Building an online store from scratch is a task that seems simple until you really get into it. Then questions arise: which platform to use, how much does it cost, how to choose the right contractor, and what will happen in a year, when the business grows?
Author: Egor Ivanov, COO of Bulltech
In this article, we’ll explore the full cycle of online store creation: from concept to launch and development.
Before you start looking for a development team, researching a CMS, or looking at pricing, there’s one crucial step. It’s often overlooked, and people later wonder why developing an online store took twice as long as planned.
You need to answer three questions:
2. Who are your customers and how do they buy? Everything depends on this: the catalog structure, the interface, payment methods, and delivery.
3. What processes are already in place in your business? If you already have 1 °C, inventory management, and a CRM, they will need to be integrated, which will impact the choice of platform and the cost of online store development.
Without answers to these questions, any solution, even the most technically advanced, will miss the mark. A good contractor will begin the project from this stage. However, if you’re immediately quoted a price and timeline without researching the business, that’s cause for concern.
Once you’ve defined the business objective, the analytics phase begins. This is often underestimated, but it determines the success of the entire project. If you make a mistake here, you’ll have to redo it later, which is expensive and time-consuming.
At this stage, we study:
• Product range. How many products are in the catalog? Are there variations (color, size, material)? How often are prices and inventory updated?
• Target audience. Who buys, where are they located, and what devices do they use to access the site? If most shoppers use smartphones, a mobile version becomes critically important.
• Competitors. Who is already selling similar products? What are their strengths and weaknesses? What can be improved?
The result of this stage is a clear understanding of user scenarios. How will customers search for products? Through the site search, through filters, through product categories? What will they do after adding them to the cart? How will they place an order?
Without this knowledge, you risk ending up with an online store that technically works, but no one can find or buy anything from it.
Based on the analysis, a technical specification (TS) is created. It describes the online store’s intended structure, functionality, integrations, design, and performance requirements.
A good TS is the foundation. If you don’t want to discover during development that the chosen catalog structure doesn’t allow for complex filtering by product characteristics, the TS must be detailed. If you want the store to withstand peak loads, this must be spelled out in the requirements.
An example of a bad TS: «We need an online store like our competitors, only better and cheaper.» Such a TS won’t help you or the developers. It will result in the project being done «by eye,» and the result will be far from what you expected.
Now to the most important and pressing question: what to use for an online store? Choosing a platform is a decision that will determine the project’s fate for years to come. And here I have a clear position.
Ready-made CMS
Ready-made CMS: OpenCart, WooCommerce, 1 °C-Bitrix, Shopify. This is the fastest and cheapest way to create an online store. The basic functionality is already there: a catalog, product categories, shopping cart, checkout, personal account, favorites. You can set it up in 3-6 weeks and launch with a small budget.
Why are they so popular at the start? Because they deliver quick results. The client sees a visually functioning store, and that’s inspiring.
But the problem is that these solutions have serious limitations:
2. Hidden costs. A CMS seems like a cheap option. But full functionality almost always requires paid modules: integration with 1 °C, CRM, delivery services, advanced filters, custom pricing, and SEO optimization. Collectively, these enhancements easily cost 80,000-150,000 rubles per year.
3. Scaling issues. As your business grows (you’re planning for growth, otherwise, what’s the point?), the platform will start to slow down. A store with over 1,000 products on WooCommerce without serious optimization starts to slow down with just 20 concurrent users. Ultimately, you have to move to a dedicated server, which is quite expensive to maintain.
4. Vendor dependency. If the CMS developer stops releasing updates or changes its policy, you’ll find yourself in a difficult situation. And if you’ve heavily customized the platform, updating the core to a new version will be a serious problem due to conflicts with your changes.
When is a CMS justified? When you have a small product range, simple sales logic, and don’t plan on complex integrations or scaling.
Custom Online Store Development
Custom online store development means creating a solution from scratch tailored to your business processes. The architecture is built to fit your business, not the other way around. You have complete control over the code, security, and loading speed. The store is easily scalable: add a new warehouse, and the system adapts within a week.
Why do I believe that custom is almost always better?
Because an online store is more than just a storefront. It’s a sales tool that should grow with your business. If you plan to expand your product range, work with wholesalers, implement complex integrations, and build a digital ecosystem, a custom solution pays for itself.
Yes, the starting budget for a custom project starts from 1.5–2 million rubles. But the cost of ownership is lower: no paid modules, fewer bugs, faster development. Such an online store pays for itself through increased conversion, automation, and flexibility. A case study from our practice: the GK Sport online store. We didn’t simply take an existing CMS and apply a design to it. We redesigned the design system, integrated it with MoySklad and SDEK, and automated product cards — they are now automatically generated with labels like «Discount,» «New,» or «Hit.» We also implemented flexible regional settings: payment and delivery methods depend not only on the region but also on the order amount. The result: increased conversion, improved SEO, and reduced manual labor.
You can read the full case study at: https://bulltech.ru/projects/gk_sport
This level of automation and integration is either impossible or very expensive to implement with existing CMSs.
Conclusion: if you’re planning a serious business with long-term growth, custom development will pay off. If you need a quick start with minimal investment, you can start with a CMS, but be prepared for the fact that after a year or two you will hit a ceiling.
The catalog structure is how products are organized on the website. It affects searchability, navigation, filtering, and, of course, SEO.
The correct structure of an online store is based on an analysis of the target audience and product range. Even if you have 50 products, you need to compile numerous categories and generate traffic to ensure product visibility in search engines. However, if you have 50,000 items with variations (color, size, material), you’ll need a complex database architecture and a well-thought-out system of categories and filters. As a result of such measures, the online visibility of your resource will significantly increase your business.
What to pay attention to:
• Product categories should be logical and understandable to the user.
• Filters and navigation should help you quickly find what you need. For clothing: filters by size, color, material, and brand. For electronics: filters by specifications, manufacturer, and price.
•The website search must work correctly and account for typos and synonyms.
•The product page is the decision-making point. It should contain high-quality photos, specifications, attributes, price, availability, reviews, and an «Add to Cart» button. At GK Sport, we reimagined the product page and automated its creation through the admin panel.
Before starting design and development, you need to create a prototype of the online store. This is a diagram that shows how the user will navigate the site, from the homepage to the checkout process.
Prototyping helps validate hypotheses before coding. We create interactive prototypes—clickable models of the future product—and test them on real users. We look at where they click, where they stop, and what causes difficulties.
At this stage, user scenarios are developed:
• How does a customer search for a product?
• How does one add to cart?
• How do they place an order?
• How do user registration and authorization work?
• How does the personal account work?
Prototyping allows you to find errors in logic and flow before the developer writes a single line of code. This saves money and stress.
Once the structure and scenarios are approved, UI design begins — creating the visual appearance of the online store. This isn’t just about making it look pretty. It’s about creating an interface that helps the user achieve their goal: finding a product, making a decision, and purchasing.
In good UI design, every element on the page has a purpose. Headlines attract attention, buttons are noticeable and clear, and forms are easy to fill out. And, of course, a responsive design is created to ensure the online store works seamlessly across all devices, from large monitors to smartphones.
Once the design is approved, online store development begins. Any website has two parts: the frontend, which is what the user sees: the catalog, cards, and shopping cart; and the backend, which processes orders, products, and users.
At this stage:
• Pages are laid out using mockups.
• Logic is developed: how the shopping cart works, how orders are placed, how user registration and authorization occur.
• The admin panel is configured to manage content, categories, products, and prices.
• Filters, site search, favorites, and product comparison are implemented.
It’s important that the backend be designed with a safety margin. Your online store will operate under heavy load, for example, during the sales season, so it won’t crash.
An online store isn’t an island. It must be connected to external systems, without which a business can’t survive:
• Integration with 1 °C or ERP. Bilateral exchange of products, prices, inventory, and orders. To prevent customers from purchasing an out-of-stock item.
• Payment system and online cash register. Online payment, online acquiring, automatic receipt sending.
• Delivery services. Calculate delivery costs and delivery times directly in the shopping cart via the SDEK, Boxberry, and Russian Post APIs.
• CRM. To automatically transfer orders to the customer management system. Without automation, managers would manually transfer data from email to Excel, lose orders, and forget to call customers back.
At GK Sport, we integrated the store with MoySklad and SDEK, set up SMS verification and automatic email messaging, and all this ensured an end-to-end process from order placement to delivery.
Integrations are a complex and crucial step. And how well they are configured determines whether your online store will be a mere showcase or a fully-fledged sales tool.
SEO for an online store isn’t something you need to do after launch. SEO is established during the design and development phase.
What’s important from the start:
• Human-Readable URLs (HRURLS). Instead of site.ru/product? id=123, use site.ru/krossovki-nike-air-max.
• Microdata. For products, prices, reviews, and breadcrumbs. This helps search engines better understand your content and display it in rich snippets.
• Sitemap and robots.txt. To ensure search engines correctly index your site.
• Canonical. To avoid duplicate pages.
• Web page load speed. This impacts SEO, user retention, and conversion. Core Web Vitals (LCP, INP, CLS) aren’t just technical metrics; they’re ranking factors.
If you incorporate SEO into your project from the very beginning, your online store will start with a good search ranking.
Before launching, an online store needs to be thoroughly tested. Testing isn’t just about checking if the «Buy» button works. It’s about checking how the system behaves under a large number of users, whether shipping is calculated correctly, and whether the server crashes while loading images.
Types of testing:
• Functional. We check that all buttons, forms, filters, the shopping cart, and checkout work correctly.
• Cross-browser and responsive. We check that the website works on all browsers and devices.
• Load testing. We check how the system behaves under peak load (for example, on a sale day).
• Integration. We check that integrations with 1 °C, payment, and delivery work correctly.
Once testing is completed and all errors are fixed, the online store is ready for launch.
What happens at this stage:
• Transfer to the production server and domain.
• Configuring an SSL certificate. Without it, customers won’t trust the site, and search engine rankings will be lower.
• Connecting analytics: Yandex Metrica and Google Analytics.
• Submitting the sitemap to Yandex.Webmaster and Google Search Console.
• Final check of all integrations and payments.
Launching an online store is not the end, but only the beginning.
After launch, data collection begins. Ecommerce analytics will help you understand what’s working and what’s not.
What to track:
• Conversion. How many visitors turned into customers?
• Sales funnel. At what point do users leave? Maybe they add items to their cart but don’t complete the order?
• User behavior. What do they search for, what do they click on, where do they linger?
• Traffic sources. Where do customers come from: search, advertising, social media?
Based on this data, you’ll improve your online store. It’s an ongoing process.
An online store isn’t complete the day it launches. It needs to be developed based on data and business objectives.
What’s included in support:
• Technical support. Problem solving, platform updates, security and loading speed monitoring.
• Product development. Adding new features, improving UX, conversion optimization, and SEO.
• Content and visual support. Updating the catalog, adding new categories, and running seasonal promotions.
At Bulltech, we always support projects after launch. Because an online store should grow with the business.
Forget promises to «build a website in two weeks.» This will either be a landing page or low-quality work that will have to be redone in a month.
The actual cycle for creating an online store is 3 to 6 months.
• Month 1: Design and approval of the specifications. This is the foundation. If you rush here, you’ll run into problems during the programming stage.
• Month 2: Design and layout.
• Months 3-4: Core and integration development. This is the longest and most unpredictable stage.
• Two weeks to a month: Testing and bug fixing.
• The never-ending process: Adding products to the website.
Custom development takes longer than a CMS. But it produces results that will last for years.
The cost of developing an online store depends on three factors:
• Platform type. Ready-made CMSs are cheaper. Custom development is 3-5 times more expensive.
• Catalog size. 100 products or 50,000 with variations—these are completely different amounts of work.
• Deep integrations. Linking a website to a single payment system is considered standard. However, setting up real-time synchronization with 1 °C, multiple delivery services, CRM, POS, and marketplaces is a separate area of work.
Estimated cost: CMS from 300,000 to 900,000 rubles, custom development from 1.5-2 million rubles and up.
But don’t just look at the development price. Consider the cost of ownership. A CMS may seem like a money-saver at the start, but after a year or two, you’ll start paying for paid modules, optimization, and customizations. Custom pays for itself through business growth.
I’ll list the most common ones so you don’t repeat them:
1. Starting with design, not analytics. It’s like building a house without a foundation. Halfway through development, you’ll discover that the catalog structure doesn’t allow for the necessary filters.
2. Wanting everything at once. An MVP isn’t a «cheap version»; it’s a minimal set of features to test a hypothesis. Start with the basic functionality and refine it based on data.
3. Choosing a platform based on the «cheapest» principle. Cheaper doesn’t always mean more profitable. Remember the hidden costs of paid modules and architectural limitations.
4. Forgetting about SEO. If you didn’t consider SEO during the development stage, you won’t get traffic and customers in the first few months.
5. Not testing. If you haven’t tested integrations and load testing, prepare for surprises on sale day.
6. Thinking that after launch, everything is over. An online store needs to be developed and maintained.
Building an online store from scratch is a complex process that requires a systematic approach. Start with analytics. Ask developers to design the structure and scenarios. Choose a platform not based on «cheapest,» but on «whether it’s right for your business.» Build SEO and integrations into the development process. Test. And remember: launching is just the beginning.
Ready-made CMSs are great for getting started. But if you’re planning a serious business, custom development will pay off with flexibility, automation, and increased sales. At Bulltech, we create online stores that function as sales tools, not just a storefront. And we always start with analytics to anticipate all scenarios. Because the success of an online store depends on planning everything from the start.
Get an estimate within 24 hours!
