The fix was a role, not a system

The problem

Stripes & Checks sold fashion online from a site the founders had designed themselves. By the time I started consulting for them in 2022, the site was dated, and underneath it sat a custom system that tied the website to their POS and their warehouse. Changing the front end meant risking everything behind it.

The brief was about growth: bring the site up to date and get more people to it. The number that mattered turned out to be a different one. Orders were taking up to two weeks to reach customers.

What I found when I did the work manually

We started where the brief pointed. The site couldn’t be rebuilt without rebuilding the system behind it, so we gave it a face lift and put the rest of the effort into service. The reasoning was that if the product and the delivery were good, customers would forgive the site’s quirks.

On social, I planned the calendar monthly and ran it with a third-party agency, who produced the posts and handled customer messages on the page. Content was approved with the client two months ahead. We ran Google and social ads alongside the revamp.

Then the digital manager left, and we took over as much of his job as we could until a replacement was hired. That is when the actual problem became visible.

He had been doing five jobs. Every order went through him for approval, and he packed it himself in the warehouse. He also ran the website, made the content and managed the ad boosting. None of it was separated. When the marketing worked, orders queued behind a man who was also writing captions.

I had seen the opposite arrangement before. Earlier in my career I worked in the Odel warehouse and learned pick-and-pack properly, and the main lesson was single-task efficiency. The person packing orders should be packing orders.

So the two-week delay wasn’t a software problem, and a better website wasn’t going to fix it.

What I built

The first thing built was a role, not a system.

We split the digital manager’s job in two. Warehouse and dispatch became one role with one job. Marketing and the website went to a separate team, which we helped the client set up.

For the warehouse hire, the client kept the job description they already had. I helped with sourcing and interviews and screened for someone who had run a warehouse before, knew pick-and-pack methods and could work to standard operating procedures. The aim was a person who could organise the warehouse and make it efficient, not just keep up with the orders.

Only then did the software change. Their dev team connected the custom POS to the delivery companies’ portals, so an order placed on the website could be booked for delivery directly instead of being approved and re-entered by hand. I didn’t write that code. I worked through it with their developers, bouncing ideas and finding workarounds where the custom system made the obvious route difficult.

The order matters. Automating dispatch while one overloaded person still approved every order would only have moved the queue somewhere else. With a dedicated person working to a defined routine, the integration had something stable to connect to.

What changed

Delivery went from up to two weeks to two days. We measured it against order records and the complaints coming in, which were the most reliable signal the business had.

The marketing side was more mixed. The face lift worked in the sense that the site stopped getting in the way. Social media was a mild success, not the lift we had hoped for.

The Australian expansion didn’t happen. The owners had connections there that could have carried a small word-of-mouth launch, so I researched import requirements and put them in touch with Australian logistics and warehousing firms. The numbers didn’t hold up. The upfront cost was too high, and shipping single orders internationally, the way the large marketplaces do, doesn’t make sense for a small operation. The client later moved towards bricks-and-mortar retail.

The result I can point to directly is that orders arrived in two days instead of two weeks, and that the warehouse no longer depended on one person who was also running the marketing.

What I would do differently

I’d start in the warehouse.

We began with the website and social media, because that is what we were brought in to fix. The real bottleneck only showed up when the person absorbing it left. I had the experience to spot it on the first day, having worked a pick-and-pack floor myself. I didn’t go and look, because the problem as described was a marketing problem.

That is the mistake I’d now warn anyone about: taking the client’s description of the problem as the problem. Doing the job before fixing it only helps if you do the whole job, including the parts nobody mentions.

The second change is expectations. We went in hoping marketing and an international launch could move a small, custom-built fashion brand a long way. Neither the brand nor the market supported that. I’d have said so earlier and put the effort where the returns were, which turned out to be operations.

Next step

If orders, tickets or approvals are queuing behind one overloaded person, the fix might be a role before it's a system. Let's talk.

Book a call↗ (opens in a new tab)