UpTrajectory Review

The marketing analytics world has a junk drawer problem, and it's getting worse. Chris Robson's piece in MarTech draws a sharp parallel between household clutter and a growing crisis in customer attribution: the '(direct)/(none)' bucket in your analytics platform, where visitors arrive with no traceable source. What once was assumed to be loyal customers typing your URL directly has ballooned into something far messier—privacy-hardened browsers, stripped referrer data, and now AI agents browsing on behalf of humans. When this residual category swells to thirty percent of your traffic, it stops being a rounding error and becomes a strategic blind spot. Robson wisely reaches outside marketing orthodoxy, borrowing from Pew Research's long struggle with the 'nones' in religious affiliation studies, where the unaffiliated category has become America's largest 'denomination.'

For small-business operators, this is not an abstract data hygiene issue. It is a direct threat to every dollar you spend on customer acquisition. If you cannot distinguish between a TikTok-driven visitor whose attribution got stripped by iOS privacy settings, an AI agent scraping your pricing, and a genuinely direct type-in customer, you are optimizing your marketing budget on a foundation of guesswork. The smaller your operation, the more dangerous this becomes—enterprises can absorb attribution fuzziness across massive spend; a local retailer or niche B2B service cannot. You may be doubling down on channels that merely appear to perform while starving the ones actually driving growth, or conversely, pouring money into platforms that show up cleanly in analytics but deliver hollow traffic.

What distinguishes Robson's framing is his refusal to treat this as a purely technical fix. The piece's most valuable move is importing the social science concept of 'residual category'—the formal acknowledgment that some classification systems contain an uncategorizable bucket by design. This is genuinely under-reported in marketing discourse, where vendors relentlessly promise 'full-funnel visibility' and operators internalize failure when their dashboards show gaps. Robson suggests the opposite: the gap is structural, not personal. We are skeptical, however, of how far the religious 'nones' parallel stretches. Pew's challenge involves self-reported identity; marketing 'nones' involve technical opacity. The solutions differ—surveys and ethnography for Pew, server-side tracking and probabilistic modeling for marketers. Conflating them risks suggesting patience where urgency is required.

The downstream effects split unevenly across the market. Large platforms with logged-in user bases—Meta, Google, Amazon—gain relative advantage as their first-party data becomes more valuable against the rising tide of untraceable traffic. Small businesses without such ecosystems become more dependent on these same gatekeepers, or on expensive attribution tooling that promises clarity it cannot fully deliver. Meanwhile, AI agent traffic, barely mentioned in most marketing circles until recently, threatens to further pollute analytics without corresponding human intent. The cost is not merely misallocated spend but strategic confusion: businesses may abandon channels, products, or even market positions based on phantom signals from their 'nones' bucket.

What to watch: the emergence of server-side and first-party data infrastructure as table stakes rather than enterprise luxuries. What to do now: audit your '(direct)/(none)' share monthly, not quarterly, and segment it aggressively by landing page, device, and behavior pattern to infer what hides inside. Most critically, invest in post-purchase surveys asking 'how did you hear about us?'—low-tech, high-fidelity, and increasingly necessary as technical attribution frays. The junk drawer will not empty itself, and pretending otherwise is a luxury no small operator can afford.

“What happens when a third of what we own is in the junk drawer?” — MarTech

Takeaway: Audit your '(direct)/(none)' traffic monthly and add post-purchase source surveys before technical attribution gaps distort your spending decisions.