MFA (Made For Advertising)
Definition & Explanation
Made For Advertising (MFA) refers to websites built primarily to capture programmatic ad spend rather than to serve an audience. The content exists to justify the ad slots, not the other way around.
The underlying business model is often arbitrage. An operator buys traffic cheaply, typically through content recommendation widgets, clickbait social campaigns, or paid search, and sends it to pages engineered for maximum ad yield: dense ad layouts, rapidly refreshing placements, and pagination designed to multiply impressions. As long as the revenue per visit exceeds the cost of acquiring the visit, the operation is profitable, and it scales for as long as buyers keep bidding.
What makes MFA distinct from ad fraud is that the traffic is usually real. These are human visitors seeing real ads on real pages. Nothing about an individual impression is invalid, which is why MFA sites pass IVT checks. The problem is the environment: users arrive with no intent, engage with nothing, and leave. The impressions are technically genuine, but practically worthless.
In September 2023, the ANA, 4A’s, ISBA, and WFA jointly published criteria for identifying MFA sites (1). The shared characteristics:
- High ad-to-content ratio
- Rapidly refreshing ad placements
- High percentage of paid traffic sourcing
- Generic, non-editorial, or templated content
- Poorly designed, templated site layouts
No single criterion defines MFA on its own. It is the combination, anchored by paid traffic sourcing, that separates an arbitrage operation from a merely aggressive publisher.
Why It Matters
The sharpest problem with MFA is that it defeats the tools buyers rely on. IVT vendors pass it because the traffic is human. Viewability vendors often score it well, since dense, refreshing layouts are built to fill the screen. A campaign can hit every quality KPI on paper while a meaningful share of its budget lands on inventory no one would defend in a screenshot.
The scale was documented in the ANA’s 2023 Programmatic Media Supply Chain Transparency Study, which found MFA sites accounted for 21% of study impressions and 15% of spend (2). The average campaign in that study ran across roughly 44,000 websites, a footprint no buyer was inspecting site by site. MFA thrives in exactly that gap, concentrated in the long-tail inventory that algorithms buy and humans never review.
MFA also distorts the market it operates in. By flooding exchanges with cheap impressions, it drags down average CPMs and makes legitimate publishers look expensive by comparison. Advertisers optimizing toward low CPMs are, in effect, optimizing toward MFA.
The definitional debate carries its own cost. After the criteria were published, several automated labeling systems flagged premium publishers on the basis of high ad density alone (3). A single-signal approach punishes legitimate publishers who monetize aggressively while missing arbitrage operations that keep their layouts clean. Getting the classification right matters as much as doing it at all.
MFA Across the Ecosystem
Industry response since 2023 has moved in two directions at once. On the demand side, awareness translated into action: ANA benchmark data showed MFA spend among participating advertisers falling from 15% to roughly 4% within a year, as buyers tightened inclusion lists and pressured DSPs for MFA controls (4). On the supply side, the inventory itself kept growing. DoubleVerify measured a 19% year-over-year increase in MFA impression volume in 2023 (5). The sites did not disappear; the spend redistributed toward buyers not paying attention.
The definitional conversation also matured. Industry discussion converged on the view that MFA should effectively mean “Made for Arbitrage,” making paid traffic sourcing a minimum requirement for the label (3). This narrowed the definition usefully, protecting content-driven publishers from false positives. It also left a gap: plenty of inventory fails the screenshot test without meeting the arbitrage bar, and buyers want visibility into that inventory too.
Verification vendors, DSPs, and curation marketplaces now offer MFA filtering as a standard control, though methodologies vary widely, and a domain flagged by one vendor may be clean by another’s definition. That inconsistency is a direct consequence of how the classification is built.
MFA in DeepSee.io Metrics
DeepSee partnered with the ANA on the 2023 transparency study and worked alongside the ANA, the 4A’s, and Jounce Media in the working groups that formalized the criteria above. We had been compiling a list of arbitrage-driven sites well before the term MFA entered wide circulation, flagging domains that met at least two of the five criteria, nearly all of which ran on majority paid traffic.
Because a single MFA label collapses distinctions that buyers care about, we classify flagged domains into three categories, each with a concrete minimum requirement:
- Arbitrage: requires a high percentage of paid traffic sourcing. This is the classic MFA model, and the category the industry consensus definition describes.
- Ad Clutter: requires a high ad-to-content ratio. These sites may not source paid traffic, but their monetization overwhelms their content.
- Template Site: requires generic content on a templated site design. These are low-effort properties, often produced at scale, usually reliant on AI, and currently the fastest growing segment of our list.
Together these form our Arbitrage, Clutter, and Template Site (ACT) list. Subscribers receive every flagged domain with its classification, so a buyer who tolerates cluttered but legitimate publishers can filter differently than one excluding anything that resembles a cash grab.
The classifications are grounded in observed behavior rather than reputation. We visit sites directly, measuring ad density and ad refresh timing as they actually occur. Those signals cluster in MFA environments: 54.4% of MFA domains run aggressive refresh schedules, roughly three times the 17.6% rate of clean domains. Content and design signals feed the Template Site classification through our content metrics. The result is an MFA determination a buyer can interrogate, criterion by criterion, rather than a black-box label.