DE
EN
1. Misconception: SEO is a project point, not a process.
What is saidPlease do some SEO, then it will be fine.
Why it's annoyingSEO reacts to content, technology, competitors, and updates. Everything is constantly changing. One time is not enough.
What you can sayWe plan SEO as a recurring task. Content, internal linking, rich snippets, loading time and indexing all require maintenance.
- SolutionMonthly content sprint, technical crawls, Search Console analysis, fixed backlog slots per release.
- ExampleProduct categories are updated quarterly. New FAQs, fresh examples, and tested facet logic.
- Mini PlaybookCrawl, prioritize, implement two levers per iteration, measure, continue.
2. Misconception: Performance comes later
What is saidFeatures first, performance last.
Why it's annoyingLater means more expensive. Performance is architecture. Images, scripts, render path, database queries. Whoever fixes that at the end ends up building twice.
- SolutionBudget for core web vitals in every sprint. Images with responsive formats, critical CSS, fewer third-party scripts.
- ExampleHomepage loads in under two seconds on 4G, product page in under three seconds. Measurement points with clear limits per template.
- Mini PlaybookMeasure, identify blockers, address Largest Contentful Paint and CLS first, then JavaScript and fonts.
3. Misconception: Headless solves every problem
What is saidWe're going headless, then everything will be flexible.
Why it's annoyingHeadless brings freedom, but also more responsibility. Two repositories, two deployments, API maintenance, caching rules, edge logic. Without a team and budget, it will be tough.
- SolutionDecision matrix: Number of channels, staff capacity, release frequency, content team. If the points fit, go. Otherwise, classic, but clean.
- ExampleHeadless for international brands with app, PWA and CMS teams. Monolith for local shops with small teams.
- Mini PlaybookStart with a small scope, decouple the product list and PDP, set up observability, and keep an eye on DDoS and cache hit rate.
4. Misconception: A plugin is always cheaper than development.
What is saidWhy not use a plugin? It saves time.
Why it's annoyingThird-party modules introduce dependencies: updates, licensing models, support. Initial savings are offset by ongoing costs or technical debt.
- SolutionTCO on accountLicensing costs for two years, update effort, security audit, exit strategy. If in-house development is more stable, then build it.
- ExampleShipping logic with special cases. A custom module was cheaper after six months than three incompatible plugins.
- Mini PlaybookProof of concept in staging, plan migration path, define fallback, activate monitoring.
5. Misconception: AI builds the shop alone
What is saidLet the AI write the shop's content.
Why it's annoyingAI helps with text, images, and code ideas. Architecture, security, law, payment methods, taxes, and migration require people with responsibility.
- SolutionAI as a copilot, not an autopilot. Style guides, prompts, mandatory reviews, tests, version control.
- Example: Pre-formulate product texts with AI, then check facts, adjust tone, and add internal links.
- Mini PlaybookPrompt templates, do and don't list, manual approval, A/B testing against manually created texts.
6. Misconception: Checkout is the same everywhere
What is saidUse a standard checkout, that will work.
Why it's annoyingCheckout depends on product range, countries, taxes, B2B rules, payment methods, coupons, and shipping methods. One size rarely fits all.
- SolutionCheckout map with paths. Guest checkout, login, addresses. ShippingPayment methods, confirmation. Measure bottlenecks, then simplify them in a targeted manner.
- ExampleAdjust field order, use Auto Fill properly, sort payment methods by conversion, and place trust signals only in the payment step.
- Mini Playbook: Analytics events, check for data entry errors, reduce validation, use inline feedback.
7. Misconception: Large pictures sell better
What is saidUse large pictures, they'll draw attention.
Why it's annoyingLarge files load slowly. Conversion suffers on mobile devices. Quality is good, but format efficiency is lacking.
- SolutionWebP or AVIF, clean sizes, correct art direction. Small thumbnails, separate zoom function, sensible lazy loading.
- ExamplePDP Gallery reduced its size from eight megabytes to under one megabyte. Clicks on zoom increased, loading time decreased, and bounce rate fell.
- Mini PlaybookImage pipeline with compression, defining breakpoints, automatic checks during build.
8. Misconception: Tracking works without consent
What is saidLet Analytics run, nobody will notice.
Why it's annoyingThe law doesn't disappear. Clean consent handling protects users and the team. Data quality improves when you measure accurately.
- SolutionConsent platform, server-side events, separation of functional and Marketing Purposes, meaningful designations.
- ExampleServer-side tagging with a clear data map. Less data clutter, better attribution.
- Mini PlaybookDocument data flows, perform regular scans, and test events against specifications.
9. Misconception: Tests are a luxury
What is saidWe skip the tests; we're fast.
Why it's annoyingWithout testing, something breaks after every release. Hotfixes cost time and cause frustration. Small tests prevent big bugs.
- SolutionTest pyramid: Unit for logic, integration for API, visual regression for templates, smoke tests for checkout.
- ExampleThree critical flows as an automated journey: shopping cart, checkout, cancellation. Errors are immediately visible.
- Mini Playbook: Set it up, run it on every merge, errors block the release, report the result in the chat.
10. Misconception: Internationalization is just language
What is saidTranslate the texts, then it will fit.
Why it's annoyingCountries mean payment methods, taxes, sizes, Legal textsShipping, address formats, holidays, product ranges. Language is one aspect.
- SolutionCountry matrix. For each country, rules regarding payment, tax, shipping, law, currencies, product descriptions, and units of measurement.
- ExampleSize conversion in PDP, country-specific payment methods above, local trust signals.
- Mini PlaybookDefine a pilot country, test the entire chain, report per country, only then scale.
11. Misconception: Search is a field and that's it.
What is said: Just add a search field, that's enough.
Why it's annoyingSearch requires synonyms, error tolerance, ranking, merch rules, facets, and analytics.
- SolutionTreat searches as a separate category. Query analysis, zero-hit list, maintain synonyms, sort facets by clicks.
- ExampleZero hits (less than two percent) from synonyms and autocomplete. Conversion via search increases significantly.
- Mini PlaybookWeekly query checks, new rules implemented, A/B tests for ranking.
12. Misconception: Design does development on the side.
What is saidBuild it first, we'll refine the design later.
Why it's annoyingWithout a design system, inconsistencies arise. In the end, the correction costs more than the initial planning.
- SolutionComponent library, tokens for colors, spacing, typography. Documented and reusable.
- ExamplePDP, PLP, and Checkout share identical components. Less CSS, faster releases, clear rules.
- Mini PlaybookDefine three core components, maintain them in Storybook, and make changes only via tokens.
Bonus: Three little truths that save projects
- Fewer options often lead to more purchases.Remove rarely used shipping methods and exotic payment methods if they disrupt the flow.
- Good defaults beat long formsPre-filled fields and logical sequences save time.
- Real-time feedback reduces dropoutsShow errors directly on the field. No surprises only after submission.
Checklist for your everyday life
- Plan for at least one performance ticket per release.
- Each new function is assigned a measurement point.
- Verify consent and tracking with documentation.
- No results found; check internal search weekly.
- Keep design tokens up-to-date and check them in pull requests.
- Align the SEO content plan with sales and support on a quarterly basis.
Short answer templates for difficult conversations
SEOWe achieve stable visibility through continuous content and technical maintenance. One-off measures do not produce a lasting effect.
Regarding performanceLoading time is part of the architecture. We plan for it per sprint; otherwise, we'll end up wasting budget.
To HeadlessHeadless is worthwhile with multiple channels and a team to support it. Otherwise, we deliver faster with a lean monolith.
Regarding pluginsWe calculate the total costs. If in-house development is cheaper and safer, we build the solution.
Regarding AIAI assists in the design process. Responsibility, quality, and approval remain with us.
Read on and delve deeper
your commitment
What misunderstandings do you hear most often in everyday life? What answers have helped you? Post an example, ask a question, or add to a mini playbook. You'll be helping other developers immediately.
FAQ in brief
I love this article! 😍 Finally, someone who says it! What really annoys me is this obsession with 'above the fold'. 'EVERYTHING IMPORTANT has to be above the fold!' Yeah, right, let's cram the logo, navigation, hero slider, all the USPs, trust badges, newsletter signup, special offers, and the first 10 products into the visible area. It definitely won't be overwhelming… Spoiler alert: Users can scroll! Mind blown! 🤯
This article is therapeutic! 😌 After 10 years in e-commerce, I thought I was the only one who got worked up about these things. Especially the part about 'innovative' features… A client recently wanted a shop where you could buy products by voice command. Cool idea, except: The target group is 50+ and they sell garden equipment. The overlap between 'Want to buy a lawnmower via Alexa' and 'Actually buys lawnmowers online' is… well, manageable. But the main thing is innovation! 🚀
FINALLY, some straight talk! What triggers me the most: 'We want to look like Apple, but our budget is 5k.' 🤡 Apple spends millions on design, employs hundreds of designers, and tests every pixel. Sure, we can just copy that. With WordPress. And a free theme. The disconnect some customers have from reality is truly appalling. This article will be my new standard reply!
Fantastic, finally someone said it! What always amazes me: customers who pinch every penny but then want to skimp on SSL encryption. 'Do we really need that?' Um, yeah? It's 2025! Chrome literally shows a warning if there's no SSL! But hey, the animated intro video for 5k was in the budget... priorities and all that. 🤷♀️
So much truth in one article! 👏 One more thing I'd add: the whole "minor changes" thing. "Could you just..." is the beginning of the end. Just the other day: "Could you just change the font?" Sure, except this font doesn't support umlauts, isn't licensed for the web, and completely ruins the corporate identity. But it's "just a quick change," right? This lack of appreciation for development work is driving me crazy. As if we just push buttons and the computer does the rest! 😤
This article should be carved in stone! 😂 The passage about feature creep is particularly apt. Last week: "Can we quickly add a wishlist, shopping list, AND comparison function?" – Two days before go-live! I then dug out the statistics: The wishlist in the old shop was used by a whopping 0,3% of users. ZERO POINT THREE PERCENT! But of course, it's "super important" and "every shop needs it." Sometimes I really wonder if clients ever look at their own analytics…
I've been working in e-commerce for eight years, and this article truly sums up ALL the pain a developer goes through. Especially the part about unrealistic time estimates… 'It's just a button!' – Yeah, right, a button that has to communicate with five systems, handle three currencies, and be GDPR compliant. 'Just a button, my ass!' 😤 But what annoys me most is when customers think, SEO It's something you just do on the side. "Just do some SEO" – as if that were the key. But it's a continuous process that takes months! Thanks for this article, bookmarked and will share if needed!
OMG this article is worth its weight in gold! 🏆 Especially the thing with the 47 product variations… We once had a client with phone cases: 15 phone models x 23 colors x 3 materials = over 1000 SKUs. Managing them was a nightmare! In the end, 5 variations accounted for 95% of sales. Lesson learned: Sometimes less really is more. But try explaining that to a product manager who insists on every single color shade… 😩
As a developer at an agency, I can confirm EVERY single point! Just last week it was: 'Can't we just copy Zalando's design?' – Um, no? Firstly, it's legally questionable, and secondly, they have a budget of several million and hundreds of developers. There are only three of us. 🙈 This article should be required reading for all shop owners!
@Robert Hartmann: That's exactly the problem! It's not about 'the latest technology,' but about functioning basics. If your shop takes 8 seconds to load, that simply won't be acceptable in 2025 – no matter how small the shop is. We just optimized a flower shop: reduced loading time from 7 to 1,5 seconds, and increased revenue by 35% in 3 months. Facts.
Oh God, I know that story about 'this has to be done quickly' all too well! 😤 Our boss recently claimed that a complete online shop relaunch can be done in just two weeks. With payment integration, inventory management connection, and everything else. I had to explain to him that the payment provider certification alone takes longer. His face was priceless! Thanks for this article; I'm forwarding it directly to management. Maybe then they'll finally understand that good development takes time.
Point 3 hits the nail on the head! I had a similar case last week. The client absolutely insisted on an "innovative" checkout with five steps and animation between each one. After two months of live operation and disastrous conversion rates, I had to revert everything back to the standard version. Sometimes, boring is simply better! 😅 Greetings from Pinneberg!
Finally, someone's said it! As a developer for a mid-sized furniture retailer, I can only confirm: most shop owners simply don't understand that performance isn't just a nice-to-have. 'Just quickly add 20 HD images to the homepage' – yeah, right, and then they wonder about the high bounce rates… 🙄
Absolutely perfect compilation!
I encounter point 2 (performance comes later) time and again. The saying "Whoever fixes this at the end ends up doing it twice" hits the nail on the head. From my own experience, I can add: It helps to include performance budgets directly in the Definition of Done. No story is finished if the loading time exceeds the agreed-upon threshold.
Another misconception I frequently encounter is: "Mobile is just responsive design." Many people think that if the site looks good on a phone, everything is fine. But it's about touch targets, reduced data usage, adapted interaction patterns, and a mobile-first approach to features. A mega menu might work on a desktop, but it's just annoying on mobile.
Regarding AI, one more point: I see the greatest danger in AI-generated content going live without review. Hallucinations in product data or legally questionable wording can be costly. Mini-playbook: Always send AI output through defined review stages, fact-checking of product data is mandatory, and legally critical areas (terms and conditions, cancellation policy) remain manual.
The checklist at the end is invaluable – I've added it directly to the team wiki!
What I'm still missing: Accessibility is underestimated. I hear "We'll do that later" or "Our target group doesn't need that" far too often. But it's not just legally relevant; it also improves the overall user experience and... SEO.
Strong point. Many believe e-commerce is a one-off project. You set up a shop, and then it runs. That's frustrating because operation, testing, and continuous optimization are the real work. Second misconception: Design solves everything. Without a clear information architecture, clean data, and stable performance, conversion rates plummet. Third misconception: Plugins fix every problem. You need processes, code quality, and monitoring. Fourth misconception. SEO You only do it once. Content ages, structures change, and so do search intents. Fifth misconception: A/B tests are only worthwhile with huge traffic. You can also learn with small experiments.
Specific checks for your team. Measure loading times and set targets. TTFB under 0,3 seconds, LCP under 2,5 seconds. Check for checkout errors daily. Clearly define events in tracking. Maintain a backlog for conversion drivers. Allocate a budget for maintenance and refactoring. What are your thoughts? What misunderstandings do you encounter most frequently in projects?