Ecomtent: AI product listings for e-commerce sellers at $100K a year

Ecomtent sells AI-generated product listing content — images, infographics, A+ Content and optimized copy — to e-commerce sellers, brands, retailers and Amazon vendors, and it was at $100K a year in revenue when co-founder and CEO Max reported that figure in a Starter Story interview in December 2023.

How it started

Two things collided in Toronto in October 2022. Max had just left six years at Amazon with a long-standing ambition to start a business; his co-founder, who holds a PhD in AI, showed him Stable Diffusion's first public model and produced an image from a text prompt in seconds. That was before ChatGPT, a couple of months after Midjourney appeared, and well before generative imaging was a boardroom topic.

The question Max asked was whether a specific product could be put into the generation, rather than a generic picture. The co-founder built an MVP over a weekend. It came with the usual early generative-AI defects — distortions, hallucinations — but it was enough to commit to.

The first version that shipped generated lifestyle images by training a custom AI model on each product. Three problems surfaced quickly:

It didn't scale — CTO Timur spent hours training a model per product.
Quality was weak — about 85% of the images came back unusable because of distortions.
It demanded 5–10 input images to train a model, while most customers arrived with a single image of their product.

A second MVP, launched in the last week of April, was built against those three findings, and it became the basis of what was on the market at the time of the interview. Revenue arrived a few weeks after the two started working together.

By December 2023 the company was a month into the Techstars accelerator, with $225k raised at pre-seed and plans to raise a Seed round after the program finished in January. The founders' working hypothesis at the outset was that more than 90% of marketplace content would be AI-generated within five years — an estimate the founder says now looks conservative.

What it sells and how it charges

On its official website, fetched on 2026-09-30, Ecomtent positions itself as AI Generative Engine Optimization (GEO) for product listing content: "Helping Sellers and Retailers win in the era of AI-powered conversational search." The product surface listed there includes:

AI product photography and lifestyle images
Infographics, including product dimension and Amazon infographics
A+ Content and optimized listing copy
Localization, plus automation for EU and USBC compliance
Channel coverage for Amazon (including Vendor Central), Walmart and eBay
Optimization and analytics aimed at AI search surfaces: Amazon RUFUS and COSMO, ChatGPT Search, Google Gemini

The homepage also carries the company's own marketing claims: that it is "the preferred Amazon listing software", that content "can increase conversion up to 30%", and that 88% of people search online for products and services before buying. Those are claims made on the company's own site.

On price, the public site as fetched does not publish plan prices; its calls to action are "Get Started Free" and "Book a demo". Ecomtent sells subscriptions, but no pricing page was available to quote, so cost is not something the website states.

How it got its first customers and what kept growth going

Nothing in the founder's account reads as a launch campaign. The mechanism he describes is conversation: he speaks to a customer on average every day, and the roadmap comes out of those calls.

Two patterns emerged from them. The first is the buyer profile — usually a Head of Ecommerce, Marketplaces or Amazon at a larger company, or a founder or owner at a smaller one, someone who needs to create and update listings but has neither Photoshop skills nor hours to spend editing. The second is past behaviour: customers were generating lifestyle images and then opening Paint or Photoshop to fix them. That attempt mattered more to the founder than a feature request, because it showed a problem customers had already tried to solve. The response was an in-app edit feature — highlight an area of a generated image and regenerate elements inside it.

The larger retention lever sat underneath the interface. An upgrade to the base model behind the AI product photoshoot methodology went live in early September, improving clarity and detail, realism of product placement and complexity of scenes. "It reduced our churn by 2/3rds."

At the time of the interview the founder also described the business as growing 25% month over month. Like the revenue figure, that is his own report for that period.

The numbers

Figure	Basis
$100K a year	Founder-reported revenue in the December 2023 interview; not current revenue
25% MoM growth	Founder-reported at the time of the December 2023 interview
9 employees, 2 founders	Team size at the time of the interview
October 2022	Month the two founders started working together
Toronto, ON, Canada	Base at the time of the interview
$225k pre-seed	Raised as of the interview; Seed planned after Techstars

The interview gives no figure for what it cost to start the business, and the website doesn't state one. The $225k pre-seed is funding raised, not a starting cost.

How it compares

The comparison below comes from 1M.chat's analysis of founder-reported revenue in Starter Story interviews — 1,997 businesses interviewed between 2015 and 2026. It is a self-selected sample of founders who chose to be interviewed, not a random sample of all businesses, so the figures describe that set of cases rather than the market.

Within the AI tools & apps category, 84 businesses have founder-reported revenue:

Median monthly revenue: $16.5K
Share at $1M a year or more: 26%
Median employees: 2
Share started by one founder: 42%
Among the 40 businesses with channel information: organic social media 60%, SEO and organic search 60%, partnerships and affiliates 35%, launch platforms 32%

By that analysis, Ecomtent's reported revenue sits below the category median. Its team was larger than the category median — 9 employees against a median of 2 — and it had two founders, where 42% of businesses in the category were started by one. One caveat on reading this: the comparison places a revenue figure reported in December 2023 against a dataset of interviews that runs through 2026.

What a founder can take from it

Choose a market whose workflow you already know. Six years at Amazon is what turned a generative imaging demo into a listings problem. If you're picking, start from work you've already done.
Ship the first version, then treat its flaws as the spec. Per-product training, the ~85% unusable output and the 5–10 image requirement were only knowable after launch.
Ask what customers already tried, not only what they want. Attempts to fix images in Paint and Photoshop pointed straight at the edit feature; evidence of past effort is a stronger signal than a wish list.
Look at output quality before adding features when churn is the problem. The two-thirds churn reduction came from a base model upgrade, not from more surface area.
Assume competitors will copy the demo, not the business. A rival joined the founder's WhatsApp group for Amazon sellers and messaged his customers offering free AI product images; his conclusion is to pick something unsexy and hard to solve.

These are one founder's numbers at one point in time. They are not current figures, not a forecast, and not a promise of income.

Sources

Revenue, team and start-cost figures are what the founder reported at the interview date, not current figures and not a prediction of what you will earn. Product and pricing facts come from the business's own website as checked on 30 September 2026. Method & sources · Disclosure