I thought launching the store was the hard part

When I started working on EVORA, I thought an online store needed a good-looking website, products, advertising and then orders would follow. The first version taught me how much I was missing. The structure was unclear, the branding was inconsistent, and I had not thought through how someone would discover a product, understand it, trust the store and decide to buy.

Looking back, that first version barely feels like the same business. Building the site was only the beginning. The real learning came from putting it in front of people, seeing where it fell short and making changes.

“You don’t really learn e-commerce by finishing a course or launching a website. You learn it by building, measuring, breaking things, fixing them and doing it again.”

More products did not make a better store

At an earlier stage, EVORA had roughly 2,500 products. It seemed like an advantage: we could offer luggage, backpacks, business bags, travel accessories and more. But that breadth made it harder to answer a basic question: why should someone choose any particular product from us?

The store felt more like a supplier catalogue than a retailer with a point of view. We began reducing and prioritizing the assortment, considering pricing, inventory, positioning and commercial potential instead of importing everything available. I started asking which products deserved a place in the store, and what value we could add between the manufacturer and the customer.

Design became part of the sales process

I used to judge the website mostly by how it looked. Now I ask whether a visitor can understand the offer quickly, find the right product, see shipping and return information, and feel confident that the item is authentic. I also think about the experience on a phone and what might make someone hesitate at checkout.

We reorganized product pages, changed image galleries, improved mobile layouts, experimented with sticky Add to Cart controls and adjusted quantity selectors. We made information about free shipping across Canada, 30-day returns and exchanges, warranties and authenticity easier to find. No single change transformed the business, but the work helped EVORA feel more like a retailer and less like a template.

The funnel made me ask better questions

One historical snapshot of the store’s funnel showed 2,029 sessions, 26 add to carts, 18 checkouts and 2 purchases. The reporting period was not specified. Those figures are not a measure of improvement; they were a prompt to look more closely at where customers were leaving.

Earlier, I might have said we simply needed more traffic. But people were already arriving, and some were starting checkout. I began asking what was stopping them and how the advertisement, product page, price, supplier costs and trust signals worked together. Looking at one number in isolation no longer made sense.

With Meta and Google advertising, I first paid a lot of attention to impressions, clicks and traffic. I learned that those numbers only matter when they lead to a commercially useful outcome. An ad can get attention, bring people to the store and even produce add to carts while still losing money.

Our work with the Rick Ross x Bugatti collection made that lesson concrete. We tested products, creative, formats, headlines and offers, and later used a clearance approach with 40% off and final-stock messaging. That discount belonged to a past campaign; it is not a current offer. The experience pushed me to consider product cost, advertising spend and margin together. Urgency cannot make up for weak positioning or poor economics.

“Revenue and profitability are completely different metrics.”

Creative testing and SEO became more deliberate

I once thought you made a good advertisement and ran it. Now I think in angles. The same backpack might matter to someone as a travel companion, a work bag or an organizational solution. We began making variations in imagery, copy, formats and calls to action, then asking whether a weak result came from the creative, audience, offer, landing page or product economics.

SEO asked me to think further ahead. We built the EVORA Journal and began planning content around search intent, the reader’s stage in the buying journey and the question an article should answer. We also ask whether a product or collection is genuinely relevant to the topic.

The less visible work mattered too. By September 2026, we had worked through technical SEO checks covering metadata, internal links, image alt text, structured data, multilingual pages and author information. An audit found 109 of 109 English URLs returning correctly, complete hreflang implementation, no missing image alt text and no JSON-LD parsing errors. Those are technical checks, not proof of higher rankings or sales.

AI made iteration faster, but judgment still matters

AI is now part of how I think through website architecture, SEO, analytics, advertising concepts, UX, product positioning, code changes and quality assurance. With coding tools, I can describe a problem, implement a change, inspect it and iterate faster than I could before.

But a technically correct change can still look wrong, feel wrong or create a problem somewhere else. AI helps me move from an idea to an implementation; I still have to inspect the result, measure what happens and decide what to correct. Often there is another correction after that.

“AI can build exactly what you ask for. The difficult part is knowing what you should ask it to build.”

What Randmar helped me learn

My experience with Randmar has not been a blueprint of exact steps to follow. Its value for me has been learning to approach the business differently: test instead of assume, measure instead of guess, understand the economics behind a sale and look at the whole customer journey rather than only the homepage.

I have also learned to treat content as an asset and mistakes as useful information. EVORA is not just a Shopify website. Inventory, suppliers, margins, advertising, email, SEO, social channels, customer support, returns, product selection and trust all affect one another. I could not see those connections when I began.

The work I am still doing

EVORA is much better than its first version, but I do not see it as a finished success story. There are conversion problems to solve, products to position more clearly, campaigns that do not perform, pages to improve and analytics I still need to understand. Some experiments will fail.

What has changed is how I respond. When something does not work, I am learning to ask what it taught us and what we should try next. I am glad the first version was imperfect: watching the store evolve one problem at a time has taught me more than a perfect launch could have. I am still building, still measuring and still learning how to make EVORA better.