Sometimes you win, sometimes you lose. This month we let one of our clients depart from MageCloud. I am choosing the words carefully, because “lost a client” is not quite right. We worked together for about two and a half years, stabilised their store, completed their open projects, and delivered a substantial amount of customisation. They came to us after leaving a previous agency, and now it was their time to move on from us. That is the lifecycle of this business, and most departures are exactly that ordinary.
What makes this one worth writing about is why they left, and what it cost me to hold my position on the way out.
MageCloud Departure Note
The Decision We Disagreed On
THE CLIENT’S PLAN
A major replatform to a modern front-end
An upgraded Magento variant with a modern front-end stack. The content team found Magento too complex and wanted faster, more stable operations.
THE NUMBER THAT WORRIED ME
10 non-brand clicks a day
Google Search Console, brand traffic filtered out. The store’s organic acquisition was effectively zero. Marketing, not technology, was the binding constraint.
MY ADVICE
Do not fund the rebuild yet
A full front-end rebuild forces most modules to be redone for compatibility. Spending that budget before fixing acquisition struck me as backwards. I said so. I did not win.
Paul Ryazanov · MageCloud · proud of the advice, sorry about the outcome
The Question I Asked Myself Before Advising Them
The client’s frustration was legitimate, and I want to be fair to it. Their content management team genuinely struggled with Magento’s complexity. Day-to-day operations were slower than they wanted, and someone credible had shown them a more modern setup. The desire was not irrational. The sequencing was the problem.
Before I gave them my recommendation, I did what I always do: I opened Google Search Console, filtered out every brand-related keyword, and looked at what was left. Ten clicks a day. That number is the whole story. A store whose non-brand organic traffic rounds to zero does not have a technology problem at the front of the queue. It has an acquisition problem, and I asked myself the only question that matters in that situation: if this were my money, would I invest in a major redevelopment right now, knowing marketing is the key? I answered it honestly. No.
A complete front-end rebuild on that platform forces most existing modules to be redone for compatibility. It is months of budget and engineering risk in exchange for operational comfort, while the actual growth constraint sits untouched. I have watched replatforming projects consume the budget that should have gone into the work that follows them, and this had the same shape.
I Could Not Convince Them, and They Left
I made the case. I showed the numbers. I explained the compatibility cascade and what the same budget could do pointed at acquisition. The client heard all of it and chose the rebuild anyway, with another provider attached to the new technology. I was unable to persuade them, and I still believe the decision was wrong for where their business actually was.
There is a version of this story where we keep the client. It is not complicated. We express enthusiasm for the new direction, position ourselves for a piece of the rebuild, and ride a large project for a year. Plenty of agencies would call that account management. I understand the logic and I am not interested. An engagement that begins with me endorsing a plan I believe is a mistake is an engagement where my advice has already become worthless, and advice is the actual product. The same principle sits underneath why I once cut a client’s invoice in half: the relationship only works if the numbers and the words stay honest, especially when honesty is expensive.
The Reminder I Set in My Calendar
I set a reminder to reach out to them in four months and ask about the return on the investment. I want to be clear about the spirit of it. It is not a told-you-so in waiting. If the project succeeds, I will be glad, I will say so, and I will have learned something about a case where my framework missed. Unlike many other times when a customer went a different way, I genuinely hope this is the success story. Their team deserves a win, whoever delivers it.
But the follow-up matters either way, because advice that never gets checked against outcomes is just opinion with confidence. I keep these reminders for the same reason I keep the Search Console screenshots. The feedback loop is how the next client gets better advice.
What Standing By Your Opinion Buys an Agency
The good news inside every departure is capacity. We now have an additional slot open, and it will go to a store we can be useful to. But the bigger asset from this episode is not the slot. It is the precedent, internal as much as external. My team watched us deliver an unwelcome recommendation, hold it under pressure, and accept the commercial consequence. That is the only way an engagement model built on trust instead of contracts stays credible. Clients stay with us for years without a contract precisely because they know we would rather lose the engagement than manage them into bad decisions.
Our doors remain open to this client at any time, and I mean that without an asterisk. If the new platform thrives, I will be reading their case study. If it does not, there will be no lecture waiting, just the same advice, priced the same way.
If You Are the Client in This Story
If you are an ecommerce founder weighing a major rebuild right now, run the one check before you sign anything: open Search Console, filter out your brand terms, and look at what acquisition is left. If the number is small, the rebuild is probably not your next investment, whatever the demo looked like. And if you want someone to walk through that data with you before you commit the budget, get in touch. I will tell you the truth, even though I now have the scar that proves what it occasionally costs.