Dimitris Panides / Ecommerce work in Cyprus
I build and run online stores in Cyprus.
Most people who build you a store have never had to live with one. I run a storefront that has taken more than 10 million euros since 2020 and grows about 20% a year. I take on a small number of outside projects where that is the point.
What I take on.
Conversion work on a store that already trades
The traffic is arriving and not enough of it is buying. I go through the journey end to end, find where orders leak, and fix the parts that cost you money: category structure, product pages, the path to checkout, the merchandising decisions behind them.
Storefront pages that ship
Homepages, category hubs, campaign and vendor pages, built as self-contained blocks that paste into your CMS without reaching into the rest of the site. Shopify, WooCommerce, Magento or custom.
English, Greek and Russian
I work in all three, which in Cyprus decides how much of the market a store can actually talk to. Copy is tested at real length before it ships, because Greek runs about a quarter longer than English.
The operation behind the storefront
Product data at volume, promotions that do not break at the edges, and the fulfilment path that decides whether an order actually reaches the customer. I have uploaded ten thousand products and shipped fifty thousand orders, so this is the part I know is never as simple as the storefront makes it look.
Performance advertising on Google and Meta
Campaigns built and run against the store's real numbers rather than platform-reported ones, so the spend is judged on orders and margin instead of clicks. I would rather run this after the storefront converts, for the reason below.
Recent work on the storefront.
Real surfaces, not mockups, from the storefront I run day to day.
Rebuilding the front door.
The homepage was one hero image doing the job of a commercial front page. I restructured it into a shopping surface: a campaign strip, a deal grid led by price, a category wall, and department tiles, ordered the way a customer actually scans a page rather than the way a brand likes to present itself.
The same weeks covered navigation, a vendor landing page for the Apple range that routes buyers to a real person instead of implying a checkout we do not have, and the seasonal campaign blocks, shipped in English and Greek.
Opening status, computed live
Store pages listed opening hours and left the visitor to work out whether that meant "now". This reads each location's own weekly schedule and answers the only question being asked. It runs in the browser, resolves the time in Asia/Nicosia rather than the visitor's timezone, and needs no server. Below is the real logic, running now.
Four locations, two different weekly schedules. The service centre opens half an hour earlier on Wednesday and closes on Saturday, which is exactly the sort of thing a static list of hours gets wrong.
Contrast measured, not assumed
Checking every text pair against WCAG AA surfaced the brand orange running at 2.2:1 on white. Text moved to a darker orange at 7.2:1; the fills stayed on brand.
Greek tested, not hoped for
Greek copy runs about a quarter longer than English. Every block was stress-tested at that length for overflow before it shipped.
Nothing leaks
All styles scoped to the block, with no unscoped selectors, so pasted custom code cannot reach into the storefront's own CSS and break it.
No claims we cannot stand behind
Stock wording, lead times and what the site can genuinely do were verified against the business before any of that copy went live.
Not only storefronts.
AMB Beauty
A separate project of mine, and not a store: a Greek-language site for a group of medical aesthetics clinics. Worth including because most of what I am asked for is not a storefront either. It is a business that needs to say what it does, in the language its customers use, and make the next step obvious.
Traffic is the last problem to solve.
Most stores that are not growing try to buy their way out of it. The order matters more than the budget, and it goes like this.
-
01
Make the store convert
Sending paid traffic into a storefront that leaks is paying to lose money faster. Before a cent goes to Google or Meta, the journey has to hold: findable categories, product pages that answer the question, a checkout that does not lose people at step three.
-
02
Then buy the traffic
Once the store converts, advertising stops being a gamble and starts being arithmetic. I build and run Google and Meta performance campaigns judged on orders and margin, not on impressions, because the platform is not a neutral witness to its own performance.
-
03
Then make the operation carry it
Volume breaks whatever was already fragile. Product data, promotion logic and the fulfilment path all have to survive the traffic you just bought, or the growth turns into refunds and support load.
An agency will happily start at step two. That is the difference: I have had to live with steps one and three afterwards.
What the channel has done since 2020.
Totals for the storefront I run, and the rate it has grown at.
- In online sales
- More than 10 million euros
- Sales growth, year on year
- 20 percent, every year
- Online orders
- More than 50,000
- Customers served
- More than 20,000
- Products uploaded
- More than 10,000
Questions people actually ask.
Do you build the store, or only advise?
Both, but the useful version is usually building. I ship storefront pages as self-contained blocks that paste into your CMS, and I do the conversion and merchandising work behind them. Advice on its own tends to sit in a document.
Which platform do you work on?
Any of them. Shopify, WooCommerce, Magento or a custom storefront. The commercial problem is the same in all of them and the platform is a detail; I run a custom storefront day to day, so I am used to working inside what a business already has rather than insisting it moves.
What languages do you work in?
English, Greek and Russian. In Cyprus retail that third one matters more than people expect, and it is the reason a chunk of demand goes unserved. I also test copy at real length before it ships: Greek runs about a quarter longer than English, which is how layouts that were only ever checked in English end up broken.
What does it cost?
Scoped per project, once I know what state the store is in. Email me what you are dealing with and I will tell you whether it is worth paying for. Sometimes the answer is that your problem is not the website.
How do you know if it worked?
We agree what we are measuring before anything changes, and it is orders, conversion rate and margin, not clicks or impressions. If a change does not move those, I will tell you it did not. A flat result you can trust is worth more than a flattering one you cannot.
Do you run Google and Meta ads?
Yes, and I would rather run them second. Paid traffic into a store that does not convert loses money faster, which is why the order above matters. Once the storefront holds up, I build and run performance campaigns judged on orders and margin rather than on clicks.
Where are you based?
Nicosia. I work across Cyprus and remotely.
Tell me what the store is doing.
What it sells, roughly what it takes, and the part that is not working. That is enough for me to tell you whether I can help.
I read everything myself and reply within a day.
Prefer LinkedIn? Message me there.