Cases · Archive · Logistics · 2026

Infinity Logistix: hypotheses in targeting

Agency-era case study (2026).

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.


Website:

Infinity Logistix

Industry:

Logistics / dispatch service (trucking, USA)

Geography:

USA

Model:

B2B / C2C marketplace

Budget:

Up to $50 per hypothesis test (iterative testing, 70+ hypotheses over 14 months)

Project start:

Not specified (case period — 14 months of work)

Result:

Stable flow of 40–50 leads per week with CPL of $13–20 thanks to segmentation of campaigns and focus on high-performing combinations (Car Hauler + B2C)

Result:

Improved lead quality and higher share of qualified leads through clear segmentation (truck driver, car hauler) and stronger focus on “warm” language audiences (Russian, Tajikistan, Uzbekistan)

Lead generation for a logistics dispatch service

1

Attract drivers who convert into signed contracts

2

Keep CPL within the $15–20 range

3

Increase the share of qualified leads

4

Rapidly test hypotheses (up to $50 per test)

5

Filter out unqualified leads at an early stage

Challenges of promoting logistics through Meta Ads

Cheap traffic

Cheap leads did not convert into signed contracts

No real deals

B2B audiences failed to generate real deals

Inappropriate traffic

Algorithms drove irrelevant, low-quality traffic

Budget without result

Broad campaigns wasted budget without delivering results

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

Optimization stages: from “cheap leads” to stable contracts

- Stage 1. Realizing that CPL ≠ results

At the start, campaigns produced “nice-looking” leads at $16–25, but the funnel data showed failure: people didn’t answer calls, didn’t match criteria, and didn’t convert into deals. It became clear that CPL is a superficial metric that does not reflect business outcomes. From that point, every hypothesis was evaluated based on lead movement toward qualification and contracts.

- Stage 2. Hypothesis testing and audience discovery

We tested B2B, B2C, different language segments, and landing pages. B2B looked good on paper but didn’t generate contracts, while Typeform proved inefficient in cost per lead. In contrast, Russian-speaking and Central Asian audiences delivered higher CPL but significantly better lead quality.

- Stage 3. Working with look-alike and quality filtering

We launched look-alike audiences and separate optimization events. Some hypotheses failed due to either high cost or lack of funnel progression. The best results came from LAL based on website visitors, where both lead volume and real movement toward deals appeared.

- Stage 4. Funnel optimization and abandoning “cheap leads”

We shifted from “lead” optimization to deeper funnel events (funnel key events) and adjusted geography. CPL increased, but lead quality improved: more qualified calls and signed contracts. This reinforced a key principle — it’s better to pay more if it leads to real revenue.

- Stage 5. Repositioning and focus on the Car Hauler segment

We restructured offers around real driver motivations and pain points. This revealed a strong segment — Car Hauler — which became a separate campaign with dedicated creatives and messaging. It resulted in scalable lead flow, gradually lower CPL, and consistent deal flow.

- Stage 6. Failed experiment with broad automation

We tested a single broad campaign with automated budget allocation (Andromeda). It led to higher CPMs and CPCs without qualified leads. The experiment showed that in niche markets, relying solely on algorithms doesn’t work — precise segmentation and manual control are essential.

Write to us on WhatsApp

Impact of systematic Meta Ads optimization

We built not just a set of campaigns, but a managed system: from segmentation to funnel design, where ads are evaluated not by CPL, but by contracts and lead quality.

ROI and key metrics: what systematic hypothesis testing delivered

As a result, advertising stopped being a source of random inquiries and became a predictable acquisition channel: clear segments, controlled unit economics, and a stable flow of drivers reaching contracts.

Over 14 months:

40–50

Leads per week

To 13–20$

CPL reduced

70+

Hypotheses tested

Segments

Profitable segments identified


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.