Why Unverifiable Rankings Betray Customer Thinking

I recently left a CX industry publication where I had been the top-ranked author for more than fourteen months. Over the years I contributed more than 150 articles, trusting that the editorial and ranking processes were applied consistently and fairly. My experience changed that belief. There are good editors there are bad ones.

I did not leave because my work declined or because readers disengaged. I left after documenting significant reductions in the engagement figures on my articles — sometimes from one day to the next — that I could not verify with the information available to contributors. I recorded the changes post by post; in some cases the reduction exceeded 60 percent.

When I asked the editor, I was told in writing that the counts could be “adjusted” to align with analytics. Yet no published policy or terms explained how those adjustments were made, when, or how a contributor could review them, and the specific discrepancies I raised were only partly addressed. Within ten days of my first questions, my account was suspended; roughly five weeks without publication followed, with minimal communication. Later, the numerical view counts disappeared from the public rankings page entirely, so positions can no longer be checked against public information.

Readers can draw their own conclusions. What I can say is that I kept everything: 348 timestamped records gathered over four months, two analytical reports, and years of correspondence — and anyone is welcome to contact me privately to see it.

Last month I wrote about the difference between AI assistance and AI authorship. That article came from principle — and, as some readers suspected, from personal experience. This is its continuation, and the second in a series on contributor-platform governance — not to settle a score, but because what happened to me could happen to any writer on a platform where rankings and metrics cannot be independently reviewed. Most contributors would never have enough information to recognise the problem, let alone prove it. Trust is essential in your editor and when is gone is gone.

So I am sharing the facts and one piece of advice: invest your work in platforms with professional editorial governance, published standards, and publicly verifiable metrics.

During my review I watched ranking positions change dramatically — sometimes overnight, with no new publication or visible reason — and watched policies shift over time, sometimes shortly after my exchanges with the editor, making it hard to know which rules applied and when.

One point deserves stating clearly. A ranking system that allows manual adjustment of its numbers, publishes no methodology, keeps no audit trail, and displays no figures cannot be evaluated by the people it ranks. I am not asserting that any improper conduct occurred — this is simply not how a credible system should work. Where contributors cannot verify how a ranking is calculated, maintained, or adjusted, it loses its value as a measure of merit. Writers competing for positions on a list deserve to understand how it is managed, and to ask reasonable questions without fearing consequences. On the platform I left,

The whole thing was so awkward that one day after I communicated that I was out for the Editor, I also got a message from the editor saying the same — as if he had not read my message.

In my view, this is the kind of arrangement that erodes trust. What I can document is the pattern: authors previously ranked around 14th moved into the top ten overnight, with no new publication to explain it; one ranked 8th later appeared at number 3 and was allowed to publish more frequently, and several others rose the same way. Meanwhile, I was told in writing — and I have the emails — to publish only once a week, and later only once a month, even as those authors published more freely. Several of those who rose had comparatively low reader engagement on their topics. As the top-ranked author, my articles carried higher engagement while those figures were still public — until the numbers were removed and the comparison could no longer be checked.

I am not claiming to know why each change happened; I am presenting what the records show and leaving readers to judge. Not every platform runs on transparency, inclusion or fairness — and you often discover which kind you are on only when you ask fair questions.

The Bargain

The contributor economy runs on a quiet exchange. Writers supply free labour — research, drafting, editing, a body of work built over years — and a publication supplies distribution, an audience, and credibility. The currency of that credibility is the ranking: the leaderboard position, the public view count beside each article, the signal that says this writer is read and trusted. For experts who do not pay for placement, a ranking earned on merit is worth more than any advertisement when it is a reliable publication — reputational capital. Yet I watched my position drop from first to second overnight while an author whose output was nowhere near the top suddenly sat at number one. Few days later I resign, my name was removed from the lists. I cannot say why.

The whole bargain rests on one assumption: that the numbers are real — that a view count reflects views, not figures changed overnight, and that a ranking reflects readership. Remove that, and the exchange collapses: the writer still gives labour but receives a signal that may mean nothing. It is the same trust any business asks of its customers: believe our metrics, and we will treat them with care. Asking for an author’s best work is fair; quietly rearranging the numbers behind their ranking and publicans numbers is not.

The Documented Record

What follows is chronological and restrained — every item documented, and available if you ask.

In the first week of January 2026 the editor told me I was number one. Right after, view counters on several of my articles began decreasing — one falling from roughly 560 to below 400. A cumulative counter can only rise. When I asked, I was told counts are manually adjusted when the platform’s WordPress figures diverge from Google Analytics. That was confirmed in writing but appears in no published policy.

I want to be fair, because fairness is the point. Manual corrections can be entirely innocent: WordPress counters and Google Analytics track, sample, and filter bots differently, so deduplication or re-syncing legitimately revises a number occasionally. But an innocent correction has three properties — disclosed as policy, applied uniformly to all contributors, and auditable. In my case, none of the three were present at least in my perspective

Ten days after I first asked about the numbers, my account was suspended — after eighteen months of ‘’cordial collaboration’’. The stated reason was a single disputed article out of 150-plus, one I had already corrected and voluntarily withdrawn. I could not publish for about five weeks, and my entire previous year’s output was placed under retroactive review — a very disproportionate response.

There was also an AI episode, and it is the bridge to what I wrote last month. A year earlier I had openly disclosed how I work: I write everything myself and use AI only for grammar and language, because I am dyslexic and a non-native English speaker — and the same editor accepted that at the time. Days after the platform’s first AI policy took effect, my posts were flagged by an AI-detection tool and moved to draft. The flag was withdrawn in writing within a day as a false positive. Peer-reviewed research in Patterns (Cell Press, 2023) found popular GPT detectors flagged most essays by non-native English speakers as AI-generated while judging native speakers almost perfectly and concluded they are unreliable for high-stakes use. I had already lived the thesis I later wrote about.

Today, who isn’t using AI? Almost no one. Yet I was asked specifically to label my work as AI-edited — framed as a general rule but, as far as I could see, applied to me and very few others. That is the difference between a uniform rule and a targeted one.

From February to April the requirements kept moving — more for me than for others. First sourcing standards (I routinely cited McKinsey, BCG, Eglobalis, ECXO, and other reliable sources and similar). Then frequency: the editor cut me to once a week, which was fair, then to once a month, while others kept publishing several times a week. In my view, a “featured” tier was treated more favourably.

By April the pattern was fully logged. I set it out in two analytical reports, sent privately to the editor on 18 and 20 April 2026, requesting an independent audit. The figures  and emails were never rebutted in writing.

New Ranking Launch

Then came the relaunch. The platform introduced a new ranking system and removed the numerical view counts from the public page. Within days the top ten changed materially — positions moving overnight with no publication trigger, lower-ranked authors appearing suddenly near the top. I dropped from #1 to #2 and within eleven days, announced I would stop submitting, and after being restricted to one post per month, resigned — having never earned a cent. The next day my contributor status was ended, and I no longer appear on the list I had led for over a year and a half. My work is not written for inspiration but to deliver value: roughly 90 percent of it draws on projects that we are doing with Fortune 100 companies

Two written admissions from the editor stand out, and I quote them exactly. The first acknowledged that the feedback I had received was “a series of declined posts without context” — despite published terms promising an editor’s note and thirty days to revise. The second, about a standard suddenly applied to my work, said it “has been part of our standards since January 2026, but I haven’t enforced it consistently.”

The Facts — Dated and Documented

The Arc, in Three Frames

What Good Governance Would Require

I write about customer experience and AI for several magazines when time allows, and the principles that govern a trustworthy customer, readers an contributors  relationship are the same ones that should govern a trustworthy contributor relationship. Four stand out.

1. Metrics that feed public rankings must be auditable. When a number is presented to the public as a measure of merit, there must be a way to check it — the logic behind The Trust Project and its eight Trust Indicators, a global transparency standard for showing readers what stands behind published work. A ranking with no verifiable figures fails by design.
2. Adjustments must be disclosed and uniform. Platforms that take rankings seriously publish how distribution works — Medium documents its Boost and curation rules; Substack explains how recommendations operate. You may dislike the mechanics, but they are written down and applied to everyone. A correction policy that lives only in private email is the opposite.
3. Ranking transparency is now a baseline legal expectation. The EU’s Platform-to-Business Regulation (2019/1150) requires providers to set out the main parameters behind rankings; the Digital Services Act adds plain-language duties for recommender systems. If you rank people publicly, you owe them an honest account of how.
4. Published processes must be followed — by the editorial team too. A promise of an editor’s note and thirty days to revise means nothing if ignored. A standard introduced in January cannot be applied to last year’s work, or to one contributor while others are exempt. And AI detectors are not fit to sit as judge: treating AI-assisted human writing as AI authorship punishes the very people — the dyslexic, the non-native — who use these tools to be understood.

A Checklist for Any Contributor Prior to Contribute

Before giving your experience or years of free work and time to any platform, ask:

• Is the ranking methodology published and clear, or does it keep changing?

• Are the metrics independently verifiable, or does a single person control them?

• Is there a disclosed, uniform policy for adjusting metrics?

• Is there a written process for declined content, and is it followed for everyone equally?

• Does the AI policy distinguish assistance from authorship — or is it an old-style standard struggling to adapt?

• Are standards applied prospectively, and to everyone?

• Does the publication run on documented, consistent process — or on one person’s judgment and mood? You will usually sense this early.

• Are you keeping your own records?

If the answer to most of these is no, you are not earning reputational capital. You are lending it, unsecured. Better to wait fo the right platform.

Closing

I left the platform, and I am at peace with it, because of something every customer-facing professional eventually learns: readers attach to writers, not to leaderboards. A ranking can be relaunched, reordered, or erased overnight. A loyal reader cannot. Reputation does not belong to the platform that displays it; it belongs to the writer who earned it. In the end, we can only change ourselves, not others — so I chose to walk away and write this instead.

You cannot control how a platform behaves or their editor, only how you respond. Platforms earn contributors the same way businesses earn customers — through transparency, consistency, honesty and a willingness to be held to their own published word. When a platform behaves well only until someone asks a hard question, it tells you how much your trust was ever worth. So, judge platforms not by their rankings, quality of material or their editorial standards, but by how they behave the moment a contributor asks to see the math. That should never be a problem and get the editor defensive. Customers think and readers and contributors think a lot; the problem only appears when the editor doesn’t.

The complete record exists — every capture dated and preserved, alongside the two analytical reports. It is available privately to anyone who messages me.

👉 Stay ahead of CX, AI, and innovation  — Subscribe to my weekly LinkedIn Newsletter “CX Insights by Ricardo S. Gulko.

Feel free to share it — and let’s connect on LinkedIn for more insights and future posts: Ricardo Saltz Gulko

My columns in several respected CX publications.

• My recent articles on Eglobalis: https://www.eglobalis.com/blog/
• My recent articles on the European CX Organisations ECXO.org : https://ecxo.org/blog/
• My recent articles on CMSWire: https://www.cmswire.com/author/ricardo-saltz-gulko/
• My German articles on CMM360: https://www.cmm360.ch/author/ricardo/
• My old articles on CustomerThink: https://customerthink.com/author/