Internet advertising works. The economic evidence is difficult to dismiss: United States internet advertising revenue reached $294.6 billion in 2025, growing 13.9 percent in a single year (IAB / PwC). Businesses would not invest at that scale if digital media produced no commercial value. But aggregate growth does not mean every campaign works, or that advertisers capture all the value flowing around them.

A person can stand in a rainstorm surrounded by water and still end up thirsty.

That is the central problem of internet advertising. The internet produces an extraordinary flow of content, attention, data, commercial signals, and advertising opportunities. But value does not arrive neatly bottled. It falls unevenly, moves quickly, mixes with contaminants, and often disappears before anyone can collect it. The marketer’s challenge is therefore not simply to find more traffic. It is to build a system capable of catching, directing, filtering, and preserving the traffic that already exists.

Media as a water system

In the Futureproof TMT water analogy, media is the water people actually seek: the news, entertainment, information, conversation, and cultural experience that draws people onto a platform, application, television service, or website. Advertising technology is an auxiliary water system: irrigation, sprinklers, pumps, canals, and controlled taps, built to redirect part of the larger flow toward a commercial purpose.

People rarely visit a publisher because they want advertising. They come for the content. Ad tech does not create the water it handles. It intercepts and redirects water created by media, communications, and consumer activity. Its usefulness depends on whether it can deliver the right amount, to the right place, at the right time, without losing most of it along the way.

The marketing funnel as physical infrastructure

The traditional funnel is usually presented as psychological stages: awareness, interest, consideration, intent, conversion. But it is also physical infrastructure. At the wide opening, the marketer catches a large flow of possible attention. As that flow moves downward, filters remove opportunities that are irrelevant, unusable, duplicated, fraudulent, mistimed, or unlikely to produce value.

• Geographic eligibility, audience and household relevance
• Device, screen, and advertising-format suitability
• Content and contextual quality
• Frequency and prior exposure
• Supply-path legitimacy
• Viewability or opportunity to see
• Attention or receptivity, and conversion probability
• Price relative to expected value

What reaches the bottom is not simply "less traffic." It should be more concentrated value. So a large top-of-funnel number is not proof of a healthy marketing system. A billion impressions can represent a reservoir of opportunity, or a flood of poorly classified, repeatedly resold, low-value auction traffic.

Harvesting inside walled gardens

Walled gardens (large social, search, retail, and video platforms) offer advertisers something resembling managed agriculture. The platform controls the content and user experience, the audience relationship, the identity system, the inventory, the auction, the interface where behavior occurs, and the feedback used to optimize future delivery.

This makes conversion harvesting comparatively predictable. The platform can observe people moving through its environment, recognize recurring patterns, and direct advertising toward users its systems consider more likely to act. The advertiser is still exposed to uncertainty, and the platform’s measurement may not be independently transparent, but the model resembles harvesting from an enclosed orchard: the land has boundaries, the crop is observable, and one operator controls irrigation, cultivation, and collection.

Foraging across the open web

The programmatic open web offers a different proposition. Instead of harvesting inside a single managed environment, advertisers forage across thousands of publishers, applications, streaming services, devices, and audience contexts. Valuable consumers do not spend their entire lives inside one garden. They move across news, entertainment, niche publications, local media, independent apps, and connected TV, and the open web lets an advertiser search across that broader landscape.

But foraging produces a supply problem. Auction opportunities arrive from many directions with different levels of quality, transparency, completeness, and duplication. The same underlying opportunity may travel through several intermediaries and appear to a buyer through multiple supply paths. The Association of National Advertisers found that only 36 cents of each dollar entering a demand-side platform translated into effective, viewable, measurable, fraud-free advertising that reached a consumer, and that "made-for-advertising" sites accounted for 21 percent of impressions and 15 percent of spending (ANA).

This is not an unavoidable property of the open web. More recent benchmarking shows top-performing advertisers converting 54 percent of spending into qualified impressions, versus 32 percent for lower performers. The gap is evidence of infrastructure and operating discipline: some systems capture and purify the flow far better than others. The open web can produce value. It just does not automatically deliver it in a stable stream.

Why the bidder should not have to catch the rain

A demand-side platform is fundamentally a bidder. It receives auction opportunities, evaluates them, and decides whether and how much to bid, formalized as a rapid exchange of bid requests and bid responses (Google Authorized Buyers). That creates an architectural mismatch when the supply arriving at the bidder is uncontrolled.

Even a sophisticated funnel struggles when the rain comes from every direction, several pipes claim to carry the same water, contaminants are only identified after entering the funnel, the pressure changes unpredictably, information is missing from the droplets, and the funnel must pay infrastructure costs merely to inspect the flow.

The industry calls poorly targeted advertising "spray and pray." But spray and pray can begin before a campaign ever launches, in the supply chain itself. Filtration should begin upstream.

The CleanTap proposition

A CleanTap exchange or supply-side platform changes where the funnel begins. Instead of asking each bidder to catch a storm of auction requests, the auctioneer organizes the supply and pours a controlled stream directly into the buyer’s funnel. The model rests on a few principles:

• Classify the opportunity before offering it for auction.
• Remove ineligible or clearly low-value supply upstream.
• Reduce unnecessary duplication and unstable volume.
• Preserve the information buyers need to evaluate value.
• Create consistent, inspectable rules for what enters the flow.
• Deliver a steadier stream aligned with the buyer’s declared needs.

This does not eliminate the buyer’s funnel. The demand-side platform should still evaluate price, audience fit, objectives, frequency, and expected performance. But it should not have to perform every stage of water treatment while simultaneously trying to bid. And the goal is not to enclose the open web. Its diversity is precisely what makes it valuable. The goal is to make open-web foraging more dependable.

From more water to cleaner flow

The industry has historically rewarded volume. Exchanges send more auction opportunities; bidders process more requests; campaigns accumulate more impressions; reports display more activity. But abundance without capture produces little value. The more useful questions are:

• How much relevant supply entered the system?
• How much was lost to duplication or intermediary complexity?
• How much reached an appropriate person, household, device, or context?
• How much had a genuine opportunity to influence behavior?
• How much produced an incremental business outcome?
• What did it cost to filter everything else?

Internet advertising works, but it works through collection and concentration. Walled gardens make this easier by controlling the reservoir, the irrigation system, and the harvest. The open web offers a much larger territory to forage, but its water arrives through fragmented and frequently unstable channels. The answer is not simply a bigger funnel, nor asking the bidder to open its mouth wider and hope for rain.

The answer is better upstream water management: organized catchment, transparent routing, earlier filtration, and a clean tap through which qualified opportunity can flow steadily into the advertiser’s funnel.

Shailin Dhar