I’ll share my experience selling branded merchandise on Amazon.com — and I’ll say it right away: the first pancake was a bit of a flop. We started this project a year and a half ago, and the product is still available online. The idea was to offer consumer goods to users of CleanTalk.org under the CleanTalk brand.
Why branded merch for an IT company?
To increase the average revenue per user and to promote the brand through everyday mentions, outside the company’s core services. The idea was to replicate the concept of souvenir shops in tourist spots — but adapted for a cloud service.
Where to sell and how to attract customers?
To keep things simple, we chose to sell on Amazon.com using Fulfillment by Amazon (FBA) — they handle the storage and shipping. We drive traffic from our website using small banner ads and suggestions that don’t distract users from our main products.
We also hoped to get some organic sales from Amazon’s internal search.
Product selection, manufacturing, logistics
As our first product, we chose a thermos bottle. We found a manufacturer in China who quickly (in less than a week) and not for “a bowl of rice” produced a batch of 50 bottles with our brand engraving. The cost per bottle (around $7) gave us hope for profit after covering all future expenses.
In March, we received two boxes of bottles in the U.S. In April, we sent them to Amazon’s warehouse and launched the product for sale on Amazon at a price of $19.99, with a promotion targeted at our user base.
Here’s the link https://www.amazon.com/CleanTalk-Leakproof-Double-Insulated-Stainless/dp/B0BRJZ3K55/
Sales in Three Geographic Zones — United States, Canada, Mexico
In total, manufacturing and shipping from China to the U.S. cost around $450.
By the way, the bottle manufacturer could have shipped the goods directly to Amazon’s warehouse, which would’ve made delivery cheaper, but we didn’t manage to set that up before sending the bottles to the U.S.
Launching Sales
Within the first few days, it became clear that the “souvenir shop” effect didn’t work out. Specifically, we didn’t see the 20–30% conversion (subjectively, based on personal experience) that we had hoped for from users who interacted with our core product. Actual sales were about one bottle every two weeks, which means the real conversion from product users was close to zero.
Presumed reasons for the failure:
- Lack of variety. That 20–30% conversion typically happens when users can choose something that suits them personally from a wide range of items. If you only have one product on the shelf, your conversion rate is limited to the demand for that exact item.
- No “souvenir shop” vibe. As much as people might downplay it, the presence of a crowd, the ability to touch and handle the items, background music — all help to create a shopping mood.
- An IT website doesn’t evoke the same emotions as visiting a historical site, museum, or amusement park. There’s probably little to no desire to associate oneself with an IT brand. So we likely lost a few more percentage points in potential conversions there.
- We didn’t want to ruin the user experience by distracting from our main product. Unlike tourist attractions, there’s no “mandatory gift shop” in the user journey of an IT product.
Sales among our users ended up reflecting the theory of interaction between two independent sets — meaning people bought the bottle if, at that exact moment, they were in need of that kind of bottle, at that price, with those specs, and those shipping terms.
You can still loosely tie these sales back to our core product, since users were introduced to the merch after using our main service. But overall, we didn’t achieve the same kind of impulse buying you see in real-world souvenir shops.
Probabilistic Model of Souvenir Sales
We can describe the observed sales pattern using basic probability theory,
P(A and B)=P(A)×P(B)P(A \text{ and } B) = P(A) \times P(B)P(A and B)=P(A)×P(B)
Where:
- P(A) is the probability of purchasing our core product.
- P(B) is the probability of purchasing the bottle.
P(B) depends on multiple factors: the bottle’s price, features, shipping cost, geography, user reviews, and overall brand strength.
Example:
If out of 1,000 users, 160 purchase our main product (P(A) = 160/1000),
and P(B) = 1/1000 (based on our real-world data),
then: P(A and B)=1601000×11000=0.00016P(A \text{ and } B) = \frac{160}{1000} \times \frac{1}{1000} = 0.00016P(A and B)=1000160×10001=0.00016
This gives us a 0.016% probability that a customer buys both — the core product and the branded bottle.
Price Experimentation
After about three months of sales, we started experimenting with price reductions. Eventually, we found a “sweet spot” at $11.99 — a price point where we started seeing a few sales per week.
Further price drops didn’t significantly boost volume. Even during Black Friday, when we lowered the price to $8.99, sales remained steady rather than spiking.
This once again highlights our miscalculation regarding how customers perceive the value of branded merchandise.
The initial price of $19.99 was in line with the market — comparable to similar bottles sold at Walmart, Target, and on Amazon. However, since the “souvenir shop” buying triggers weren’t working, our unknown brand held much less perceived value, and customers weren’t willing to pay market rates for it.
Sales via Amazon: Storage and Fulfillment Costs
Let’s not forget — we had a Plan B. Amazon! The go-to marketplace in the U.S. for selling almost anything. But even there… no breakout success. Our bottle entered a highly saturated category — over 6,000 competing listings for “32oz Water Bottle” on Amazon:
https://www.amazon.com/s?k=32oz+Water+Bottle&i=garden&crid=28IYBNFK0EAYS&sprefix=32oz+water+bottle%2Cgarden%2C206&ref=nb_sb_noss_2
It quickly became clear that we were just one more bottle in an ocean of options — and not a particularly visible one, either. Meanwhile, Amazon’s FBA model comes with its own costs:
- Monthly storage fees
- Fulfillment fees per unit sold
- Referral fees on every transaction
These expenses further ate into our already narrow margin, especially at the lower price points we had to adopt in order to make any sales at all.
Visibility, Pricing, and Amazon Realities
As expected, our bottle simply doesn’t show up in search results — and therefore, doesn’t sell. Lowering the price does not improve ranking. Even after getting two 5-star reviews, there was no noticeable impact on visibility or conversion.
We also tried tweaking the product title, descriptions, and other listing details — again, no effect.
At this point, I suspect that in order to get organic visibility on Amazon, a seller might need to generate the first 100–500 sales at a loss. This would help accumulate reviews and push the product up in rankings based on sales volume — Amazon’s favorite performance signal.
The Hidden Costs of Amazon FBA
And let’s not forget — Amazon FBA isn’t free.
Yes, they store your product and handle fulfillment — but they charge for both.
Here’s how our costs broke down per unit:
- $8.75 per bottle during the first six months
- $9.04 per bottle during months 6–9
- And it gets even more expensive over time, because longer storage means higher monthly fees
Storage and fulfillment fees are based on product size and weight — and they scale up the longer your inventory sits unsold in Amazon’s warehouse.
And the Bottle Itself?
Honestly — it’s a solid product.
It’s something people use every day, and the quality is excellent.
We’re proud of it. It’s not the kind of item you’d feel embarrassed presenting to a client or using in public.
The issue wasn’t with the product — it was with positioning, visibility, and customer motivation. In other words, a marketing misfire, not a manufacturing one.
Leave a Reply