I published my last essay on this website five years ago. What have I been doing all this time? Having kids!
I’m now the proud Dad of a four year old boy and a two year old girl. They’re very cute and very defiant. We’re working on that.
I launched this website to share my work and to point out weird patterns.
Though I haven’t published anything new in these past several years, I have been secretly updating the Hand Dies First post. I’ve even been able to add a subgenre to the list: time travel movies!
As for my work: we’ve A/B tested our Custom Shirts page many more times. I’ll be honest, the version I featured in my original blog post was hard to beat. It took several attempts.
We’ve also been doing a lot more Facebook advertising, which renders obsolete my post titled Why Facebook Fails. In my defense, the Facebook Ads platform has evolved in leaps and bounds over the past five years. Remarketing lists. Lookalike audiences. Dark posts. A million types of targeting.
Facebook is a still a dangerous platform for beginners. If you run a decent ecommerce site and if you install the Facebook tracking pixel and if you run any ad targeted toward conversions… then you will see an unbelievable conversion rate. Literally. I mean you should not believe it.
Your first impulse might be to congratulate yourself. But can you really trust a $2 CPA? Is that even possible?
If you’ve installed the tracking pixel, Facebook knows exactly who is about to checkout on your site. If any of those people are also inside the target audience for your ad, then all Facebook has to do is “show” the ad in that user’s feed. By default, Facebook will count “view-through conversions”.
At first glance, it looks like your ad performance is beautiful. But dig deeper. Did those users even see your ad? Or did they just scroll right past it? Either way, Facebook will claim credit for that conversion when they eventually checkout. They were checking out anyway and as I said earlier, Facebook knows that.
To Facebook’s credit, they also give you the tools to combat this. Just like adding negative keywords to your Adwords campaigns, so too must you exclude certain audiences from your Facebook campaigns. For example, we have audiences of 7 day, 30 day, 90 day, and 180 day visitors. If we’re targeting cold or warm traffic, we excluded these audiences from our ads, which unskews our conversion data.
Amazon
We knew that more and more customers were starting their shopping experience on Amazon rather than Google. But we also knew that it would take a lot of development resources to sell directly on Amazon. After all, we print on demand. All designs are created within our interactive design center. That design center is a fundamental part of our organization and it doesn’t exist on Amazon.
Do you remember Amazon Product Ads? You could upload your product feed to Amazon, your products would appear in search results, and you would pay per click, just like Google. The click would actually take the customer away from Amazon and onto your site.
In early 2015, our Amazon Product Ads were buggy. Half our feed would be rejected. Traffic would drop off completely. We were getting an error message that said:
We believe that you have violated our policies by attempting to redirect Amazon customers to another website.
Uhhh, yes. That is exactly what we were doing. Because that is the entire point of your platform, Amazon Product Ads.
The writing was on the wall. Amazon wanted to kill Product Ads, but they didn’t want to say that they were killing Product Ads.
The only choice was to sell on Amazon directly. We couldn’t wait for the development resources, so we built our own specialized Amazon product feeds by hand and we placed every order by hand. Within a few months, it was clear that this was a successful channel. We got the development resources. We started automating all processes.
Sidenote: when I say “we”, I mean my team. The Amazon project succeeded due to the hard work of a lot of people, but mostly Jason Kienbaum.
Unfortunately, it wasn’t all sunshine and rainbows.
In the spring of 2016, the counterfeiters arrived. They attached themselves to our listings. They undercut our prices. When an order was placed, they would take our thumbnail image, blow it up, use that as the print file, make a terrible product, and ship it out.
All of our designs were created by our designers in-house. We printed everything in-house. When we uploaded our feeds to Amazon, the brand column was filled with our brands: Customized Girl, Bridal Party Tees, and FunnyShirts.org.
Now there were terrible reviews popping up on our products, under our brand. They said things like:
- Shirt is from some Chinese company and the writing is blurry.
- I opened it and now my entire office smells like vinegar.
- I never got anything.
Re-read that last bullet point. Some of these “counterfeiters” weren’t even counterfeiting. They were literally just taking your money and not shipping anything. If you visited their seller profile, it would just be ten pages of reviews saying “I placed my order and I got a fake tracking number. Never received the package.”
We contacted our Amazon reps. Seems like an open and shut case, no? Sadly, no.
Somehow, it was our word against the “counterfeiters”. For Amazon, it was possible that these other sellers were buying from us, and re-selling our product.
But we don’t sell wholesale. No one is buying our product at retail for $25 and selling it on Amazon for $10. These other sellers provided zero proof that they had actually placed an order from us.
Instead, the burden of proof was on us. We had to buy the counterfeit goods and, in the event that we actually received the product, we had to take photos, document all the ways in which the item was clearly a knock-off, and send the reports to Amazon’s copyright department.
We tried doing this. It didn’t work. Here is why. By the time we even got the goods, that seller was already banned. Every time. Amazon has strict rules about “on time delivery percentages” and “customer satisfaction ratings”. These counterfeiters were producing garbage and their customers hated them. After all, they’d been duped.
In theory, this would be great for us. Counterfeiters were getting busted! Yay! But it didn’t matter. The sellers were using software to automate the creation of new seller accounts. The seller names were essentially a random string of alpha-numeric digits. They had thousands of accounts. As soon as a hundred got killed, no big deal, they moved on to the next hundred.
With the burden of proof on us, it was impossible to keep pace.
There had to be another way. At this point, we were selling custom products on Amazon. You could enter your own text, and we would print it and ship it. With custom products, Amazon recognizes that you might be the only company that can produce the customized print you promise. Other sellers cannot attach to those listings. So we knew Amazon had methods of preventing these attachments.
It was clear that our product pages should be ours and ours alone. No other sellers should be able to attach to our listings.
I talked to our reps. I made it clear that I was going to email them every week with one of two potential outcomes:
A) Our account would be freed from other sellers.
B) I would die.
So I emailed them. Every week. For months and months. And eventually… it worked. We found someone to help us. And that person maybe just saved the livelihoods of the 85 people who work at eRetailing.
Storefronts
One of my early essays on this site was titled Deep Thoughts on the Print-on-Demand Industry.
In that post, I compared CafePress and Zazzle to what we were doing at eRetailing.
CafePress and Zazzle were the leaders of the POD storefront industry. On those sites, anyone could open a store, post designs for sale, and earn a royalty when those designs sold.
They excelled at building a platform. But they did not excel at customization. Those storefront designs were basically static images that users had created in photoshop and uploaded to the site. The designs themselves weren’t truly customizable.
We did excel at that. And we continued to launch new businesses aimed at specific niches, offering customization: Bridal Party Tees and FunnyShirts.org. (Example design: Buy Me a Shot, I’m Tying the Knot: [Name Here]’s Bachelorette Party.)
It seemed to me, the future was a combination of both business models.
The ultimate POD website would be a platform where you could create your own custom apparel business aimed at your own favorite niche, while offering genuine customization.
Well, we built it. We launched a storefront platform on each of our three sites in 2014.
We were trying to create a platform where anyone could join up and launch their own Bridal Party Tees.
Guess what? That’s exactly what happened.
The best example might be a storefront on Customized Girl called Mom Means Business.
Just like Bridal Party Tees, she has a very specific audience: sports moms.
And just like Bridal Party Tees, she sells truly customizable designs. Usually they have a cute and clever graphic on the front and on the back it will say: MOM OF 00. You customize the jersey number to match your son or daughter.
This has been extremely exciting for me. I had spent so much time working on (and thinking about) the storefront platform project. It seemed obvious to move in this direction, but I didn’t know for sure if it would work. It took three years from the date of my original blog post. I had hoped for this exact type of store to pop up and then, almost immediately after we launched, there it was.
Leadership
I love writing about our various marketing channels. But the best part of these past five years? Building a team.
I now head up the Marketing department at eRetailing. The people on my team are smart and efficient and talented and I’m proud to be associated with each of them. I’m really looking forward to the next five years.