How iRunFar Grew Traffic 249% While Affiliate Sites Collapsed

The Google Helpful Content Update didn't discriminate much. It landed on content farms and legitimate enthusiast sites alike, and by most measures, the affiliate web took the hit harder than almost any other category. Traffic evaporated. Revenue followed. Whole media companies that had been built on gear guides and commission links quietly stopped publishing.

Against that backdrop, iRunFar — a trail running site founded in 2006 by a former Washington D.C. attorney named Bryon Powell — grew organic traffic by 249%.

That number comes directly from Ahrefs data, verified on camera in a recent video from the SEO analytics company. The video, published by Ahrefs on their YouTube channel, walks through Powell's operating model in enough detail to function as something closer to a case study than a profile. It's worth parsing carefully, because the lessons cut against some conventional content marketing wisdom, and at least one of them has implications beyond affiliate publishing.

The Contraction That Wasn't a Retreat

When All Gear Digital (formerly Lola Digital Media) acquired iRunFar in January 2021, they did what acquirers of content properties routinely do: they scaled. More guides, more categories, more surface area. The affiliate content industry circa 2021 was running a volume playbook — more pages meant more chances to rank, more longtail keywords captured, more revenue potential. The math seemed straightforward.

Powell says it nearly killed the site.

After the acquisition, he eventually returned with a mandate to cut the guide portfolio down to roughly 20 titles. That's not a trim — that's a demolition. He described his framing for this to Ahrefs: "I would feel bad for anybody in the affiliate world who's trying to manage 75 buyer guides. You're a content farm."

He called the underlying principle "the attention dividend." The logic is compact: every page you publish is a commitment. It needs to be monitored, updated, refreshed when the competitive landscape shifts, and maintained when rankings start to slip. If you can't give a page that kind of attention, it will decay. And decay, in Google's current ranking environment, isn't passive — it actively drags the rest of your domain down with it.

Powell didn't cut to 20 immediately. The mandate hit during holiday buying season, so the team spent November updating 10 guides, December updating another 10, January another 10, then cut roughly 25 guides at the end of January. By that point they'd updated 30 of the 50 remaining titles. Organic traffic climbed 249%. Revenue per buyer guide rose 140%. During the worst content downturn in a decade.

The counterintuitive claim — that cutting inventory increased revenue — deserves a moment's scrutiny. It's consistent with what we know about how Google's algorithms now treat thin or stale content: a site with 75 mediocre pages can perform worse overall than the same site with 30 excellent ones. The page-count-as-moat thesis was always more about capturing long tail volume than building genuine authority. What Powell's numbers suggest is that Google's recent updates have substantially reweighted toward the latter.

What "Testing" Actually Means Here

The most interesting structural feature of iRunFar's operation isn't the content pruning strategy. It's the separation between the people who test products and the people who write about them.

Every shoe that appears in an iRunFar buyer guide has been run for a minimum of 100 miles by an actual ultrarunner in real conditions. Many are tested for considerably longer. The testers aren't the writers. Real ultrarunners accumulate field data over days, months, sometimes years. That data then goes to writers who turn it into guides built to rank and convert.

This matters for a few reasons. First, it makes the testing hard to fake and hard to replicate quickly. A competing site can't spin up 100 miles of shoe data with a content brief and a freelancer. Second, it establishes a credibility floor that Google's systems — and readers — can detect. The specificity of a guide written from genuine field data reads differently than one assembled from manufacturer specs and competitor summaries.

Powell checks rankings on his key pages daily or weekly, adjusting update frequency by category competitiveness. Trail shoes, he said, might get updated six times a year. The schedule isn't arbitrary — it's built from historical traffic data, mapped against how quickly the competitive field turns over in a given category.

This is portfolio management, not content publishing. The mental model is closer to how a fund manager thinks about position sizing than how a traditional editorial team thinks about story counts.

The Gaps AI Found

The Ahrefs video has a second layer that's harder to evaluate cleanly, because it's also a product demonstration. The interviewer shows Powell a tool called Agent A — Ahrefs' own AI agent — and uses it to audit iRunFar's content strategy in real time. The product placement is obvious. But what Agent A surfaced is interesting regardless of who built it.

Finding one: iRunFar has reviewed nearly every major trail running shoe individually — Hoka, Salomon, Brooks — but has published essentially zero head-to-head comparison posts. Not one "Hoka Challenger vs. Speed Goat" article. Not one "Salomon vs. Brooks" piece. Individual reviews exist. The connective tissue between them doesn't. The Ahrefs data showed 800 monthly searches for just one of those comparison queries. Powell's reaction — "there's purchase intent right there, and it doesn't cannibalize at all" — was immediate. He recognized it.

Finding two: iRunFar publishes extensively but almost none of its content is designed to attract backlinks. It's built to convert. Those are different goals that require different content formats. A "State of Trail Running" annual report — the kind of statistical reference document that journalists, coaches, and race organizers would naturally cite — could compound domain authority in ways that another gear guide never will. For comparison, a competitor site called Run Repeat has accumulated 373 referring domains to its state-of-running report. As Powell put it: "It's the long play. And that's part of iRunFar's moat — we've been doing this for 20 years."

Both gaps are real, and both are the kind of thing that's easy to miss when you're inside a site. Powell had clearly been focused, reasonably, on defending his existing rankings. The comparison and authority-content opportunities were sitting adjacent to that work, unconverted.

The Automation Question

Powell currently spends 10 to 15 minutes every day manually checking for ranking declines. Annualized, that's more than 60 hours of monitoring work. Agent A built him an automated decaying-pages detector — scanning the top 1,000 pages, flagging traffic drops by time window, charting lost keywords, identifying which competitors now hold the positions iRunFar lost — for $3.52 in compute costs.

The number is a demo flex, and Ahrefs is selling something. But the underlying point stands independent of the tool: a lot of what operators like Powell do by hand every day can be systematized. The question isn't whether to automate monitoring — it's whether the person doing the monitoring has the domain expertise to act on what the monitoring surfaces. Powell does. Someone running a generic gear site with no real subject-matter depth probably doesn't.

That distinction matters more than the automation itself. Automating the detection of a decaying page is straightforward. Knowing why it's decaying — whether it's a Google update, a competitor refresh, a seasonal pattern, or a structural content problem — and knowing what to do about it requires the kind of judgment that Powell has built over 20 years of ultrarunning and a decade and a half of publishing.

The affiliate sites that collapsed weren't generally failing at monitoring. They were failing at the part that comes after: having something worth ranking in the first place.

Powell's numbers are real. His method is replicable in its logic, if not always in its specifics — most affiliate publishers can't field a team of ultrarunners to test trail shoes for 100 miles each. But the underlying structure — fewer commitments, higher attention per page, testing that actually generates proprietary data, content designed to build authority rather than just convert — is available to anyone willing to make the tradeoff.

The tradeoff being: you have to actually know what you're talking about.