Cases · Archive · E-commerce · 2024

FurniFlex

Agency-era case study (2024).

This case was first published on the old humanswith.ai WordPress site between 2019 and 2024 — before the 2025 platform pivot. Performance, SEO, CRM, content, or growth work from that era. Modern AI-search cases (with per-engine Hermes data) live at /cases. Preserved here as written at the time.


Results from using advertising on Meta

Reduction in CPL after changes were made (chart from Alytics by week)

Website

http://furniflex.ae/

Niche

Data-driven marketing, B2C

Location

Dubai

Target region

UAE

Budget

39,500 AED for 3 months

Project start

June 2024

Result

successful combination found for scaling

Result

CPL reduced by half (from 145 to 27 AED)

Result

CPL reduced by half (from 145 to 27 AED)

Result

ROI 1.87-2.49% at 10% conversion rate

Challenges faced by the team

1

Optimize lead cost

2

Improve lead quality

3

Test and implement effective hypotheses within 3 months

4

Set up full-cycle analytics

5

Automate processes

Challenges faced by the team

Lack of existing demand

New product without established market fit in the region.

High lead cost at launch

Initial CPL of 145 AED required urgent optimization.

Limited time to profitability

Need to make the channel profitable within 3 months.

Traffic source tracking complexity

Absence of configured analytics for proper attribution.

Take the first step towards business growth - leave a request!

Consult Now

Book a Free AI Marketing Strategy Call

Filling out the form takes less than 1 minute and helps us prepare a tailored proposal specifically for your business.

1 Information

2 Contact

Website

Annual revenue of your business

1-year average revenue per customer (ARPU/ARPC)

Location

Next

Get Free Strategy in 24h1

The form takes less than 1 minutes to complete and helps us prepare a personalized offer for you and discuss it

1 Information

2 Contact

Your name

Your WhatsApp Number

Your email

How soon would you like to start working with us?

—Please choose an option—We have an urgent taskWithin 1-3 monthsWe are just gathering proposals, no urgencyOther

I agree with the privacy policy.

Back

The scope of work was divided into 3 stages

- Stage 1. Systematic hypothesis testing

Each hypothesis was tested using a clear algorithm:

Hypothesis

For example: «Launching a landing page will generate 32 leads at a CPL of 100 AED».

  • Metrics

CPL, conversion rates, statistical significance (e.g., 570 clicks for analysis).

  • Mockup

Visualisation of the idea before launch (Figma → Tilda).

  • Audiences

Broad and targeted segments (e.g., parents, renters).

  • Budget and timeline

No more than 6 hours to prepare a hypothesis.

  • Result → conclusion → new hypothesis

  • Testing 3 types of landing pagesQuizzes (Marquiz)

Result

67 leads at 66 AED each (within cost KPI).

  • Issue: Low lead quality, 0 sales.

  • Conclusion:

Quizzes are not suitable as a main landing page, but were kept as pop-ups on the website.

  • Action: Traffic was redirected to the main site

Website

  • Result:

20 leads at 145 AED each.

  • Plus: Improved lead quality.

  • Minus: High CPL.

  • Conclusion:

The website can be scaled, but requires optimization.

  • Action: Created a separate landing page

Landing Page (LP)

  • Result:

61 leads at 27 AED each (with a budget of 14,600 AED).

  • ROI: At 10% conversion rate, profitability is 1.87–2.49%.

  • Conclusion:

Facebook* delivers cheap and high-quality leads. Found a channel for scaling.

Screenshot of metrics from the hypothesis test on improving advertising campaigns in ads leading to the landing page

- Stage 2. Development and launch of the LP

Development of the landing page structure and creation of a mockup in Figma

  • Transfer of the mockup while preserving all layers and styles of the client's website using a plugin

  • Coding the landing page while maintaining the design of the main website, adapted for mobile version

  • Creation of a vertical video for the first screen of the landing page

  • Assembly of necessary blocks and transferring them to the Tilda platform

  • Adaptation of blocks for various screen resolutions

  • Publication of the landing page on a subdomain

Stage 3. Setting up advertising

Integration with CRM via WhatsApp

Setup of UTM-tag and ClientID tracking

  • Configuration of target events in Meta Ads

Tracking valuable actions: form submission, video view, session duration

  • Automated lead qualification

Setup of filters in CRM based on behaviour and responses

Write to us on WhatsApp

Project results: how we achieved an effective lead channel

The team developed and implemented a rapid landing page launch algorithm, which reduced the cost of applications by half while maintaining high lead quality.

Key results:

CPL reduced by half

Optimization from 145 AED to 27 AED per lead

70% qualified leads

Lead quality growth after filter implementation

ROI 1.87-2.49%

Profitability at 10% conversion rate


End of archived case

Want a modern case? Hermes data, not Google Analytics.

The seven modern cases at /cases use a different proof model: per-engine AI mention rate, citation tracking across nine engines, and weekly delta reports. Different category, different proof, different era.