About Viasat
Viasat is a publicly traded global satellite communications company (NASDAQ: VSAT) with roughly 7,000 employees at the time of the engagement. Its business spans residential and business satellite broadband, in-flight connectivity, government and defense contracts, and international mobile communication services, with operations across NA, LATAM, EMEA, and APAC. For public market context only: annual revenue was reported at roughly $1.92B when the engagement began and roughly $4.64B in the most recent reported year, and the share price moved from around $9 in January 2025 to around $80 in August 2026. These are publicly reported figures describing the environment the work ran inside—they are not outcomes claimed by Stack Finder.
Who we worked with, and the ask
Stack Finder was brought in to own digital growth strategy and performance marketing across the global Fixed Broadband business, covering both DTC and B2B acquisition channels. The scope widened into the performance marketing operating model itself: customer journey optimization, experimentation governance, data modernization, AI-enabled optimization, and cross-functional alignment across Marketing, Product, Technology, Analytics, Finance, Sales, and Executive Leadership.
Spotlight
Global campaigns for Viasat's new satellite launch and service, built to compete head-on with Starlink while the residential DTC funnel and the enterprise, government, and airline motions kept running on the same operating model.
Context & Challenge
What breaks when one company runs two opposite acquisition motions?
A high-ACV enterprise motion measures pipeline over quarters. A residential DTC funnel measures conversion over days. Run both without a shared operating model and you get two marketing organizations wearing one brand: different definitions, different reporting cadences, and no way to compare where the next dollar should go.
Why does a hardware and infrastructure launch need a different campaign model?
Launching a new satellite service is not a feature release. Capacity comes online by region, availability differs by market, and a well-funded competitor is already in the consideration set. Campaign infrastructure has to be regionally aware from day one instead of being localized after the fact.
Why is experimentation governance the prerequisite, not the polish?
Without a shared standard for how tests are designed, sized, and read, every region produces results nobody else trusts. The output looks like a testing program and behaves like anecdote. Governance is what turns regional experiments into a compounding asset.
Our Solution
Global Performance Marketing Rebuild for B2C and B2B
Stack Finder rebuilt global marketing across both sides of the business: the residential DTC broadband funnel and the B2B acquisition motion covering enterprise, government, and airline contracts. One operating model, one set of definitions, and channel strategy set per motion rather than per team preference.
Satellite Launch Campaigns in a Starlink-Contested Market
Global campaigns were built for the new satellite launch and service, positioned directly against Starlink. Beyond rural broadband availability, the campaigns carried the wider service story—in-flight, maritime, and international mobile communications—so the launch was not reduced to a single rural-coverage message.
Paid Acquisition and ROAS Program for International Mobile Communications
Large-scale paid acquisition across international mobile communication services was managed to return on ad spend rather than to lead volume. Spend was allocated against contribution by region and service line across NA, LATAM, EMEA, and APAC.
Customer Journey Optimization Across DTC and Enterprise Paths
The DTC path and the enterprise path were optimized as separate journeys with separate conversion definitions, then reported through a common frame so leadership could compare them without pretending they behave the same way.
Experimentation Governance and Data Modernization
Experimentation was given a defined standard—how tests are proposed, sized, run, and read—and the underlying data was modernized so results were comparable across regions and channels instead of being rebuilt per campaign.
AI-Enabled Optimization and Cross-Functional Alignment
AI-enabled optimization was layered onto the modernized data rather than bolted onto a broken one, and the operating model was aligned across Marketing, Product, Technology, Analytics, Finance, Sales, and Executive Leadership so the same numbers were used in every room.
The Results
Reach: one model across four regions
Acquisition across NA, LATAM, EMEA, and APAC now runs on a single performance marketing operating model rather than regionally improvised programs.
GTM: both motions on shared infrastructure
The residential DTC funnel and the high-ACV enterprise, government, and airline motion operate on shared infrastructure with motion-appropriate measurement, so budget conversations compare like with like.
Efficiency: governed experimentation and modernized data
Tests are designed and read to one standard, and the modernized data layer supports AI-enabled optimization instead of manual reconciliation.
Durability: the environment the model kept running in
Publicly reported figures show revenue moving from roughly $1.92B to roughly $4.64B and the share price moving from around $9 in January 2025 to around $80 in August 2026. Those are market figures, not attributed outcomes—but they describe the scale and pace the operating model had to keep working at.
Frequently asked questions
How do you run performance marketing for a self-serve funnel and an enterprise motion at once?
Keep one operating model and two measurement frames. Shared definitions, shared data layer, and shared governance—but conversion, attribution window, and payback are defined per motion. Forcing one metric across both is what breaks the budget conversation.
How do you launch against an entrenched, well-funded competitor?
Compete on the parts of the service the competitor cannot match rather than on the shared claim. Build campaign infrastructure that is regionally aware from the start, because availability and capacity differ by market and a single global message will misfire in most of them.
Why manage paid acquisition to ROAS instead of lead volume?
Volume targets reward the cheapest impressions, not the most valuable customers. Allocating against contribution by region and service line moves spend toward the segments that actually pay back.
What has to be true before AI-enabled optimization is worth adding?
The data has to be modernized and the experimentation standard has to exist. Optimization on inconsistent inputs produces confident, wrong answers faster—the governance layer is the prerequisite.