The many hats of an Ecommerce Solopreneur

During the pandemic I, like countless others, found myself in a strange situation and was enticed by the thought of starting a side hustle.

At this time the amount of side hustle ideas bandied around the internet was booming.

TikTok was alight with young influencers boasting about their monetary gains from day trading, or dropshipping, or faceless YouTube accounts, or some other spurious notion that didn’t seem quite legit.

I’m no stranger to running my own business. I started up as a freelance designer whilst in my last year of uni - class of 2000. And I kept that going for a good decade as my side hustle.

Those were the golden days of freelancing. I could design, code, and host a basic website. Back then that’s all my clients wanted and it was a wonderful way to make some cash on the side.

Back to 2021 - now the ideal side hustle wasn’t based on selling my time for money. The pandemic made this kind of business model challenging as our kids were at home more and my time was in short supply.

I needed something that used my skillset, that could start small, and could be scaled if successful - just to keep it interesting.

The business model that stood out to me was ecommerce as it seemed to fit my criteria:

• It requires broad use of my skills in web design, product design, and marketing
• Many online shops focus on one product - they can really be that small to begin with
• Revenues are gained by the number of products sold, which can scale well if the shop becomes popular

And so it began. The business, plus an experience that taught me some new skills and some harsh truths.

A hat for each department

The premise of running a simple online shop is, in reality, far from simple.

Every time I needed to do a task, I imagined putting on an imaginary hat and got to work.

I also set up a Notion database to manage activities in each department.

• Research
• Design
• Engineering
• Marketing

I actually had a couple of extra cards, Planning to capture brain dumps, and a Finance card to help me work out all the economics. These weren’t juicy enough to write about here so let’s concentrate on the other four.

Research - Finding the balance between economics and ethics

Within the scope of an ecom model, I needed to figure out how to make this a viable business.

What kind of products could I sell?

• Producing something myself was an interesting direction, but it does require a fair amount of upfront time and cost, which is risky given that I was trying ecom for the first time.
• Re-selling other people’s products felt like more of an experiment - fast to set up and to pivot

There are a few different approaches to selling other people’s stuff, again each come with pros & cons.

Affiliate marketing

Affiliate marketing means promoting products clearly as sold by others.

I’ve dabbled in this before and got nowhere. It seems to be a volume game.

If you want to make decent money here you need to either litter the internet with thousands of your links & hope people trip over them, or build up a large following of real people that trust your brand & will buy things because you say they’re worth spending money on.

It needs this kind of volume because the commission per sale often very low.

Example: Amazon will pay you 1% - 8%. So unless you’re selling high ticket items, you’re making pennies of revenue.

I’m oversimplifying… there are likely people making a success of this model out there. But I’m neither a litterer nor a cult leader so it was a no from me for this approach.

Dropshipping

With dropshipping you become the retailer of often cheaply manufactured products from China.

The lure of this approach is that you don’t have the cost of self-managing logistics: Buying stock in bulk, storing them, or distributing them around the world.

Here you’re simply waiting for a customer to order a product from your web page, and then placing an order for it to be shipped directly form China to the customer’s home. You never actually see the product.

If you can identify a decent product that your target market will actually buy then profit margins can be good.

Example: An unbranded Levitating Moon Lamp costs £5 to buy, plus £10 shipping. You sell for £50 and bank £35. A decent markup.

I avoided dropshipping for the following reasons:

• Finding and validating a single product can take several attempts and cost you startup capital
• The barrier to entry is low which makes this space is highly competitive
• Shipping from China takes 3-4 weeks, which customers hate
• The products are of low quality, often creating high return rates
• It all just feels a bit wrong to be that guy selling snake oil to unsuspecting citizens

Print on Demand (POD)

This is the approach I did take.

In many ways it’s no different from dropshipping. A production partner to makes the thing, while you’re busy marketing the thing as a retailer.

The added part of the approach is that you get to customise white-label products with your own design.

The there are a few key differences that made this a appealing to me:

• I found production partners based in North America - far more ethical and with faster shipping times
• Their products were already validated and selling - this cuts out a big chunk of the risk
• While still a crowded space, custom designs can differentiate your own products to the point where it successfully beats your competition - my background in design was well suited to this challenge

So, what products did I sell?

Initially I spent some time thinking about products in the areas I personally find interesting: music, cycling, cats etc.

These are all valid niches, with paying customers. But with POD, niches work in different ways. You can take any item, no matter how bland, and customise it for your target niche.

Example: T-shirts with custom graphics relating to different school years, marketed at teachers who run group activities with kids. A clear use case with a niche in mind.

I chose two types of product, and sold them across two separate periods of time:

• Dress jewellery in presentation boxes
• Lazer-cut metal signs

Both were available from North American production partners, and had good potential profit margins given the high perceived value of these items.

The metal signs were relatively new to the POD world and so had lower competition that other products. If I acted fast, I could capitalise on that.

Design - Aiming for instant emotional appeal

It’s the customisation of POD products that makes them appealing to sell. My background is in design so this part was fairly simple.

Designing the product

Let’s talk about how I tackled the jewellery products.

The customisable part was a piece of card shown behind the piece itself.

They’re primarily bought as a gift, and so that card would typically have a message from gifter to recipient, presented in a visual style of some sort.

It’s the message that my competitors obsessed over - finding a phrase that evoked enough emotional sentiment in gifters that they would like to present that message to someone alongside some jewellery.

To determine what words & phrases resonate with people, I decided upon a research method based on social proof.

I searched Instagram for posts with a quote or phrase that seemed relevant to some niche, and logged them all in a Notion database alongside the number of likes & comments they had at that point in time.

This quickly amassed hundreds of phrases to work with:

• I filtered by those that I felt may not be legal to use such as song lyrics or book quotes
• I tagged them by relevance to some gift-giving event like Xmas, Birthdays, Valentine’s Day, Mother’s Day etc.
• I ranked the remaining phrases by social proof metrics & chose the most successful for the next stage

By combining these phrases together I shortlisted a handful of phrase permutations that read well together, applied them to a few different visual styles.

This gave me artwork for the my production partner and, more importantly, some options to test.

Qualitative testing isn’t the right method for this, I needed higher volume of feedback, and so I took the lean UX approach of throwing out options to the real world & measured what happened. More on that in the Marketing section of this article.

Designing the brand

With products designed, I now needed to expand their visual styling into some kind of visual identity and apply it to other assets that running an ecom business needs.

The products would be seen in marketing artwork (I just can’t bring myself to erroneously use the word ‘creative’ to describe what is a piece of artwork) and on a website where customers could buy them.

All of that needs to look consistent enough to communicate that this is a trustworthy brand to buy from.

Another part of the brand worth thinking about is the shop name. I have no formal training nor experience in this kind of exercise (other than suggesting the business name for Life Moments). All I knew was that I had designed products which seemed relevant for family members to buy to each other on certain occasions.

And so I simply thought up a list of 40-ish ideas inspired by random meanderings on the internet, rendered them in a suitable font, and picked my favourite.

Quite likely a flawed approach here but I highly suspect that so long as your shop name is somewhat relevant and sounds trustworthy or relatable then that’ll do just fine. No need to over egg that pudding.

It was time to put on the engineering hat and set up the website.

Engineering - Saving time with pre-built software

I knew that ecommerce tech would be a huge challenge to code, and so licensing an off-the-shelf checkout process was essential. I looked at several to compare:

Shopify

The most obvious option. Shopify was apparently easy to set up, with lots of front end themes to choose from, and decent analytics to track user behaviour. Small downside was that it’s the most expensive at £20 per month.

Wordpress & Woocommerce

I’m well-versed with building Wordpress themes so this felt like a no brainer. However, upon closer inspection I wasn’t keen on working with the Woocommerce plugin - reviews stated it was a technical challenge to set up correctly and I was more keen to get something live on the internet to learn from real data.

Big Cartel

A bit of a strange choice but given that it’s free to use & didn’t require much setup I thought I’d explore it. Fast and cheap it is, but Big Cartel was simply not fit for purpose when doing anything more than simply publishing web pages: no analytics, a cap on the number of products, and far too limited branding control. Although perhaps it’s matured since then.

Website setup

I did look at a handful of other options, but chose Shopify in the end as my production partners had integrations with them and offered help on how to set it up.

These integrations help automate the creation of product listings. I was aiming for operational speed to experiment with different designs here, so integrations were a winning factor.

Building a Shopify site isn’t particularly hard. Mostly due to the theme store.

Here merchants can pick one of hundreds of themes that come with predefined visual styles and some specific features that may be more relevant to the kind of products you’re selling.

Many themes cost between £200-£300, but Shopify offer free themes to get new users started with no cost barrier.

I chose their free theme named Dawn for two reasons:

• It used the latest theme tech that included drag & drop of content sections onto any page
• I discovered a way to inject custom CSS into themes so I could totally override the existing visual design with just this feature

After a few new graphics and some hand-written CSS code, I had the basics of a shop styled in a way that looked like a brand.

I actually didn’t change much of the default styling, it was so generic that just by changing the font & adding a sprinkle of more specific colouring the whole site quickly looked unique enough.

‘Enough’ was fine for me, as I was chomping at the bit by now to load in some real products and see it all come to life.

Connecting products

The production partners I chose had built apps that were available in the Shopify app store.

For each, I installed their app, and set up a list of products to sell. It’s a clever system that does two things at once:

• Publishes a product with your chosen wording & promo images onto the site
• Connects that product listing to the production partner’s backend

This means that every time a customer buys a product, the site automatically sends a request to your production partner to create that product using the official artwork you’ve associated with that product listing, and post it directly to the customer.

With dress jewellery your design is printed on card and placed into the box. With metal signs your artwork is used to guide the lazer that cuts the actual product.

As a merchant you have no involvement with this process, which is exactly what you need in the case of high volumes of sales - the ideal situation to find yourself in 😉

Legal pages

Every ecom shop needs to declare its terms, conditions, and policies.

This is important as a retailer. Real people will be spending real money on real products through the store and it won’t always be plain sailing.

I found an online tool which writes a Privacy Policy for you based on a few key inputs. This was a good starting point and certainly enabled me to tick this box faster than figuring it out myself.

If done well, these pages can do more than simply cover legal challenges. They’re also an opportunity to extend the brand tone of voice and position answers to frequently asked questions where people expect to find them.

It’s the boring stuff, but may just tip the balance between a visitor & a customer.

Analytics

Shopify handle the payment as part of their platform. They offer analytics screens to help you track how many customers have spent how much, on which products etc.

The analytics screens are close to ideal. Shopify is a platform entirely focussed on ecommerce sites so they know exactly what metrics are important to retailers and they present those in an intuitive way.

There’s even a mobile app with a widget to show when sales drop in. More on that in the marketing section.

Like many website builder platforms, it’s easy to integrate Google Analytics by pasting a code snippet into an input box somewhere deep in the site’s settings.

Shopify also plays very well with social media metrics. You can connect various accounts to your store to help make sense of the customer data.

Setting up the Facebook Pixel was somewhat more challenging (to put it mildly) but Meta’s tech changes so rapidly that there’s no point in critiquing its usability. If you’re going to make use of it, just dive into the most recent YouTube tutorial on the subject.

Marketing - Where the real hustle reveals itself

So there I was. Online store operational and shiny, products ready to be shipped but, as the very successful ecom store is aptly titled: Who Gives A Crap?

I needed visitors to convert into customers, and there are two ways to find them:

• Free traffic
• Paid traffic

Free traffic

By far the most important way to gain free traffic is Search Engine Optimisation.

If someone finds your store through searching Google for a product then it’s likely they have at least some intent to buy. That’s traffic worth gaining.

I’d already set up sensible SEO metadata when building the store. Beyond that it’s all about writing regular articles and gaining backlinks to improve authority.

Posting regularly on social media is the other main approach to gaining free traffic. TikTok changed the world here and provided opportunities for huge exposure.

I searched for example of how POD stores were using video content to promote their store, but all I found were cringe adverts with scathing comments - not a world I was keen to join.

Successful TikTok videos have a vibe that the content is genuine. And I’d seen some candlestick makers or screen printers etc. documenting their production process to great effect.

However, I wasn’t making any of this myself, and so to even try using the medium of video felt like I’d most likely fall into the cringe club.

Social media viewers also don’t necessarily have any intent to buy whatsoever. So only a small proportion of those who engage with your content are comparable to those searching Google for the kinds of products you sell.

Free traffic is most often a slow approach. I wanted to rapidly experiment first, and that costs money.

Paid traffic

Most paid traffic comes in the form of hiring an influencer or running your own advertising campaigns.

The influencer effect is best left for when you know you’ve figured out product market fit and want to scale up. I was far too early in the journey for this approach so, advertising it was.

The first question I asked myself was where to advertise?

• Google ads - Known to be expensive but, as previously stated, searchers have high intent to purchase. I considered this but given the visual nature of my products I decided against for now.
• TikTok - Audience is generally too young to be the kind of person buying jewellery as a gift, or metal signs for their home. A no from me.
• Instagram - Average audience age is slightly older, has a creative vibe. Another one to consider.
• Facebook - An older audience still, the best fit demographic for my products of choice. This is where I went to advertise.

I’d never advertised before so a quick bit of digging on YouTube unlocked all the how to knowledge I needed to get going.

I set up a business account under my personal FaceBook account and used what they called their Business Manager’ – an overly generic name for a product that only does advertising.

At this point it’s not that hard to just whack an ad onto FaceBook and call yourself a marketer. But it’s all too easy to spend more than you need on ads that produce no results.

You need a strategy.

There are three main areas to get right when advertising on FaceBook, I thought of them like a funnel:

• Target audience
• Ad copy and artwork
• Tracking

Target audience

Of all the 8Bn people in the world, I’d chosen Facebook users to see my ads.

That narrowed it down about 3Bn – Progress!

FaceBook has an audience tool to target showing your ads to people of different locations, speaking specific languages, age ranges, genders and, most importantly, who have shown interest in certain topics.

It’s easy to zero in on specific groups by adding more and more filters. FaceBook will even instantly show an estimate of how many people are in your target audience - it’s powerful stuff.

However, the overwhelming advice is to leave your audience much broader than you initially think, with at least 100k people in the audience, and then let FaceBook’s machine learning figure out the pattern of which kind of people are likely to buy.

So why target at all? Because the larger your audience, the more you’ll need to spend to train the ML model.

I set up a few ads with a different audience interests to differentiate them, leaving all other targeting as the controlled elements.

Ad copy and artwork

Remember in the product design phase I was aiming for an instant emotional response?

That response happens at the moment someone sees your ad. It’s a critical moment to design for as this opportunity can be over in seconds.

When I design a new UI component I don’t draw it on its own and hope it works. I like to place it inside the layout, in the position it will live, and assess it in context.

For ads, I went one step further and built a single screen prototype that enabled me to scroll through a mock FaceBook interface and see my ad ideas appearing on the screen amongst other normal posts from real users.

Once my prototype was set up it was quick and easy to swap out different ad options and experience them as a real mobile user would.

The interaction that’s happening here is called a glance. I wanted to dissect which parts I noticed first in this glance to inform my ad design.

Once again, this kind of sequential, funnel-like analysis gets deeper into the detail of the user’s experience:

• At first my eyes tend to look at the profile picture - high contrast ones seemed to be most arresting as they’re quite small and could fight for attention.
• The image is my next area of focus. I scrutinise what it is for a few seconds, it’s obvious this is a product for sale. What I need users to feel is that this product could help them solve a problem eg. A happy partner through finding a great gift. The product image also needs to deliver on the promise of this solution.
• Then I figure out that it must be an ad & I read the copy above it. A simple list of benefits is appreciated if I like the product image.
• By this point I’ve pretty much made up my mind whether I’m going to click it, or scroll past

All that in a few seconds. It’s worth dissecting. For example, there’s a space for live text above the image but in my opinion that’s the weakest part of the whole ad.

Thinking of this like a funnel, it’s best to get the elements highest up the funnel optimised first so you have more chance of people continuing through to the next part and learning if you’ve got it right or not.

The final part to consider in the design is the journey people who click take when from the ad to your landing page.

Of course, I used my specific product pages on my site as the landing pages for each ad. People are goal-driven when buying online and expect the fastest possible journey.

What’s critical to understand in this moment of navigation is that you’ve promised something on the ad, and now the landing page needs to deliver on that promise in order to convert an interested visitor into a paying customer.

Assuming your imagery is consistent between ad and landing page, this is an exercise in copywriting. I even needed to double check my spelling as I was targeting American audiences so words like ‘jewelry’ are different to how it’s spelt in the UK.

Each of my ads suggested a solution to a problem, and listed a few benefits. So I wrote the landing page copy to expand upon those benefits in more detail, making sure that the text flowed on naturally from ad to page.

Once set up, you choose a budget and start your campaign.

Tracking

Having live ad campaigns is the really exciting time about running an ecom business. I was constantly refreshing FaceBook Business Manager to see how things were going.

There are lots of metrics available, but I focussed on these:

• CPM - Cost Per Mille. This is the spurious cost that FaceBook charges you for every 1,000 Impressions (people it shows your ad to). I say spurious because it changes like the wind. Different with every ad, at different time periods, for different audiences. It’s one to watch early on to see if it’s even worth completing your campaign. If the cost looks too high, best to cancel & choose another audience.
• CTR - Click Through Rate. The percentage of people who saw you ad and actually clicked it. This figure tells you how well you’ve designed your ad for the target audience. It sounds incredibly low but, for ecommerce, 1-3% is considered a successful ad.
• CR - Conversion Rate - Connecting your FaceBook Pixel tracking to your Shopify store means the number of sales made can be pulled directly into your FaceBook metrics so it’s easy to check on the general success of a campaign.
• ROAS - Return On Ad Spend. The Return being total Revenue attributed to that campaign. It’s an important number to monitor as if the Return is consistently lower than what you’re spending that shows the campaign is currently failing. It may improve over time, in which case just keep running the ad. That’s your call as Marketer.

When I say the metrics were exciting to watch, I think that’s because advertising, using my own hard-earned cash, seems no different than gambling. There’s a strong element of risk and reward going on here.

I would leave my campaigns running for a few days and in the evenings, when America had woken up, I’d wait for my Shopify iOS widget to notify me of each sale.

After all this research, design, website build and marketing setup, every ping on my phone brought a sense of euphoria that a real person had found value in one of my creations.

Then I’d check the campaign metrics and see, on occasion, the CPM had risen to levels so high that I had to shut it down early.

Ultimately, you don’t really know for sure what profit you’ve made until the returns period has passed for each sale.

I kept a tidy Notion database of all these metrics in my Finance card in order to inform my next steps. Otherwise it all feels slightly disconnected.

The Shopify App Store is a dark place

Once my stores had been running for a good few months, I was looking for more ways to optimise.

At one point I ended up on the Shopify App store having learnt about how certain apps claim to improve conversion rates.

It may sound dramatic, but this app store is a slippery slope into the dark side of consumer psychology.

Much of the advice to ecom store owners focusses on deceitful design tricks that lure people into clicking the buy button:

• Every store has a permanent sale. First-time visitors don’t know this, and so it seems attractive…
• Announcing that the perma-sale will end soon, even adding a countdown timer to that moment in the future when prices will suddenly increase and the chance to purchase at today’s price will be lost forever…
• Sticky ‘Buy now’ buttons that remain on screen regardless of where users have scrolled to…
• Fake user reviews aiming to prove the product is loved by those who buy it…

The list of dark patterns is endless, and I didn’t want to employ any of these tactics.

As a person with solid history in user experience, I can think of a counter argument to every one of them. And that’s exactly what I did.

Some may say that it doesn’t matter if you agree with these tactics or not. That the proof is in the conversion rate. And that’s partly true.

I definitely wanted to make my stores profitable, but only in an ethical way. That was more important to me than making bogus claims or the animating the buy now button every 10 seconds…

How it all ended

After a loss-making start, I was able to make the venture profitable with one complete rebrand and an improvement in my campaign running skills.

After lots of experimentation with different products, ads, and audiences, I had two specific products that were selling well.

The next logical step was to scale up the operation, but something held me back.

This thought that advertising is gambling felt all too real, and to increase the risk was worrying at that moment in my life.

Privacy laws for consumers changed, reducing the efficacy of FaceBook targeting, which meant CPM costs rose, therefore making it harder for ecom shops to grow.

I’d managed to develop one store into the profitable side of the business, but wasn’t making an exciting amount of money. So I began to question whether all the hard work it takes to keep this cycle of activities going was really worthwhile.

The real value had been in everything I’d learnt along the way.

My third child arrived, once again my wife and I faced becoming sleep-deprived for months on end, and neurodiversities began to emerge in the elder brothers.

At this time I wanted to focus on my kids and my work at Life Moments.

So I took what had been a fantastic journey as a solopreneur, kept all the notes I’d made along the way, and have only just gotten around to telling the tale in this article.

If you’ve got this far, thank you for reading and well done for sticking with what has become a ~5,000 word piece of writing.

Once family life becomes more manageable I’ll be venturing back into solopreneurship waters once again.

I’ve got this idea for a new SaaS product…