Performance marketing case studies

Case 01 · Telecommunications · Meta Ads + Google Ads

Meta Ads System for Prepaid Telecom | BAIT Walmart Case

10x acquisition scaling in maximum competition environment · 21 months

21 months

10x

Lead scaling

100K

Monthly leads at peak

27–30%

Conversion rate

+100% · from 15% to 30%

$3.5M

MXN spend/mo

Context and problem

BAIT is Walmart's MVNO focused on prepaid portability. At the start of the operation, the digital acquisition system generated 8,000 leads/mo with a ~15% conversion rate and no defined scaling structure. The competitive environment includes Telcel, AT&T and Movistar — operators with 10x higher acquisition budgets. The business model imposes a very restrictive CAC because revenue per prepaid portability is low and the error margin is minimal.

System architecture

Meta Ads

Main volume engine. Full funnel: cold acquisition via Lead Ads with behavioral segmentation (portability, data usage, specific telecom interests) + warm retargeting + unconverted lead reconversion. Lookalikes built from successful portabilities fed back from CRM.

Google Ads

High-intent complement. Search for active portability queries ('switch to BAIT', 'BAIT Walmart plan') that capture the user at the exact decision moment. Lower volume but higher CVR. Display retargeting for site users without conversion.

CRM Loop

The system's differentiator: successful portability feedback from BAIT's CRM to the platforms. The algorithm optimized not just for captured lead but for lead that completed portability — this improved audience quality and CVR over the 21 months.

Escalamiento

Gradual budget progression: from ~$800K MXN/mo initial to $3M–$3.5M MXN/mo without losing CPL efficiency. Method: identification of ad sets with best volume/quality ratio → duplication → audience expansion without saturating optimization signals.

Strategic decisions

Not optimizing for CPL — optimizing for real portability

Most telecom lead gen systems optimize for cost per lead. The error is that a cheap lead that doesn't port has a higher real cost than an expensive lead that does. Integrating the CRM to feed successful portabilities back to the algorithm was the decision that most impacted CVR in the long run.

Gradual budget, not jumps: protecting optimization signals

Meta Ads loses efficiency when budget rises more than 20–25% in a week because it restarts the learning phase. Abrupt jumps generate CPL spikes. Gradual scaling protocol (+15–20%/week on well-performing groups) kept CPL controlled while volume multiplied by 10.

Segment by real behavior, not demographics

Demographic audiences generate volume but low quality. Segmentation by mobile data usage behavior, operator switching frequency and specific portability interests generated leads with higher porting intent — reflected in CVR increase from 15% to 27–30%.

Metrics evolution

Metric	Start	System implemented
Monthly leads	8,000	80,000 – 100,000
Average CPL	$20 MXN	$35 – $45 MXN
Conversion rate	~15%	27% – 30%
Monthly spend	~$800K MXN	$3M – $3.5M MXN
Duration	—	21 sustained months

Does your operator have CPL that rises when scaling or out-of-control CAC? I review the current architecture and tell you where the efficiency leak is.

Let's analyze the architecture

Case 02 · Telecommunications · Meta Ads

Building Postpaid Lead Gen from Scratch | BAIT Postpaid Case

Lead quality over volume in higher-value-per-user model · 6 months

6 months

35K

Leads/mo from zero

$25

Avg CPL MXN

20%

Conversion rate

$1.2M

MXN spend/mo

Context and problem

BAIT's (Walmart) postpaid product had zero structured digital acquisition operation at the start. The postpaid model fundamentally differs from prepaid: higher value per user, higher closing friction (requires identity validation, credit score and point-of-sale visit) and lower tolerance for low-quality leads — an unqualified lead consumes sales team time without generating revenue.

System architecture

Meta Ads

Segmentation differentiated from the prepaid model: audiences with higher apparent creditworthiness (consumption behavior, personal finance interests, C/C+/B socioeconomic level). Forms with basic qualification questions to filter prospects without postpaid profile before reaching CRM.

Lead Quality System

The main KPI was not CPL but lead-to-activation conversion rate. This oriented all decisions: creatives with more specific message (less volume, more qualification), landing pages with onboarding process info (reduces closing friction), and basic lead scoring before passing to the sales team.

Strategic decisions

Defining success at activation, not capture

Many lead gen systems are measured by CPL. For postpaid with high closing friction, CPL is a misleading metric — a cheap lead with 5% activation is worse than an expensive one with 20%. Orienting all optimizations toward activation rate (not CPL) was the system's most important decision.

Differentiating segmentation from the prepaid model from day 1

A common error is using the same audiences from the prepaid product for postpaid. The buyer profiles are different: postpaid requires a credit-capable profile with stable income and less price sensitivity. Building audiences specific to this profile from the start avoided budget waste on incorrect leads.

Metrics evolution

Metric	Start	System implemented
Monthly leads	0 (from scratch)	30,000 – 35,000
Average CPL	N/A	$25 MXN
Conversion	N/A	18% – 20%
Monthly spend	N/A	$800K – $1.2M MXN
Time to result	—	6 months

Case 03 · Telecommunications · Meta Ads + Google Ads

Dual Prepaid + Postpaid Acquisition System | Telefónica Movistar Case

Scaling in high-competition market with multi-product system · 18 months

18 months

Prepaid 12 months

Scaling

20K

Leads/mo

Metric	Start	System
Leads/mo	~5,000	~20,000
Avg annual CPL	~$90 MXN	~$75 MXN
CPL 2026	—	~$30 MXN
Conversion	—	~18%–20%
Spend/mo	—	$1.5M–$1.8M MXN

Postpaid 6 months

17K

Leads/mo from zero

20%

Conversion

Metric	Start	System
Leads/mo	0 (from scratch)	+17,000
Average CPL	N/A	$85–$90 MXN
Conversion	N/A	18%–20%
Spend/mo	—	$1.2M–$1.5M MXN

Dual system architecture

Meta Ads Prepago

Scaling from 5K to 20K leads/mo with portability behavioral segmentation. Creative rotation system to avoid audience saturation: format rotation every 3–4 weeks. Lookalikes of active prepaid customers with higher retention.

Google Ads Prepago

High-intent search for active portability queries. Contribution to total CPL thanks to leads with higher CVR that offset the higher CPC. Performance Max to capture demand across Google's entire network.

Postpago desde cero

System built with different logic than prepaid: focus on C+/B profile, less volume, more quality. Forms with pre-qualification questions. 0 → 17K leads/mo in 6 months maintaining 18–20% CVR. Budget independent from prepaid with its own KPIs.

Case 04 · Pharmaceutical Retail · Meta Ads + Google Ads

From Unstable ROAS to 835%: Pharmaceutical Omnichannel System | YZA FEMSA Case

Transformation from fragmented digital operation to consolidated Always On system · 24 months

24 months

$25.6M

MXN total attributed revenue

835%+

Sustained avg ROAS

+294K

Purchases generated

+2,289%

eComm revenue growth

Context and problem

Farmacias YZA, Farmacon and Moderna (three brands under FEMSA Salud) had a fragmented digital operation: Meta and Google operating without unified strategy, ROAS fluctuating between 42% and 147% with no consistency, no Always On structure and no attribution model connecting eCommerce with in-store traffic. The result was total dependence on seasons (Buen Fin, Christmas, Mother's Day) to sustain positive numbers.

System architecture

Always On

Campaign structure with fixed base budget oriented to the highest-margin, fastest-moving products. These campaigns operate 365 days a year, regardless of seasons. They are the base of constant revenue that eliminates monthly volatility.

Catálogo Dinámico

SKU catalog integration with Meta Advantage+ Catalog Ads. Specific product remarketing for users who visited SKU pages without buying. Automatic prioritization of SKUs with highest historical CVR for dynamic budget allocation.

Atribución Omnicanal

Unified dashboard connecting eCommerce revenue + attributable in-store traffic estimation from digital campaigns. Data-driven model to evaluate real contribution of Meta vs Google vs organic traffic. Without this model, reported ROAS undervalued the system's real impact.

Temporadas Incrementales

With the optimized Always On base, seasonal budget increases generate efficient (not chaotic) scaling. Additional budget is injected on already validated audiences and SKUs — not on new campaigns from scratch that require a learning phase at peak demand.

Strategic decisions

Prioritizing Always On over seasons — against team inertia

The internal team was used to activating big campaigns in seasons and pausing the rest of the year. Changing this logic required demonstrating with data that acquisition cost in peak season (when everyone competes for the same ad inventory) is 2x to 3x more expensive than in normal season. The Always On base is more efficient in terms of MER.

Building the attribution model before scaling budget

The initial ROAS of 42–147% was not just a campaign problem — it was a measurement problem. A significant portion of revenue generated by digital campaigns was realized in-store and not attributed to the digital channel. Correcting the attribution model before optimizing campaigns avoided making decisions based on incomplete data.

Optimizing for ROAS in the long run, not CPA in the short

The team wanted to optimize by CPA (cost per purchase) because it's the most direct metric. The problem: optimizing by CPA in pharmaceutical retail with a wide catalog tends to concentrate budget on low-ticket SKUs that have better CPA but lower total revenue contribution. Changing the optimization objective to ROAS (with a realistic target ROAS) resulted in total revenue stabilization.

Full 24-month evolution

Metric	Jan–Jul 2023	Aug 2023–Jun 2025
ROAS	42%–147% (unstable)	650%–800% (consistent)
Avg purchases/mo	~3,100	6,000 – 7,000
Total revenue	N/A	$25.6M MXN
eComm revenue	~$749K MXN	$17.9M MXN
Model	Fragmented / reactive	Always On / omnichannel

Does your pharmaceutical retail have volatile ROAS or depend on seasons for positive numbers? Let's analyze the current structure and see what system can be built.

Let's analyze together

Case 05 · Luxury Automotive · Google Ads + Meta Ads

High-Intent Lead Gen in Luxury Automotive | Mercedes-Benz Case

Quality over volume: lead volume was the problem, not the solution · 12 months

12 months

+3x

Qualified leads/mo

-38%

CPL reduction

24%

Scheduled test drives

$420K

MXN spend/mo

Context and problem

Authorized Mercedes-Benz dealer in Mexico. Luxury segment with avg ticket $800K–$2.5M MXN. The initial situation was a system generating ~180 qualified leads/mo with $1,850 MXN CPL and only 8% test drives scheduled. The diagnosis was counterintuitive: the problem was not the low lead volume — it was that 70% of captured leads had no real luxury buyer profile, which saturated the sales team with prospects without real intent.

System architecture

Google Search

High-intent capture: transactional keywords exclusively ('[model] price Mexico', 'Mercedes dealer CDMX', 'Mercedes-Benz for sale'). Informational keywords (comparisons, reviews, specs) that brought volume without purchase intent were excluded. Higher CPC but 3x higher conversion than general traffic.

Meta Ads

AB/A+ segmentation: luxury interests (business class travel, golf clubs, premium wines), premium consumption behavior, lookalikes from historical buyers. Retargeting at 30, 60 and 90 days for the extended decision process — a luxury buyer can take 3 months between first contact and dealer visit.

Calificación de Lead

Forms with qualification questions before passing to the sales team: specific model of interest, approximate budget, purchase timeline. This reduced sales team time on non-profile prospects and increased scheduled test drive rate from 8% to 22–26%.

Strategic decisions

Lead volume was the problem, not the solution

The team requested more leads. The diagnosis was that the system already generated leads — just that 70% had no luxury buyer profile. More volume of the same type of lead only further saturated the sales team. The solution was not to scale volume but to redesign segmentation to capture fewer but correctly-profiled leads.

Reducing budget to improve quality

Counterintuitive: when redesigning segmentation, reach dropped and CPL initially rose. The decision was to stay the course because the real KPI was not CPL but cost per scheduled test drive (proxy for real sales opportunity). After 6 weeks, the total CPL dropped 38% because the system learned to identify the correct profile.

Aspirational creatives, not offer/price

In luxury, discount or price messages generate incorrect leads — they attract price-sensitive people who are not the target buyer. The creatives that worked for Mercedes-Benz avoided any mention of price and focused on experience, status and brand positioning.

Metrics evolution

Metric	Start	System implemented
Qualified leads/mo	~180	550 – 600
Average CPL	$1,850 MXN	$1,140 – $980 MXN
Scheduled test drives	~8%	22% – 26%
Monthly spend	~$260K MXN	$380K – $420K MXN
Duration	—	12 months

Does your company generate high lead volume but low closing rate? The problem may not be volume — it may be the quality of the captured profile. Let's analyze the segmentation.

Let's analyze the system

Experience you can audit.

Cinco operaciones. Distintos mercados. The same system judgment.

Meta Ads System for Prepaid Telecom | BAIT Walmart Case

Context and problem

System architecture

Strategic decisions

Metrics evolution

Building Postpaid Lead Gen from Scratch | BAIT Postpaid Case

Context and problem

System architecture

Strategic decisions

Metrics evolution

Dual Prepaid + Postpaid Acquisition System | Telefónica Movistar Case

Dual system architecture

From Unstable ROAS to 835%: Pharmaceutical Omnichannel System | YZA FEMSA Case

Context and problem

System architecture

Strategic decisions

Full 24-month evolution

High-Intent Lead Gen in Luxury Automotive | Mercedes-Benz Case

Context and problem

System architecture

Strategic decisions

Metrics evolution

Otros proyectos documentados

Does your company have CAC, CPL or scaling problems?