About Mo RezaAli

I’m Mo RezaAli. I work as a Growth Marketer at Yakkyo S.p.A., and I live in Messina, on the Sicilian side of the strait.

Yakkyo is a listed company. Ticker YKY, ISIN IT0005573966, traded on Euronext Growth Milan, which you can check on Borsa Italiana, on Euronext, and on the company’s own investor relations page. Its consumer-facing product is Yakkyofy, a sourcing and fulfilment platform for e-commerce sellers. Yakkyofy is a product of Yakkyo S.p.A., not a second employer. I have one employer.

The work is acquisition and retention for an e-commerce product: paid and organic channels, landing pages, lifecycle email, and the measurement setup underneath all of it. My title is Growth Marketer. Not senior, not head of anything.

What went wrong on this domain, and when

Before you read anything else here, read this.

From 31 October 2025 to 27 April 2026 this site published on an hourly schedule without me. A Cloudflare Worker I wrote pulled five RSS feeds, passed 450-character excerpts to gpt-4o-mini, and committed the output straight to main. Nobody read it, including me. The repository records 842 Add post: commits over that window. What was left standing at the end was 739 English articles, and 2,922 markdown files once the machine translations into Italian, Arabic and Simplified Chinese are counted, all under my byline.

The validator I built to keep quality up is what broke it. It required every post to contain at least one number and at least two named entities, extracted from a 450-character excerpt that usually contained neither. So the model supplied them. That is not a bug in the model. That is a specification that rewards invention, written by me.

What that produced, counted across the 739 English articles:

• 180 of the 739 English posts cited a source called “Internal Analysis” that does not exist. There never was an internal analysis.
• 470 of the 739 carried a before-and-after metrics table belonging to a company that is never named anywhere in the article.
• 91 of the 739 credited a figure to Forrester without linking or naming a report. 46 did the same to HubSpot.
• 158 of the 739 shipped with a title that stops mid-sentence in an ellipsis, because nothing in the pipeline ever read the output before it went live.
• A first-person account written by a HubSpot writer was republished as if it were mine.
• Case-study style performance metrics about my own work, none of which I can substantiate. I pulled them from the English pages in March 2026, and they survived in the Italian, Arabic and Chinese translations for months after that, which is worse than the original error.

Publishing stopped on 27 April 2026. The archive stayed up until 5 August 2026, when I deleted all 2,922 files. Not noindexed, not archived quietly. Deleted, with the reasoning published. The full account is in the postmortem.

I’m telling you this on the About page rather than burying it in an archive, because the alternative is that you find it yourself and then correctly discount everything else I’ve written.

What I can actually do

Stated as capability, not as results. I don’t have public numbers about my own work and I’m not going to invent any.

• Set up measurement so decisions survive scrutiny. Event schemas, UTM conventions that don’t drift, and a definition of each metric written down before anyone argues about it.
• Run acquisition experiments end to end. Hypothesis, sample, stopping rule, and a written result even when the result is “no effect”.
• Write and diagnose landing pages. Message clarity, objection order, and the gap between what a page claims and what it proves.
• Automate the repetitive parts. Static site generation, scheduled jobs, build-time validation. I’m better at this than most marketers and worse at it than most engineers.
• Run a multilingual site without trusting the translations blindly. This one shipped in four languages, and the failure described above survived longest in the two I don’t read.

How I work: I fix instrumentation before I raise budget. I prefer a small claim I can source to a large one I can’t. I write down what I don’t know, on the page, next to the thing I do know.

The record, with links

• March 2026 to now: Growth Marketer, Yakkyo S.p.A., full-time, hybrid, Italy.
• September 2019 to March 2026: Marketing Specialist, Jumbula, B2B SaaS for registration and class management. Six years and seven months, contract full-time, remote, for a US company. Growth and product marketing end to end: strategy, analytics, experimentation, automation, launches, content.
• 2013 to 2016: Android Developer, Cafe Bazaar, remote. I wrote software before I marketed it, which is the reason the rest of this site exists as a hand-built Hugo project on Cloudflare rather than a page on someone’s platform.
• September 2025 to now: BSc in Computer Science (Data Analysis), Università degli Studi di Messina. In progress, not finished.
• January 2023 to April 2024: Computer Programming, Seneca Polytechnic, Toronto.
• January 2022 to January 2023: English Language and Literature, Seneca Polytechnic, Toronto.
• Certificates: eleven online non-credit courses from UIUC Gies and UVA Darden, each with its Coursera verification link. A verification link is the only part of a certificate worth anything, so that page leads with them.
• Find me: LinkedIn, X, Telegram.
• Languages: Persian, English, Italian. An earlier version of this page claimed five. I cut it to the three I can hold a working meeting in.
• Name: see the note directly below. Three forms, one person.

On my name

Three forms, and they do different jobs.

Mohammad Mahdi RezaAli is my legal name. It is what appears on documents, which is why the Seneca Polytechnic certificate reads that way, and the Coursera certificates print Mohammad RezaAli in capitals. If you are checking one of those against this page, that is the same person.

Mohammad RezaAli is how I introduce myself formally.

Mo RezaAli is what I go by, and it is what I would rather you call me. Mo is fine. So is morezaali, which is how it reads in my email address and my handles.

That is the whole list. Every byline, heading and citation on this site says Mo RezaAli, and a build check refuses to compile the site if any other form appears outside this page and the certificates page. The version of this site that this one replaced drifted across nine spellings and shipped an eighteen-entry list of name variants in its structured data, one of them a typo, which is a different thing entirely: that was an attempt to be found, and this is an attempt to be identified.

What I can’t prove to you

• Anything about my results. No client outcomes, no revenue figures, no before-and-after tables. My employers’ performance data isn’t mine to publish, and unpublishable numbers are indistinguishable from invented ones. So there are none here.
• My enrolment and my employment. Yakkyo’s listing is a matter of public record; my presence inside it is not. Take the role as a self-report and weigh it accordingly.
• That I’ve fixed my judgment. I’ve fixed the pipeline: nothing publishes here without a human writing it and a build step that checks every cited URL still resolves. That’s a structural fix, not a character reference.

Contact

Email: contact@therezaali.com. LinkedIn: linkedin.com/in/mo-rezaali.

If you find something on this site that’s wrong, tell me and I’ll correct it with the date on it.