Building a Simple Product Research Workflow for E-Commerce With Python

Most product research advice in e-commerce is vague:

• “Find a winning product”
• “Look for viral trends”
• “Copy what competitors are doing”

The problem is that none of this gives you a repeatable process.

I have been experimenting with a simpler approach: collecting product ideas in one place, assigning a few basic signals to each one, and using a lightweight Python script to make the first round of filtering less random.

This is not a magic “winning product” finder.

It is just a small workflow that helps turn scattered ideas into something easier to compare.

The problem with manual research

When you research products manually, you usually end up with dozens of browser tabs, screenshots, saved TikToks, supplier links, and notes in random places.

A week later, it becomes difficult to answer simple questions:

• Why did I save this product?
• Is the price range realistic?
• Is it already saturated?
• Does it solve a clear problem?
• Can I explain it in one sentence?

So I started using a basic spreadsheet as a database.

The fields I track

For every product idea, I add a row with fields like:

Field	Why it matters
Product name	Helps keep research organized
Estimated cost	Needed for basic margin calculations
Expected selling price	Helps identify unrealistic pricing
Problem solved	A product should have a clear use case
Competitor activity	Shows whether there is active demand
Creative angle	The main hook for ads or content
Notes	Any important details or concerns

The goal is not to make the spreadsheet complicated.

The goal is to avoid relying only on “this looks cool.”

A very simple scoring model

After collecting a few ideas, I use a basic score from 1 to 5 for each category:

• Problem clarity
• Price potential
• Creative potential
• Shipping simplicity
• Competition level

Here is a small Python example:

python
products = [
    {
        "name": "Portable desk lamp",
        "problem_clarity": 4,
        "price_potential": 3,
        "creative_potential": 4,
        "shipping_simplicity": 5,
        "competition_level": 3
    },
    {
        "name": "Pet hair remover",
        "problem_clarity": 5,
        "price_potential": 4,
        "creative_potential": 5,
        "shipping_simplicity": 4,
        "competition_level": 2
    }
]

for product in products:
    score = (
        product["problem_clarity"]
        + product["price_potential"]
        + product["creative_potential"]
        + product["shipping_simplicity"]
        + product["competition_level"]
    )

    print(f'{product["name"]}: {score}/25')

This does not prove that a product will sell.

But it makes the decision process more visible.

Instead of choosing products based only on instinct, I can compare ideas using the same criteria.

What I am learning from this

The strongest product ideas are usually not the most unusual ones.

They are often products that are easy to understand in a few seconds:

They solve an obvious problem.
They have a clear before-and-after effect.
They can be demonstrated visually.
The price feels reasonable without needing a long explanation.

Automation is useful here because it reduces repetitive work.

For example, it can help with:

Cleaning product data from CSV files
Comparing price ranges
Grouping product ideas by category
Calculating basic margin estimates
Creating a shortlist for deeper research

The important part is still human judgment.

A script can sort data, but it cannot fully understand whether a product has a strong emotional hook, whether a creative feels believable, or whether the customer experience will be good after the purchase.

My current stack

Right now, I am keeping it simple:

Google Sheets for collecting ideas
Python for simple scoring and cleanup
CSV files for moving data between tools
Manual research for product validation
Basic notes on creatives, pricing, and competitors

No complicated dashboards yet.

Just a system that is easy enough to use consistently.

Final thought

The goal is not to automate every decision.

The goal is to spend less time organizing messy research and more time evaluating ideas properly.

For anyone learning e-commerce or automation, this is a good small project to start with:

Build a product research spreadsheet.
Add a few simple scoring fields.
Export it as CSV.
Write a short Python script that ranks your ideas.
Improve the system as you learn.

A simple workflow used every week is much more useful than a perfect system you never open.