This is a very special case study for me. This is the story of my first experience building a business. But also the first time I had to shut something down. Let me explain you how me and partner Fabricio Cortina were able to scale in 5 months our e-commerce business with just 1 product.
I am going to start showing you quickly our main results and achievements:
$86.861 in 5 months
2707 orders
173535 to our website
56 order orders in just 1 day (almost 2k in revenue)
Conversion rate went really down because of virality and there was a reporting bias between Analytics and Shopify
Video Ads with over 2.5 million views (not reach)
Image Ad went Viral with 41k shares in just 4 days (170.000 people reached in 1 day)
Went from spending in ads $50-$80/per day over a $1000/per day
Have a much better understanding of how Facebook Ads works in Ecommere after spending $46k (Testing a LOT of things)
I discovered this new way to approach E-commerce in 2017 but I experimented with it seriously in early 2019.
You sell products as yours but you don’t touch the product. You are in charged of the brand identity, pricing, strategy... everything except shipping.. Normally the products you sell are stocked in China and you can sell them almost worldwide. Many people starts selling products and sourcing it from Aliexpress. Firstly we did that, but as we scale we hired a Chinese Agent who owned a Warehouse and shipped our products.
Basically we started promoting products via Facebook Ads. We tested 3 or 4 until we discovered our true winner. Our winner was a very strange product. It was a Samurai Katana Umbrella. How can you determine if your product has the potential to be winner?
This kind of business works very well with Shopify thanks to their integrations. We downloaded a theme from www.themelock.com and we customized so it didn’t look as anything before.
We installed very few Apps. We only got Oberlo, SMS Bump (recover cart with SMS), Reconvert(Email Recovery Cart) and SweetUpsells. And we tweacked some parts with HTML, CSS and Javascript
This is a very unique type of business and for starters you should focus in being profitable as soon as you can because of your cashflow. Firstly, we ran an Engagement Campaign (cheaper CPM’s) to test ad creative and engagement with the products. The campaign structure was the following:
Demographics (44 e-Packet countries), Placements, Creatives and Budget across the adsets were all the same. The only thing that was different between them was the interest targeting. In that way we could identify quickly which type of audience worked best for us. In the very first day we’ve got some very good results. There was very good engagement and people commenting. And some of them proceeded to checkout. That was what we were looking for.
The ads needed to grab people attention quick. This business is based on impulse buying. Person who clicked. Person who needed to buy.
We’ve tested different thumbnails and videos and we decided we were going with 2: Video A and Video B. There were very similar videos but their first 3 seconds weren’t equal.
I implemented a strategy that I was using with another client. I didn’t create new ads for every adset. I created only one ad and I connected them to all the adset. This means that every engagement we got from an adset will replicate in the others.
We realized this product had potential. We duplicated the engagement campaign and we turned it into a Purchase Conversion campaign. The first day we’ve got $280 sales spending only $80. We were getting something there.
So the following weeks we kept doing some tweaks. Our KPI’s were cost per purchase and ROI. So if an adset wasn’t doing ok, we would turned it off and look for more audiences.
Something that worked really well for us was breaking down an adset by demographics, location, age & gender, platform; and duplicating new adsets with that variables changed.
Once we’ve got the ball rolling, we were gathering enough data to create some custom Audiences. The guidelines we followed were like these:
Combined with different Periods of Time based on how many data we had
We were using Paypal & Stripe as our payment processor. During the first week we’ve received a Stripe notification. They told us they were going to holdback 25% of our revenue for 3 months. And that happened.
That forced us to improve our Customer Service and also to be really careful with scaling too fast.
Then we had a problem with our EIN number and stripe. And they holded 100% of our revenue until we fixed it. That hold lasted about 2-3 weeks. So we were really chill about scaling.
Once we solved the payment processor problems, we started to think about scaling. First of all we need to define what scaling is in Paid Media.
"Scaling is spending more. If you are able to spend more, and get proportionally more revenue, you are doing it right."
That being said, the actions we took for scaling were the following (Not all of them worked, and there weren’t tested all at the same time):
As you can see, we tested a LOT of ways. Overall I think that for our product the best ways to scale were:
And our sweet spot was spending around $150-$200 per day. When we spent that normally we generated between $650 to $900 in revenue.
Important: Take into account that we weren’t working on any type of branding campaign. We didn’t upload anything to our social media organically. We didn’t do content marketing. We were just trying to get new customers every single day. I will talk about it a little bit more in the conclusion section of the case.
We had been changing ads creatives over a while. We’ve created a 2 new videos where the main characters were woman instead of men. We also have created some specific videos for stories with footage that our Chinese Agent sent us via skype.
But I had an hypothesis:
If I could explain in a single image what the product is, I will get people’s attention quicker.
That’s what I did. I researched some customer images wearing the umbrella and I formated the ad as it wasn’t an Ad. (Native Ads)
The first days were ok. But something happened. Suddenly, on 15th of November around 4pm Argentinian time, numbers began to rise exponentially. In just 10 hours our image was being shared 11 thousand times.
We were getting hundreds of people all over the world into our website, without us targeting them with our ads. The next day was even better, we had more than 12k sessions in a single day.
The virality lasted around 4-5 days. It ended up having 40 thousand shares. IT WAS INSANE!
I tracked down the shares. It became viral thanks to Facebook friendliness with viral stuff. Some random guy shared the ad in a Facebook Anime Group with over 170k members. And it became viral inside the group.
Also another random guy, shared it in another Facebook Anime Group. And it became viral in there too. In that way, I had 2 viral post shares + facebook ads . And that’s how I realized the social networks reach potential.
After the insane results we’ve got from the viral ad, everything fell apart. I was working at 2 jobs at that time as a Growth Marketer. And the year was coming to an end, I was really tired. I stopped giving the business the attention it needed.
We were spending way too much in ads but the results were getting worst every day. I tried making some changes but they didn’t work. So me and my partner decided we were going to take a break. We shipped the last units, and we closed the store.
During these journey we learned a lot of stuff. We made some very good decisions and some very bad ones. It was really fun and exciting to watch analytics all day, and testing and creating new strategies. This is what I took from it:
What did I learned? 📚
What thing did we do right? 👍
What thing did we do wrong? 👎
What thing should we have done? 🧐
In conclusion, considering we weren’t an existing brand and 96% of our customers were new, we did an incredible job. I am 100% sure that if we implement some of these techniques to an existing brand it would work very very good.
Absolutely. But with a longer term perspective. I think this is the best way of starting an ecommerce and then transform it into an actual brand. Also a very cheap way for brands to test new products to sell, or new markets.