VaryLife real estate in Czechia: 1,506 leads and €2.71M in sales

Next, we tested an offer for investors: the campaign generated 7 leads at €29 each, 2 of which were qualified (cost per qualified lead: €104).

We also tested new audiences: Russian-speaking expats in Germany, along with German- and English-language ads for local audiences. These hypotheses did not deliver the expected results (1 lead at €15,34 and no return from the other tests).

By the third month, we were running short of new promotional materials, which reduced campaign effectiveness.

We reworked the existing photos and videos, using AI and stock assets to create original creatives. These supported limited-time offers:

• penthouse discounts (1 lead at €39,61);
• an offer for women aged 50+ (8 leads at €8,84 each).

The campaign for women aged 50+ was based on the client’s internal analysis, but it produced weak results.

We analysed several months of CRM data in detail. It showed that the typical buyer:

• was usually a man;
• was buying for investment;
• used a combination of personal funds and a mortgage.

These insights informed subsequent campaigns, including revised offers and creatives with static floor plans and emotional videos about life in the development.

After six months of working together, we concluded that Meta was the main paid source of contacts for the sales team.

Of 15 deals, 6 were closed with contacts generated through paid social, with a total value of €2 724 361,99.

Going forward, we planned to post more actively on social media to maintain brand awareness and grow an audience within the profile for continued communication.

After reviewing the results, we decided to focus upcoming campaigns on people considering an apartment as an investment and explain the financial benefits of investing in more detail.

We also offered help with finding the best mortgage terms through an in-house mortgage specialist.

We recommended adding a mortgage calculator to the website and posting more actively on social media to maintain brand awareness and build an audience for ongoing communication.

By September 2025, the best-performing campaigns were beginning to fatigue: frequency was rising while lead volume was falling. The strongest format at that point was a static creative showing an apartment floor plan and its features as numbers. To improve incoming lead quality, we added the location name to the main ad copy, helping filter out people for whom Karlovy Vary was too far away. A separate campaign targeting Prague only did not work. Late in the month, video campaigns performed better than static creatives on cost per lead.

In October 2025, we achieved the best weekly result of that period: 42 leads, including 8 priority leads, with two meetings scheduled. We used AI videos and static creatives from our existing asset library; “floor plan + features” was again the best-converting format. We also tested a new form with a question to filter out leads from unsuitable locations. The longer version generated 3 leads; the shorter version generated none. One additional insight: we immediately stopped the campaign featuring apartment B-06 after learning that the property had already been sold.

In November 2025, we consistently generated 21–27 leads a week, including 4–7 qualified leads. Video V-16 performed best, generating 11 leads at 187 CZK, compared with an average account-wide cost of around 350–400 CZK. Retargeting generated no leads and had a high CPM, so we turned it off. The first “viewing done” and “meeting scheduled” statuses were recorded.

In December 2025, we launched campaigns with new offers for apartments A-29 and A-03, discounted units and videos for stage 3. Weeks with 40 leads became normal, including 5–7 priority leads. The “Old Form” generated inexpensive leads at 150–210 CZK, but many were unqualified, so we stopped it. By the end of the month, all discount offers had ended, the campaigns were switched off and preparations for the new year had begun.

In January 2026, we focused on offers for stages 2 and 3. One unexpected insight: a single static A-29 creative with an attractive price on the banner generated 17 of the campaign’s 19 leads, showing how strongly the offer influenced conversion. By the end of January, the best result of the month was 45 leads at 168 CZK in one week. We also tested website landing pages: campaigns optimised for several forms generated no leads, so we relaunched with a single form.

In February 2026, we actively tested website advertising. Video creatives were too expensive for this channel: cost per lead reached 1200–1500 CZK, and the campaigns were paused one by one. Facebook lead forms continued to perform steadily, generating 15–18 leads a week at 155–200 CZK. At the end of the month, we switched to broad targeting without interests. It generated 5 website leads at a lower cost, including 1 qualified “high” lead, and the first viewing-done status was recorded.

March 2026 was the best month of the collaboration. In the week of 12–19 March, we generated 55 leads at 259 CZK, including 10 qualified leads (2 middle and 8 low), and scheduled a viewing. Open Day campaigns generated 17 leads at 378 CZK, though the CRM revealed that some were duplicated. Lead forms generated 38 leads at 162 CZK, while the best video generated 21 leads, 8 of them qualified. In the second half of March, the Open Day campaign stopped generating leads and was switched off, while Facebook forms maintained volume at 21 leads at 230 CZK.

In April 2026, we wrapped up Open Day activity and began testing a new video format, B-19. The first launches generated 2 leads at 168 CZK, one of them qualified. By the end of April, B-19 was generating 7 leads a week, but campaigns dropped off completely on 29–30 April, with no leads for two days, so we relaunched them. We also prepared a 10/90 offer (10% down payment and 90% mortgage) for the stage 2 audience.

May 2026 started with expensive, unqualified leads from mortgage and stage 2 offers; the week ended with 12 leads and none qualified. We switched the stage 2 campaign to retarget people who had already shown interest but had not qualified on budget. In mid-May, we tested a form with two filtering questions: it reduced location-mismatched leads but cost more and generated less volume. The breakthrough came in the last week of May: an AI creative showing the apartment’s minimum price generated 22 leads at 126 CZK, the best result of the month. At the same time, a simplified form with one budget question generated 34 leads at 240 CZK in a week.

In June 2026, results stabilised at 36–40 leads a week at 310–317 CZK, including 6–10 qualified leads.