Chaos doesn’t scale.

We put e-commerce in order so your ads, content and product data turn into profit, not just revenue. We test new sources of growth in Google, AI assistants and social commerce with small, fixed-price experiments.

Maciej Szymczuk, AI eCommerce Advisor

Let’s talk about your store.

Send me a link to your store and tell me what you want to improve. In 30 minutes we’ll go through where AI can help, what to check first and how to tell whether a change pays off.

Email me

maciej.szymczuk@wehuman.io

ROAS of 5. And the business is losing money.

Your ad dashboard counts revenue, not profit. Once you subtract cost of goods, shipping, returns and fees, a campaign that looks great can lose money on every single order.

That’s why every engagement starts with margin, not budget. We measure what the business actually keeps, not what the dashboard shows.

Example: a month in a store with a “good” ROAS

Sales revenuePLN 100,000
Cost of goods−PLN 65,000
Shipping, payments, returns−PLN 10,000
Advertising−PLN 20,000
Agency fees−PLN 6,000
Result−PLN 1,000
ROAS in the dashboard: 100,000 / 20,000 = 5. Example figures.

What we do

One team, from cleaning up your catalog to selling on TikTok Shop. Everything runs on the same well-organized product data.

Order in your store and catalog

Store structure, UX and categories, with product data managed in one place with help from AI. See how it works.

Migration to Shopify

When your current platform is holding you back, we move your store without losing your Google rankings and connect your sales channels.

SEO and GEO

Clean data, a site search engines can access, and expert content that helps customers choose. We measure traffic and sales from Google and AI assistants, not one-off chatbot answers.

Social commerce

Facebook and Instagram catalogs, TikTok Shop, and Meta and TikTok campaigns that sell without anyone leaving the app.

AI content and video

Product shoots, campaign videos and social content created with AI, faster and cheaper than traditional production.

Growth experiments

A new audience, a new value proposition, a new market. We test the direction in 1–2 months, before you commit a bigger budget.

All your product data in one place.

Descriptions, photos, variants, SEO, EAN codes and versions for every channel. AI prepares them in line with your communication strategy, and you approve.

These aren’t mass-produced, one-click descriptions. AI works from your product knowledge and brand rules, and everything goes through quality control. You don’t need thousands of products for this to pay off. With 50 products you save time and gain consistency; with 50,000 it’s the only way to keep the catalog under control.

  • One brand voice everywhere. We set naming rules and description style once, and AI applies them to every product.
  • EANs without spreadsheets. GS1 codes prepared and registered for every variant, consistent with your store data.
  • One change, every channel. Your Shopify or Shoper store, Google Merchant Center, Facebook, Instagram and TikTok Shop all get the same up-to-date data.
  • Ads promote what makes money. We connect the feed to margin and stock levels, so your budget doesn’t go to products that get clicks but don’t earn.
See how AI eCommerce works

An experiment, not a big transformation.

We don’t build a months-long project only to find out at the end whether it works. Every experiment has a fixed scope, timeline and price.

  1. 1. Hypothesis

    We name one specific source of growth to test.

  2. 2. MVP

    We build only what the test needs. Usually 4–6 weeks.

  3. 3. Traffic and data

    Real customers from Google, Meta and TikTok. We measure behavior and sales.

  4. 4. Decision

    We scale, adjust or stop, before you spend a big budget.

Scale

The result held up. We add catalog, campaigns, content and more channels.

Adjust

The result is unclear. We change the offer, price or messaging and test again.

Stop

The hypothesis didn’t hold. You know where not to invest, and you save a big budget.

We tested it on our clients’ brands.

A woman reading in an armchair, next to a dog in a gray LivaMio bed

LivaMio

same products, new audience

Beds and accessories for dogs and cats. Instead of rebuilding Royal Vision, a brand that already worked, we created a parallel brand for customers who buy design for their homes. New positioning, content, campaigns and a separate Shopify store.

result: ROAS of 4.5 after the first 8 weeks

See the new-audience experiment
A man in a SUAP.M T-shirt with a Full Chaos Mode print, with a dog on his lap

SUAP.M

a clothing brand built with AI

A fashion brand for dog owners. A catalog of 30+ products, photo shoots and product videos generated 100% by AI, plus the store, social media and campaigns.

2 days instead of 4 weeks compared with a traditional shoot

The Human team has worked with Payback, Avon, Douglas, Medicover, Eurocash, Traficar, Grupa Pracuj and others.

Have a team and AI tools? Great.

A good e-commerce manager with AI can do a lot on their own today. That’s why we don’t sell hours, we sell results, and we adapt to how much you want to keep in-house.

You do it

You get a diagnosis, a plan and access to our AI eCommerce tool for managing product data. Your team implements, we advise.

We do it together

We plan, prepare the assets and handle quality control. Your team keeps the decisions and part of the implementation.

We do it for you

Full delivery: marketing and content at Human, UX at ExtraHut, technology and integrations at SoftwareHut. One contract and one person accountable. You can also choose an implementation partner in your market.

Questions before we talk

Why not just do it ourselves with AI?

Often you can, and if that’s true for you, we’ll say so plainly in the diagnosis. We bring what tools don’t: a margin model built on your data, marketing, UX and technology in one team, our own product data tool, and accountability for implementation, not just recommendations.

Do you guarantee we’ll show up in ChatGPT answers?

No, and no one can honestly guarantee that. Model answers change from day to day. We do what visibility really depends on: solid product data, a site search engines can access, and helpful content. We measure the effect in traffic, inquiries and sales.

How do you measure whether it works?

At the start we agree on how margin and costs are defined, and on a baseline. Where possible, we compare against a control group, for example a part of the catalog we leave unchanged. What counts is incremental profit, not revenue attributed to a campaign in a dashboard.

Do I have to move my store to Shopify?

No. We only suggest migrating when your current platform is genuinely blocking growth. Our product data tool also works with Shoper.

What if the diagnosis shows an experiment doesn’t make sense?

Then you’ll hear exactly that, and you keep the margin model and a 90-day plan you can implement yourself. If you order an experiment within 30 days, we deduct the cost of the diagnosis from it.

Let’s talk about your store.

30 minutes, no sales pitch. We look at your store, data and campaigns and tell you which experiment to start with.

Maciej Szymczuk

AI eCommerce Advisor

maciej.szymczuk@wehuman.io

Tomasz Kulczycki

Head of UX, eCommerce expert

Runs the store diagnosis and designs the experiment.

We’ll get back to you within 1 business day.