Share Your E-Commerce Journey: Niches, Turnovers, Challenges and Lessons Learned

[Mehdi Bourechka]

Hello everyone,

I thought it would be valuable to open a discussion where we can share our e-commerce journeys, not only technical details, but also the real stories of how we started, what challenges we faced, and how we overcame them.

Here are a few prompts to get the conversation started:

What niche did you begin with?

What turnover did you manage to reach?

Which markets are you in?

What challenges slowed you down, and what clever tricks helped you move forward?

To get the ball rolling, I’ll also share my own experience. Besides running an agency, I’ve developed more than 50 PrestaShop modules and a theme, and I also operate my own e-commerce stores. This gave me insights into both the development side and the day-to-day challenges of managing stores, from scaling sales to handling customer expectations.

My experience

I started in e-commerce through dropshipping. Right from the beginning, I noticed the main strategy was heavily based on Facebook Ads because they are so precise for acquisition. The first challenge was that I was using PrestaShop, while most dropshippers around me were using Shopify. That made it harder, as I was facing tough competition and higher acquisition costs.

At the start, the numbers were very tight:

Acquisition cost: around 20 dollars

Selling price: 35 dollars

Product cost: 9 dollars
So the margin was only a few dollars. Some days, scaling meant no margin at all.

The first big unlock came when we added volume discounts and quantity breaks:

1 product = 35 dollars

2 products = 60 dollars

3 products = 85 dollars

Around 20 percent of customers started buying more than 2 units. Suddenly, margins went from 6 dollars per sale to about 20 dollars per sale because shipping and acquisition costs were spread over multiple items.

We scaled progressively from 50 dollars per day campaigns to 100 dollars per day and later to 200 dollars per day. As spend increased, cost per sale decreased, but visibility also triggered more competition since dropshipping has a low barrier to entry. Acquisition costs rose again.

So, we built our own solution: an Upsell Pop-up Module. When someone added a product to cart, a pop-up suggested a related item with 20–30 percent discount. For example, people buying girl toys were offered a boy toy as an upsell. Since many buyers were parents or grandparents with both girls and boys, this was very effective. Profits increased another 30 percent thanks to more multi-product orders on a single acquisition.

Later, we faced the logistics nightmare of sending data to suppliers, importing tracking numbers, and handling client communication. Customers often complained if they saw “China” on tracking info, leading to refund requests.

We developed another module, connected to an API covering more than 3,000 carriers worldwide.

Clients could track their orders directly on our site by entering order reference and email.

We automatically removed mentions of “China” or airports to avoid confusion.

The module mapped carrier statuses and directly updated PrestaShop order states such as shipped or delivered.

All we had to do was bulk-import tracking numbers via Excel, and everything else was automated.

This was all in the first quarter. In just 3 months, we did around 30,000 dollars turnover. That was our first year of trying dropshipping with PrestaShop.

And that was just the beginning. I would be happy to share what happened in the following years if you are interested.

Your turn

Now I would like to hear from you:

How did you get into e-commerce?

Which markets and niches did you test?

What were your biggest challenges, and what solutions helped you scale?

Let’s make this a thread where we can all learn from each other’s experiences.

[cocos.codes]

Hi,

Thanks for sharing this very important information. Were the social media ads mainly on Facebook? Was social media the main source of conversions?

For us, the good old technique works very well: Google positioning. Social media ads are an addition, but we are still working on them. Just adding an ad is one thing, but choosing the right audience and setting it up is another matter, so that it translates into conversions 🙂

[Mehdi Bourechka]

6 hours ago, cocos.codes said:

Hi,

Thanks for sharing this very important information. Were the social media ads mainly on Facebook? Was social media the main source of conversions?

For us, the good old technique works very well: Google positioning. Social media ads are an addition, but we are still working on them. Just adding an ad is one thing, but choosing the right audience and setting it up is another matter, so that it translates into conversions 🙂

Hi,

Yes, the conversion ads were mainly on Facebook Ads, since that’s the platform we have the most experience with. We did get some organic sales too, but not through structured SEO or positioning, rather, it happened because of the heavy traffic generated by Facebook campaigns. Interestingly, Google picked up and ranked us for some products as well, so the combination of both worked out nicely.

That said, when you’re selling physical products, the shipping context and product lifecycle (usually around 6 months) make deep SEO investments less practical. However, even small SEO efforts can still generate some extra sales over time.

On the paid side, we’ve tested many channels like TikTok, YouTube, and others, but Facebook consistently delivered highly targeted, qualified traffic. Even with Advantage+ campaigns, letting the algorithm optimize, the results were strong. One big benefit of Facebook is scalability: since ads run on impressions rather than limited keyword searches (like in Google), increasing the budget directly increases reach and sales. For example, we went from 5 sales/day to 14 sales/day just by scaling the budget.

So, in short:

Google SEO = long-term strategy, slower to build but sustainable.

Facebook Ads = short-term, scalable, almost like a “black box” where you put money in and profits come out.

Both can work well together, but for fast traction and growth, Facebook has been our main conversion driver.

[cocos.codes]

Thank you for another interesting piece of information. Tell me what elements of creating a Facebook campaign you consider most important.

I think it's the content of the ad itself and the target group. However, we've recently had a problem with selecting the target group – you can choose a source of interest, e.g., Shopify, Magento... but not PrestaShop.

 

Many people also do the following:

- they launch ads on Facebook for a general target group

- after collecting data on who clicked on the ad, they bombard those who clicked in the previous campaign with ads

[Mehdi Bourechka]

How Meta (Facebook) ads actually work in 2025, before you spend a penny

Quick history: why “targeting hacks” don’t win anymore

Era 1  Interests only: In the early days, Meta grouped people by page likes and on-platform behavior. CPMs were cheap, hyper-targeting/crossing interests felt effective.

Era 2  Pixel learning: The Facebook Pixel (a small JS snippet in your site header) started sending events (AddToCart, Purchase, Lead…). With enough volume, the algorithm learned who buys and got very precise.

Era 3  Privacy + CAPI: iOS/ATT and browser restrictions reduced trackable data. Meta responded with Conversions API (server-side events) to recover accuracy by matching browser events with server events.

 

Takeaway: Today, the algorithm does most of the targeting heavy lifting. Your job is to feed it high-quality signals and great creatives.

 

What you must set up before running ads

Pixel + Conversions API (CAPI): Implement both. Use event_id for deduplication between browser and server events so Meta doesn’t double-count.

Event hygiene: Send the right standard events (ViewContent, AddToCart, InitiateCheckout, Purchase) with clean parameters (value, currency, content_ids).

Domain verification & event prioritization: Make sure your domain is verified and your web events are prioritized properly.

Quality check: In Events Manager, confirm Event Match Quality is good and use Test Events to validate every step (landing → purchase).

Data consistency: Ensure cart value, currency, and order IDs match between front-end and back-office. That’s what lets Meta reconcile browser vs. server purchases.

If your tracking is sloppy, the algorithm is blind. Fix tracking first.

 

“Testing the market” vs. “testing the platform”

Budgets under ~$1,000/day are not a true market test; they’re a platform/offer test. Real market tests (2–3 weeks, ~10% population reach) are multimillion-dollar campaigns.
With smaller budgets, expect noisy outcomes across countries. A product might crush in Canada and stall in Australia at $50/day for 48h, that doesn’t prove “the market is bad”; it often means insufficient signal/time for the algo to learn.

Campaign setup that’s working now

Start broad with Advantage+ (ASC) for Sales/Purchases.

Limit targeting to location + language and let the algo find buyers.

Minimal fragmentation: Fewer campaigns/ad sets; let each reach 50+ conversions/week if possible to exit the learning phase.

Catalog on? If you have a catalog, use it. Dynamic formats + broad signals = stronger learning.

Retargeting: Smaller than before; still useful, but avoid over-segmenting.

 

Creatives: where the actual leverage is

You win (or lose) on creatives. Build a simple creative system:

Angles: Problem/solution, demo, social proof/UGC, founder talk, unboxing, “why us,” offer/guarantee.

Hooks: Test 10+ opening hooks per concept. The first 3 seconds decide CPM and CTR.

Formats: Ship 9:16, 4:5, 1:1, and 16:9; with sound and no-sound/captions.

Lengths: 8–15s (short), 20–30s (medium), 45–60s (long).

Variants: Swap intros, CTAs, overlays, subtitles. Expect 3–5 concepts × 4 hooks × 3 lengths to find 1–2 winners.

Don’t forget static images and carousels, cheap, fast signal on angles.

We run a video/marketing production team and have seen ugly sites sell when the offer is great, and gorgeous sites fail when the offer is weak. Offer > design.

Reading the numbers (so you know what to kill/scale)

Creative-level (early): Thumb-stop (3s view), hook retention (0–3s→6s), outbound CTR, cost per ATC/IC.

Business-level (decisive): CPA/CPS, ROAS (or MER), AOV, CVR (session→purchase).

Decision windows: Give a creative $50–$150 and 1–3 days (depending on AOV) before judging. Kill losers quickly; re-invest in winners.

 

Scaling without nuking performance

Vertical: +20–30% budget increments on winners (don’t edit daily).

Horizontal: Duplicate winners into a new ASC or new country; launch fresh creative variants weekly to fight fatigue.

Stability: Avoid constant edits that reset learning.

Common pitfalls I see all the time

Over-targeting and interest stacking (starve the algo).

Broken deduplication (no event_id parity).

Optimizing for the wrong event/window (e.g., top-funnel event when you want purchases).

Judging a market off 48 hours at $50/day.

No creative pipeline, running the same few videos until fatigue hits.

 

To sumup:

Set up Pixel + CAPI correctly (dedupe + match quality), start broad with Advantage+, and build a creative testing machine (multiple angles, hooks, formats). Judge by business KPIs, not vanity metrics. Small budgets test the platform/offer, not the whole market.

[cocos.codes]

I have to agree with you completely – PIXEL does indeed send a lot of data that allows the algorithm to learn how to target people who will actually be interested in the offer. I was preparing the PIXEL WordPress + PrestaShop module, with data sent with deduplication and hybrid (JS + PHP/API). Thank you for opening my mind!

 

Unfortunately, we don't have millions to spend on advertising 🙂

 

Do you think there is a minimum budget below which it doesn't make sense to run META ads? We believe there is, and we have informed many customers about this.

[Mehdi Bourechka]

1 hour ago, cocos.codes said:

I have to agree with you completely – PIXEL does indeed send a lot of data that allows the algorithm to learn how to target people who will actually be interested in the offer. I was preparing the PIXEL WordPress + PrestaShop module, with data sent with deduplication and hybrid (JS + PHP/API). Thank you for opening my mind!

 

Unfortunately, we don't have millions to spend on advertising 🙂

 

Do you think there is a minimum budget below which it doesn't make sense to run META ads? We believe there is, and we have informed many customers about this.

That depends on your market, offer, pricing … if you can share more insights I can give a reasonable range budget accordingly 

[cocos.codes]

I cannot provide a link on the forum because it is the client's website. I hope this information is sufficient for you. Another company is running his campaign (we do not deal with this, although we have basic knowledge on the subject) and we have a slightly different opinion from the “agency” that is running it.

In short: the market is global, as the products sold are virtual (e-books), translated into many languages, and the margin on the products is practically 100%.

[Mehdi Bourechka]

If you send people directly to a $29 e-book page, most won’t convert. The reason is simple: they don’t know you yet, and their past purchases have probably been from well-known authors at much lower prices. So the trick is not to push the full book right away, but to start with a small, irresistible offer.
 

Something like a 10-page “mini version” priced at $1.99. It’s cheap enough to be an impulse buy, and it gets you past the biggest barrier: having their card on file and earning a first layer of trust.

The real profit comes after that first step. Instead of sending them to the thank-you page, you redirect them to an upsell page. That’s where you offer either the full version of the e-book (hundreds of pages) or even a bundle of five strong titles in the same niche for $29.99.
 

Some people will say yes, and that’s great. Those who say no shouldn’t just be dropped, this is where you introduce a downsell, the same offer but with a 20 to 30 percent discount (for example $19.99). Either way, whether they accept or decline, you then move them to the thank-you page.

This kind of funnel works because it matches buyer psychology: low friction entry, then higher ticket offers when trust has been built. The upsell and downsell stages are where you recover ad spend and generate profit.
 

Speaking of ads, since you’re targeting global markets, I’d suggest splitting the budget roughly 70 percent on cold traffic for the $1.99 offer, 20 percent on retargeting (people who clicked or almost bought but didn’t), and 10 percent on awareness/testing creatives.
 

A starting daily budget of $50 to $100 is enough to gather data, and from there you can scale by reinvesting part of the profits.

It’s not about the $1.99 sales at all, it’s about building a funnel that takes someone from “never heard of you” to “just spent $30 with you” in a single flow. That’s the power of this strategy.

[cocos.codes]

Thank you, that's really interesting information. We'll pass the idea on to the client and let you know the results. The idea seems very good 🙂

If you want to share more ideas like this, I'd love to read more!