10 phases to launch
4–5 months of real preparation
70 % of carts abandoned globally
How to start an e-shop in 2026 — the complete guide

Five years ago I thought launching an e-shop was "just picking a template and uploading products". Then Lukáš and I really got stuck into it. And within the first week I discovered there are hundreds of things nobody warns you about, things that decide whether in a year's time you'll be counting profit or cancelling your trade licence.

This text is everything I wish I'd known back then. It's not a theoretical manual or a 6-page blog full of vague phrases, but a complete roadmap split into 10 phases. Every recommendation in this guide stands on something concrete — either on our own experience across dozens of e-shops, on a legal obligation, or on research from behavioural economics and neuromarketing that tells you why it works.

If at any point along the way you feel it's a lot — you're not alone. It is. That's why we turned each phase into a standalone deep-dive article (you'll find the links further down), and this text holds the overview of the whole journey together. Start at the top, walk through the map, and then dive into whatever is burning most right now.

What's ahead of you: 10 phases over 4–5 months

A realistic timeline from idea to grand opening is 4 to 5 months. It can go faster if you already have a brand, photos and don't need to sort out a trade licence. It can also go slower if you're doing everything yourself in the evenings after work. This plan is a benchmark, not dogma.

Phase

What it covers

Month

1. Preparation and strategy

Business plan, legal, brand

1

2. Graphic design

Photo, video, materials

1–2

3. Technical setup

Shoptet, domain, template

2

4. Products and content

Stocking, copy, SEO

2–3

5. Shipping and payments

Logistics, gateways, ERP

2–3

6. Orders and communication

Emails, invoicing

3

7. Social media

Profiles, catalogue, content

3

8. Tracking and marketing

GA4, pixels, campaigns

3–4

9. Testing and launch

QA, soft launch, grand opening

4

10. Growth

Retention, CRO, scaling

5+

Now, step by step. Feel free to skip around — you won't earn a badge for it, but the reading goes faster.

PHASE 1: Preparation and strategy (month 1)

This is where it's decided whether in a year you'll be drawing growth charts or paying your accountant to wind up the business. It's not a phase you want to rush.

The business plan: three questions you have to know the answer to

1. Who are you selling to? Not "women aged 25–45". That's a description of half the population, not a target group. Here's where Robert Collier's classic advice comes in handy: "Always enter the conversation already taking place in the customer's mind." In other words, until you can describe what keeps your customer awake at night, you don't have a target audience, you have a demographic. The difference between "women 25–45" and "mums on parental leave looking for chemical-free cosmetics who want to support local brands" is the difference between an ad everyone ignores and an ad that speaks directly to someone who instantly recognises themselves in the description. In real campaigns this typically shows up as a multiple difference in conversion rate.

2. Why should they buy from you and not the competition? Eugene Schwartz, one of the legends of direct marketing, put it plainly in his book Breakthrough Advertising (1966): copywriting can't create desire, it can only take existing desires and channel them onto a specific product. Your job isn't to "create demand", but to be the better recipient of an already existing need — either because you offer something the competition doesn't, or because you do it measurably better.

3. How much will it cost and how much will you have at the end? Margin, fixed costs, customer acquisition cost (CAC), average order value (AOV), Customer Lifetime Value (LTV). A simple equation every e-commerce consultant will tell you: if LTV/CAC < 3, your business doesn't make mathematical sense. Over the long term, a single customer has to earn you at least three times what you paid to acquire them. Less than that = you're diluting, not scaling.

Practical tip: Open a Google Sheet and map your 3–5 main competitors — prices, shipping, payments, communication, reviews. Come back to it once a month.

Legal: what you have to have done before you even launch

This isn't a section anyone wants to read. But it's the only section whose neglect can wipe out your business overnight. The specific legislative changes for 2025/2026 (DSA, GSPR, the accessibility act, green claims, PPWR, the "button" amendment) I break down in detail in the article on the legal obligations of e-shops — here just the bare minimum, without which you won't move an inch:

Sole trader or Ltd. An Ltd. makes sense from a turnover of around €40,000 a year (roughly 1 million CZK), when the tax advantages start to outweigh its fixed costs (accountant, register filings). Below that, a sole trader is cheaper and simpler.

Company ID, VAT number, a business bank account with a bank that offers a decent API and integration with accounting software (Fio, mBank, Raiffeisenbank). Accept that you'll enjoy it for about a week, and after that you'll only do it because you have to.

An accountant with e-commerce experience. This is not the place to save money. An ordinary accountant won't correctly set up the OSS regime for VAT on cross-border sales, won't advise you on imports from third countries, and will start to sweat at your first audit. An accountant specialised in e-commerce costs 30% more and saves you ten times that in fines and corrections.

Liability and goods insurance. When your warehouse burns down or a client cuts themselves on your product, there's a difference between having a policy and staring at the end of your company.

Terms and conditions and GDPR from a lawyer. Please, not a template downloaded off the internet. A consultation with a lawyer experienced in e-commerce costs €200–600 and saves you six-figure fines. In its inspections, the Czech Trade Inspection finds faults in the majority of e-shops, and penalties run into the millions.

Brand: the most underrated part of the preparation

People don't buy products. They buy the feeling that they made the right choice. And that feeling is given to them by the brand — from the logo through the colours to the tone of voice. This isn't a motivational platitude, it's neurology.

In 2004, a team led by Samuel McClure at Stanford University ran a famous experiment with Coca-Cola and Pepsi (McClure et al., Neuron, 2004). Participants first drank both drinks blind, and the fMRI scan showed they preferred Pepsi by taste. Then they repeated the experiment, but this time the participants knew which drink was which. The preferences reversed in favour of Coca-Cola. And the most interesting part: brain activity changed too — beyond the reward centre, areas linked to memory and cultural associations lit up. The brain wasn't drinking sweet fizzy pop. It was drinking memories, stories and decades of marketing. The brand literally changed the sensory experience of the product.

This is why a brand with a strong identity earns more than an identical product in generic-looking packaging. It's not about having a nice logo, it's about firing up a network of positive associations in the customer's head before they even start looking at the price.

What you need as a minimum:

PHASE 2: Graphic design and materials (month 1–2)

Product photos are the most underrated conversion driver on the entire e-shop. It's understandable: the brain processes visual information far faster than text, and the first impression of a page is formed within a fraction of a second. And the only thing the customer sees at that moment is the photo.

More photos from different angles typically convert better than a single thumbnail — not only because customers "want more information", but because a detailed inspection triggers the endowment effect, the ownership effect described by Kahneman, Knetsch and Thaler in the Journal of Political Economy (1990). Once a customer "picks up" the product in their head — zooms into the texture, turns it around from every side, clicks through the gallery — they start attributing greater value to it than if they had merely seen it as a thumbnail. And higher perceived value = higher willingness to pay.

What you have to have before launch:

Practical tip: Professional photography of 50 products comes to €1,000–2,000. Homemade photography on a shoestring looks like homemade photography on a shoestring, and it costs you more in conversion rate than you save on the photographer. AI generation still can't reliably replace a product shot on a white background; for lifestyle and mood photos it already can — but that's a different story for a different article.

PHASE 3: Technical setup of the e-shop (month 2)

In the Czech market, Shoptet is the de facto standard for an e-shop — it has complete integration with comparison sites, carriers, payment gateways and accounting. Unless you have a very specific reason to go elsewhere (Shopify for international expansion, a custom solution for large B2B), Shoptet is the default choice and this guide assumes it.

Technical setup steps (short version)

  1. Register a trial account at shoptet.cz. You get the trial period for free and the full admin is available straight away.
  2. Choosing a plan — since 1 September 2025 Shoptet has Free, Basic, Business, Profi and Enterprise plans. For most new e-shops, Basic or Business makes sense (depending on the number of products); you move up to Profi once you need more advanced functionality or more than 1,000 products. Always verify current prices and limits on the official shoptet.cz pricing page.
  3. Domain — register a .cz for the Czech market, a .com for international. DNS records can be managed directly in the admin under Settings → Hosting → DNS.
  4. SSL — automatic, HTTPS has to be lit up.
  5. Email addresses — info@, orders@, complaints@, support@ on your own domain. Without this you look like a garage operation and, according to the psychology of processing fluency, you raise suspicion even in cases where you're a perfectly legitimate business.
  6. Template and design — under Settings → Appearance you'll find the Template Designer, where you set the logo, colours, layout, favicon and banners. Shoptet offers free basic templates as well as premium variants from partners (Bluefox, Goodweb and others). Premium ones typically convert better thanks to better UX, and the investment usually pays back quickly.
  7. Product categories — Products → Categories. A logical hierarchy, SEO-friendly URLs without diacritics, category descriptions of 200–300 words (crucial for SEO).
  8. ⚠ VAT and taxes BEFORE products — Settings → Basic settings → Taxes. VAT payer/non-payer, tax zones and rates, the OSS regime for sales within the EU. If you change your VAT status later, you recalculate all prices manually. It's no joke.

The whole Shoptet setup step by step — from registration to the first test order — is in the deep-dive article Shoptet from A to Z.

PHASE 4: Products and content (month 2–3)

A rule nobody told me at the start: the product description is 80% of the copywriting on your entire e-shop, and most e-shops dispatch it in two sentences. Yet this is precisely where it's decided whether the customer adds to cart or goes back to Google.

The anatomy of a good product description

One rule that transforms product descriptions: Instead of "Weight 200 g" write "Weighs just 200 g — fits in any handbag". Feature vs. benefit. People don't buy grams, they buy the feeling that the thing won't be a burden. David Ogilvy already formulated this approach in Ogilvy on Advertising (1983), and every generation of copywriters since has reconfirmed it: a concrete, verifiable detail sells better than a general claim. His legendary headline for Rolls-Royce — "At 60 miles an hour the loudest noise in this new Rolls-Royce comes from the electric clock" — is a textbook example to this day. Instead of an abstract "luxury", one concrete detail that proves the luxury.

And one more rule you'll only believe once you test it: stick to three key claims, no more. The study by Carlson and Shu, "When Three Charms but Four Alarms", published in the Journal of Marketing (2014), showed that with three claims about a product, consumer scepticism stays low. The moment you add a fourth, the level of counter-argumentation shoots up and people start looking for the catch. The rule of three works because it's cognitively optimal — enough to feel comprehensive, not so much that it sets off the alarm.

The mandatory pages everyone forgets

These pages won't earn you any likes, but missing terms and conditions or a non-compliant returns policy will earn you a fine from the Czech Trade Inspection. Have them done by a lawyer, not downloaded off the internet.

And don't forget the small texts in the places most e-shops stay silent — the texts in the cart, at checkout, on the 404 page, in confirmation emails, in the footer. In Shoptet these are edited via templates, pages and HTML snippets under Settings → Appearance, or in the email templates. The default texts sound neutral to the point of corporate. When you rewrite them in your own voice, the brand shines even where the customer doesn't expect it. This is where what Daniel Kahneman in Thinking, Fast and Slow (2011) described as the peak-end rule comes into play — people remember an experience by its most intense moment and by its ending, not by its average. The footer, the 404 page and the order confirmation are the "endings" of micro-interactions. When your brand speaks here too, the resulting impression is fixed in a completely different way than if you'd let the defaults do the talking.

PHASE 5: Shipping, payments and logistics (month 2–3)

This is where it's decided how many carts complete the journey from "add to cart" to "order and pay". Badly set up shipping and payments are one of the most common reasons for an abandoned cart. According to long-term data from the Baymard Institute (baymard.com/lists/cart-abandonment-rate), the average cart abandonment rate in e-commerce is around 70%, and the main reasons are unexpected additional costs (above all shipping), forced registration and a complicated checkout.

The minimum set you have to have:

The detailed procedure for setting up shipping, payment gateways and connecting accounting/CRM is step by step in the deep-dive article Shipping, payments and logistics.

PHASE 6: Orders and communication (month 3)

Transactional emails are the most-opened communication you'll ever have with a customer. Open rate typically 50–70%, which is 4–5 times more than your best newsletter. And most e-shops send them in the default Shoptet template, which sounds like something out of 2014.

The emails you have to have ready and tested:

The last point deserves special mention. An abandoned cart isn't just a "customer error", it's a sunk cost moment in practice. The sunk cost fallacy (described, among others, by Arkes & Blumer in Organizational Behavior and Human Decision Processes, 1985) says that the more effort we've invested in something, the harder it is to abandon. A customer who has already spent several minutes choosing a product and filled in half the checkout is in a psychological position where a single well-timed reminder can rescue the conversion. According to ongoing reports from email providers (SaleCycle, Klaviyo, Omnisend), abandoned-cart automation — above all a quickly sent first email followed by a series of two or three messages — is among the campaign types with the highest revenue-to-effort ratio in all of e-commerce.

A copywriting rule that holds across the board: The email subject line must be personal and short. "Jane, your parcel is with the carrier 📦" works far better than "Your order number 12345 has been dispatched". A personalised subject line with a name, according to ongoing benchmarks from email providers, consistently lifts open rate. The reason is neurological: your own name is one of the strongest cognitive priorities — in psychology it's called the cocktail party effect. Even in a noisy room you snap to attention when someone says your name. In the same way, the brain scans a crowded inbox looking for relevant cues.

And don't forget SPF, DKIM and DMARC in your DNS. Without correctly set DNS records, half your emails end up in spam and you won't even know why. Test it via mail-tester.com — the goal is 9/10 or 10/10.

PHASE 7: Social media and sales channels (month 3)

Set up your social media BEFORE launching the e-shop, not the day before launch. Algorithms need time to start distributing you, and you need time to find your voice. Ideally 2–4 weeks ahead with regular posts.

The minimum set:

Connecting to Shoptet: Meta catalogue (Commerce Manager), an Instagram widget on the e-shop, Pinterest Rich Pins. Details in the deep-dive article Social media and XML feeds.

And a rule that holds across all platforms: consistency > perfection. Three posts a week for six months will deliver a bigger result than ten perfect posts in one month followed by nothing. Algorithms reward regularity, not genius, and the audience gets used to you gradually. There's one more principle behind this: the mere exposure effect (Robert Zajonc, Journal of Personality and Social Psychology, 1968) — repeated, non-aggressive exposure to a stimulus increases liking for it. The more times your audience encounters your brand without pressure to buy, the more natural it later feels for them to buy from you.

⏸ Stuck on one of the phases? That's where we come in.

If you've made it this far and you're thinking "am I supposed to do all this myself?" — the answer is: you don't have to. There are three types of people we help with this most often:

  1. Starting completely from scratch — we have a product, we have an idea, but everything on this page is sending us into a panic. We'll sit down with you, put together a complete launch plan and either walk you through it or do the work for you.
  2. Launched the e-shop themselves and got stuck — it all more or less stands up, but revenue isn't growing. An audit and a marketing setup that scales.
  3. Have a finished e-shop but want to get it going properly from the very first second — and know that one specialist for everything isn't enough. Complete marketing under one roof.

We take on only as many clients as our team can fully handle — we work slowly and deeply, not broadly. If we happen to have no capacity, we'll say so honestly and agree on a date.

I want to talk to LK Media →

PHASE 8: Tracking, marketing and advertising (month 3–4)

A rule that cost me two unpleasant conversations with clients: what isn't measured doesn't exist. If you don't measure where your customers come from and what they do on the site, you're running the e-shop blind. And you can't scale blind.

The stack you have to have before launch

  1. Google Tag Manager — a container for all the other codes. Deploy it first.
  2. Cookie bar and Consent Mode — legally compliant wording (not informational only), a "Reject all" button as visible as "Accept all" (mandatory under the Czech data protection authority).
  3. Google Analytics 4 — purchase, add-to-cart, checkout as conversions.
  4. Meta Pixel + Conversion API — not just the pixel. CAPI is a necessity today, not an add-on — without it you lose 20–30% of data from the iOS environment because of Apple ITP and App Tracking Transparency.
  5. Google Merchant Center + Google Ads — Shopping campaigns are the highest-ROI channel in the first year for most e-shops.
  6. Microsoft Clarity — heatmaps and session recordings. Free. It'll open your eyes.
  7. Sklik + Zboží.cz — still an important search channel in the Czech market (Seznam has long held around a quarter of the Czech search market), don't ignore it.
  8. Heureka, Glami, Google Shopping — XML feeds generated automatically by Shoptet, but you have to have your EAN codes filled in and the correct Google categories.

The detailed procedure for setting up the whole tracking stack is in the deep-dive article Tracking, GA4 and pixels from scratch.

Marketing channels in the real order of priority for a new e-shop

  1. Google Shopping + PMax — products are shown directly to people who are already searching for a product like yours. That's the Petty and Cacioppo ELM model (1986) in practice: the customer is on the central route of processing, actively looking for a solution, highly motivated — getting in front of them at the right moment is the highest ROI of all channels.
  2. Meta Ads — Facebook + Instagram, the main acquisition channel for brands that need to spark interest (the customer isn't actively searching for the product yet).
  3. Email marketing — automation + newsletters, the highest ROI over the long term from relatively low costs.
  4. SEO — an investment over 6–12 months, but stable organic traffic is the holy grail, because you don't pay click prices for it.
  5. Organic social media — builds the brand and community, it's not about direct sales.
  6. Influencer marketing — micro-influencers (1,000–50,000 followers) typically generate a higher engagement rate than big celebrities, because they come across as more authentic and activate the unity principle (Cialdini, Pre-Suasion, 2016) — the feeling of "one of us" is more powerful than the feeling of "someone famous".

PHASE 9: Testing and launch (month 4)

Test everything. Twice. On mobile and on desktop. Every error in live operation costs money and — worse — trust. And trust is far harder to rebuild than to lose, because according to prospect theory (Kahneman & Tversky, Econometrica, 1979) the pain of a loss is psychologically about twice as strong as the joy of an equivalent gain. When a customer is let down on their first purchase, they leave with an emotional loss they remember far longer than if they'd been satisfied.

What you have to test before launch

Soft launch → Grand opening

  1. Soft launch — a week for a close circle of people (family, friends, beta testers). Collect feedback via Google Forms.
  2. Fixing bugs — the things you didn't see in testing always crawl out here.
  3. Grand opening — a coordinated campaign: a 15% launch discount for 48 h, an email to your collected contacts, social media with a teaser, a PR release, paid ads on a low budget, an influencer campaign. Time-limiting the discount here isn't a marketing gimmick — it's time scarcity activating loss aversion. fMRI studies have shown that time pressure increases activity in the amygdala (the emotional centre) and decreases activity in the prefrontal cortex (logical deliberation). In other words: the customer switches from "Do I need it?" mode into "I have to have it before it's gone" mode. It works. But only if the urgency is real. False urgency (a countdown timer that resets when you refresh the page) is a dark pattern and irreversibly destroys trust.

On D-day, be there and watch the numbers in real time. The first day will tell you more than a whole month of planning.

PHASE 10: Post-launch and growth (month 5+)

Launching an e-shop isn't the goal. It's the start. The real work begins here. And it's the more enjoyable part, because you're now working with real data instead of estimates.

The three pillars that growth stands on

1. Retention — a new customer costs, on average, several times more than a returning one. This isn't a marketing phrase, it's a long-term benchmark documented, for example, in the analyses of Frederick Reichheld (Harvard Business Review, 1990; The Loyalty Effect, 1996). A loyalty programme, VIP discounts, birthday coupons, win-back emails after 60–90 days, review requests 7–14 days after delivery. If your retention rate is below 25%, the problem isn't in acquisition, it's in the product or the experience.

2. CRO (conversion rate optimisation) — A/B tests of headlines, CTAs, layouts, photos. Heatmaps in Microsoft Clarity (free) will show you where people click, where they get stuck and where they leave. The maths is brutal in a good way: improving the conversion rate from 1% to 1.5% is +50% revenue without a single extra crown into advertising. CRO is the highest-ROI tool in all of e-commerce, and paradoxically most e-shops ignore it because it isn't as "sexy" as launching a new campaign.

3. Scaling — expanding the product portfolio, international expansion (Shoptet supports multi-language), automation, increasing ad spend on successful campaigns. The rule goes: first find one channel that works, take it to the maximum, and only then add another. Diversification without the first pillar is scattering your energy.

The detail of the whole phase — from customer support through loyalty programmes to the CRO playbook — is in the deep-dive article Post-launch growth and CRO.

Three things that make the difference between an e-shop that survives and one that scales

After five years of running this process with dozens of clients, there are three things I see consistently in successful e-shops. And three things that are consistently missing in the ones that don't work:

  1. The brand has a soul, not just a logo. It's not about how good it looks. It's about whether the customer walks away from the first contact with the feeling that your brand means something. That's the Coca-Cola effect from McClure's study — the brand adds value to the product that genuinely changes the sensory experience. Find your why.
  2. The tracking stack is deployed from the very first minute. GA4, pixels, Clarity, Merchant — not in a month, not in six months, from the very first second of live mode. Without data you can't optimise, and without optimisation you can't scale. Full stop.
  3. The first three months are all about one channel that works. Not about five that sort of work. Find one marketing channel that brings you customers at a healthy CAC, take it to the maximum, and only then add another. Diversification without the first pillar is scattering your energy.

The whole series: How to start an e-shop 2026

This article is part of a seven-part series on starting an e-shop in 2026. The other parts:

FAQ

How much does it cost to launch an e-shop in 2026?

The minimum for a functional start on Shoptet is roughly €2,000–4,800 (plan, template, legal terms, product photography, basic marketing). A realistic budget with compliance, branding and first campaigns runs between €8,000 and €20,000. The key point is that the biggest line item isn't technology, it's time — 4–5 months of work before the first order comes in.

How long does it take to get an e-shop up and running?

With serious preparation, 4–5 months. First month strategy and legal, second month brand and design, third month products and technical setup, fourth month marketing and testing, fifth month soft launch. Anyone who promises an "e-shop in a month" is either skipping compliance or launching something that will fold in 3 months.

Is Shoptet, Shopify or WooCommerce better?

For the Czech market, Shoptet is the default choice — it integrates Balíky (Czech Post), Heureka, Zásilkovna and Shoptet Pay, it has Czech legal templates and it handles VAT correctly. Shopify is strong globally, but the Czech checkout, shipping and VAT require extra apps. WooCommerce is cheap, but it needs your own developer for updates. We launch 90% of new Czech e-shops on Shoptet.

What's the most common mistake when starting an e-shop?

The wrong order. People start with products and graphics, and then retroactively deal with VAT, tracking, the cookie bar and legal terms. The result — manually recalculated prices, missing data in GA4, a fine from the Czech Trade Inspection for a badly run sale. The right order: legal and VAT → tracking → products → marketing.

When does it make sense to hire an agency?

Whenever the cost of lost time outweighs the cost of the agency. If you're building the e-shop in the evenings alongside your main job, the first 3 months will eat up 300+ hours. With an agency that's compressed to 50–80 hours of your own time. If your hour as the owner is already worth a lot of money, it's simple economics.