This Former Amazon Employee Built A $100K/Year Tool That Uses AI To Generate Product Listings
Hello! Who are you and what business did you start?
Hello! I am Max, co-founder and CEO of Ecomtent. Ecomtent enables e-commerce sellers to generate optimized product listings with AI.
With Ecomtent, sellers can generate lifestyle product images, infographics rich brand content, and optimized copy.
We started to work together in October last year, were making revenue a few weeks later, and currently are growing at 25% MoM with $100K ARR.
See a 3-minute YouTube demo here:
If you are thinking about it, do it. You will never regret trying and failing, but you will always wonder ‘what if’ if you don’t.
What's your backstory and how did you come up with the idea?
My Co-founder and I met on a founder speed dating program in Toronto. About a month into this program, in October last year, Stable Diffusion released its first public model. This was pre-ChatGPT, Midjourney had only been around for a couple of months and nobody (including me) had heard of it when the technology was really in its infancy and felt like magic.
My Co-founder, being a PHD in AI, showed me on Stable Diffusion how you could write a text prompt, and create an image in seconds.
Having spent the six previous years working at Amazon, I immediately saw how transformative this would be for the e-commerce industry. I asked him if it would be possible to put a specific product into the generation, rather than generating generic pictures, and he built a quick MVP over a weekend.
That MVP had all the problems typical of Generative AI - distortions, hallucinations, etc., but the potential was there.
From that point onwards, we were all in, with a vision of bringing this new technology (Generative AI) to an industry I knew very well (e-commerce).
Understanding how efficient it was to generate merchandising content with AI - not just images but copy and infographic visuals, we developed the hypothesis that +90% of content on marketplaces would be AI-generated within the next 5 years, an estimate that looks conservative now.
In many ways, the timing was incredibly lucky - the stars aligned that at the very same time, I was leaving the corporate world to satisfy my lifelong ambition of starting a business, and this once-in-a-generation technological shift occurred. I found myself at the starting line of the race that is playing out right now, and I felt (and still feel) that good execution will win.
My biggest advice to other founders looking to launch a business is to pick something unsexy and hard to solve. Our initial product’s critical flaw is that it is too sexy and cool, and thus many other entrepreneurs have been drawn to it.
Take us through the process of building the first version of your product.
The first iteration of our product was just focused on generating product lifestyle images. We would train custom AI models on each product to enable e-commerce customers to place their products in any scenario.
We learned three things from launching quickly:
1) This first version was not scalable, as it required my CTO Timur to spend hours individually training a model for each product;
2) The current methodology had a key technological problem of hallucinations, as ~85% of the images produced were not usable because of distortions; and
3) We didn’t solve a customer's pain, as we required 5-10 input images from the customer to train each AI model (a fundamental flaw, as most customers were coming to us because they only had one image of the product in the first place).
We took on these learnings, and launched a second ‘MVP’ in the last week of April, which fixed these three key problems. This second MVP is the foundation of what we have live on the market today.
Describe the process of launching the business.
Our initial sales came from my network. As a commercial founder, you must have strong industry expertise in the area you are building - (i) so you have experienced problems you are solving first hand (and at a very least multiple times second hand); (ii) so you can pick up your contact book and make these first sales with people who are willing to purchase based on personal trust built up over time, and (iii) so you have a good answer to the “why you” question that investors will later inevitably ask.
Following selling into my network, I broadened the outreach - I drew up a list of our top ~100 target customers in Excel, went through LinkedIn, and noted anyone that I had a 1st or 2nd-degree connection at these companies, and I leveraged my ex-colleagues for these introductions.
We grew through this B2B founder sales strategy and then also through word of mouth. This is a key growth channel - Wise acquires 1 million new users a quarter, with 70% of this acquisition driven by word of mouth. For an early-stage start-up, it’s one of the critical signs to demonstrate that you are getting close to product market fit.
Our final growth lever I focused on in the early days was SEO. This was a new topic for me, and I learned a lot from Google’s podcast ‘Search off the Record’. It helps to contextualize Google’s approach to ranking content, which in their words, favors “fresh, helpful and people-first content”. If you deliver this, it will align with Google’s nuanced, evolving, and intelligent and evolving signals in their algorithms.
I experienced this first hand, with one of the blogs I wrote which was particularly helpful and unique going viral, resulting in the number of clicks on our homepage (bold lines below) jumping from ~50 a day to 2.7k at peak, and improving the average rank of our homepage from ~17 to 7 on Google.
Since launch, what has worked to attract and retain customers?
It sounds corny, but one of our core values at Ecomtent is to learn every day from our customers (a principle drilled into me from my Amazon days, which we have taken to another level as an early-stage company). It’s something we live and breathe - I speak to a customer on average every day, and we are constantly evolving and iterating our product roadmap based on these conversations.
What are really powerful insights for me is not just what customers say they want (although this is still very important), but when I can see they have attempted to take action in the past to resolve such problems, rather than it being a ‘nice to have’.
For example, I saw lots of our customers using our tool to generate lifestyle images, and telling me they tried to use Paint or Photoshop to edit them. Frequent customer conversations also help to understand the specific profile of my customers we derive - usually a ‘Head of Ecommerce / Marketplaces / Amazon’ at a larger company, or a founder/business owner at a smaller ones.
These people are looking for tools to make their workload more efficient in creating and updating product listings, and are not skilled at using Photoshop and do not have the bandwidth to spend hours editing.
From identifying these patterns and user profiles in conversations, we decided to prioritize launching our own ‘edit’ feature, where customers can highlight an area of the image generated and regenerate elements (e.g. adding a cow in the background) in a second. It’s a much simpler, quicker, and cleaner feature than our closet competitors, and it comes from spending a lot of time listening to our customers and analyzing their past behavior.
The biggest impact on our churn rate however was a significant upgrade in the base model of our proprietary methodology to generate product photoshoots with AI, which went live in early September. This upgrade comes with a significant and notable improvement in the quality of the lifestyle images customers can generate, specifically: i) greater clarity and detail, ii) more realistic product placement, and iii) more complex scenes (see below). It reduced our churn by 2/3rds.
How are you doing today and what does the future look like?
My co-founder, my team, and I are determined to build an industry-defining business, building a new type of E-commerce-enablement company in the age of Generative AI. As I write this, I am currently in San Francisco, a month into the Techstars accelerator.
The program has encouraged us to hold a torch and magnifying glass to the business, seeking an even deeper understanding of our customers in a more structured way. This process has led to us deciding to work on a new, more automated MVP, where customers can generate content for 1,000s of product listings at once. I’d highly recommend it.
We are currently piloting this new product with several enterprise customers. Following the completion of the Techstars program in January, we will be looking to raise our Seed round, having currently raised a $225k pre-seed.
Through starting the business, have you learned anything particularly helpful or advantageous?
My biggest advice to other founders looking to launch a business is to pick something unsexy and hard to solve. This is a learning we are taking fully into our latest product. Our initial product’s critical flaw is that it is too sexy and cool. Generating beautiful product images instantly with AI makes for a pretty glamorous product demo, and thus many other entrepreneurs/intrapreneurs have been drawn to it. The fact they may not have any prior experience in marketing or e-commerce didn’t dissuade them. Two events drilled home this lesson for me:
Firstly, in one of our many applications to future accelerators, one of the principal VCs ended up almost laughing at us when he read our pre-submitted company one-liner, simply because he had heard it so many times before. We had evolved and refined our vision since this application form we submitted a couple of months before the meeting, but this didn’t matter.
You could tell that he was bemused by the fact that his team had even suggested he take the call with us, and the damage was done with this first impression.
The second event was perhaps even more brutal. As part of my continuing journey to obtain customer feedback, I started to run a WhatsApp group for Amazon sellers and customers. This group was billed as a group for sellers to share the latest AI tools, which it is, but it also serves as a very useful way for me to talk directly to my customer base, get feedback, and get volunteers to test the latest product features pre-launching.
My friends who are Amazon Sellers are on it, and whenever I go to an event, I add the Sellers I speak to, as it serves as an easy way for them to remember me and our service. One of our competitors sneakily joined this group (I advertised the link publicly on all of our blog posts), and proceeded to message every single one of my customers on there, offering to generate AI product images for free.
I guess that this competitor probably did not gain any customers from such a spammy message. However, the illusion of uniqueness that I had crafted with my customers was shattered, as they directly saw others offering a similar service. (This competitor has since stopped operating, illustrating such tactics are not sustainable in the long term).
My advice however remains - don’t get tempted into jumping onto the latest trend, even if you think it plays perfectly to your strengths. It’s a lazy strategy. Be smarter. Do something unsexy, complicated, and boring; and disrupt a traditional non-tech-enabled industry.
What platform/tools do you use for your business?
Our website is built on Webflow, we track user data on Mixpanel, we take payments through Stripe, we manage internal projects through Asana, and get customer feedback in surveys through Qualtrics.
What have been the most influential books, podcasts, or other resources?
Books
- How to win friends and influence people, by Dale Carnegie
- Lean Startup, by Eric Ries
- 7 Powers, by Hamilton Helmer
- Pitch Anything, by Oren Klaff
- Levers: The Framework for Building Repeatability into Your Business, by Amos Schwartzfarb
YouTube
- YC Channel
- Talks at Google Channel
Podcasts
- 20VC podcasts
- Lenny’s Podcast
- All-In
- How I built this
Advice for other entrepreneurs who want to get started or are just starting out?
If you are thinking about it, do it. You will never regret trying and failing, but you will always wonder ‘what if’ if you don’t.
Are you looking to hire for certain positions right now?
Not currently, but look out for us on Linkedin!
Where can we go to learn more?
If you have any questions or comments, drop a comment below!
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
- 4,818 founder case studies
- Access to our founder directory
- Live events, courses and recordings
- 8,628 business ideas
- $1M in software savings
