to collect all Taplink visitors (so we can build Taplink visitor audiences and lookalikes and run ads to them).
- In addition to Google Analytics, we reviewed “exit pages” in Yandex Metrica. It showed that a large share of drop-offs happen on collection pages. That means a huge chunk of any traffic is lost at this stage.
- We are working on quizzes. At this stage, we created a quiz skeleton. Quizzes will help warm up users before they visit the site. They will also help users choose a suitable collection on an emotional level.
Proposals to improve website conversion
- Add more selling and emotional description for each collection. Embed the same idea we plan to test via quizzes. In quizzes, we want to implement less “choose lingerie,” more “choose an identity / feeling this lingerie gives.” In plain terms: improve how the lingerie is presented.
This can also be tested faster via traffic (apply the most clickable messaging to the site).
Right now, the site shows only images and video previews before you click into a collection. But inside a collection there is no emotional supporting info. I scroll model photos, pick a size — and that’s it. No text, no video, no media that tries to sell the lingerie on the collection page itself.
- In Yandex Metrica recordings, we can see false triggers sometimes. That suggests the UI might need to be more user-friendly.
- On delivery and payment pages there are Telegram and WhatsApp buttons. We suggested adding them to collection pages too — where most visitors drop off.
Platform-related proposals
- Communicating in the customer’s language
Sometimes people from ads message managers in their native language. Managers are not always ready to respond in that language. Also, English-speaking customers sometimes get sent Stories from Highlights that are in Russian.
Right now this creates uncomfortable situations for customers. Foreign customers who are ready to buy do exist — but Instagram service is not adapted to them.
We proposed two options to the client:
Focus only on Russian-speaking customers — from Russia and from Europe. Managers can handle these requests smoothly, with no language issues and less wasted budget due to communication gaps.
Continue targeting the international market, while actively improving customer experience by communicating in the customer’s language. This can increase the conversion rate from Instagram DMs to purchases. Possible solutions:
- If managers use scripts, create scripts in multiple languages. This is easy now with AI tools such as ChatGPT and Gemini (by Google).
- For non-scripted conversations, use translation tools / AI tools / Telegram bots with built-in translation. These tools can sit “next to” the manager, not inside the chat window.
- Create Highlights for different languages (at least major European ones, or those where ad performance is strong), and label them clearly — France, Germany, etc.
Right now, the “Catalogue” Highlight is in Russian and prices are in rubles. For foreign customers, it is not useful. Also, when a foreign visitor sees a Highlight with their country name, it triggers recognition — “this is for me.” They are more likely to click and buy. This raises personalisation, which often lifts AOV. - If multi-language Highlights are not feasible, at least publish Stories that can be translated inside Instagram. For that function to work, the text must be typed in the Instagram editor itself. If you overlay text in another app and upload as an image, translation won’t work. We also recommend showing prices in multiple currencies. This is a workaround, though. Separate country-specific Stories are better.
Advertising proposals
We tested all relevant interests across Europe. Each time, we simply rotate creatives and show them to the best-performing audiences at that moment.
We noticed that PR brings girls from Dubai and the US. So we can test those countries too. We can test Dubai with English, try hotel geo targeting or broad female audiences. We also had a lead from Australia — DHL delivers there — so demand likely exists in other countries as well.
Now, the key focus with the client is website conversion — rebuilding the site towards better customer usability.
The client liked our report — we keep working.
Credits
Bogdan Zozulya, strategy, marketing, and team lead: https://vk.com/the_redbeard
Anastasia Kiseleva, project manager since late November 2023: https://vk.com/stacykiselyova
Lyubov Simanova, traffic manager: https://vk.com/aimoonsmm
Rita Peryshkina, project manager until late November 2023: https://vk.com/hikaru_snow
To order a service, message us via any of these links: