The final sum for a website depends not only on what it is made of. Two other things affect it, and the client largely controls both: how ready the materials are when the work starts, and how quickly decisions are made. So the simplest way to save without losing quality is right here, not on the design and the checking.
There is a second path — to take part of the work out of the project. Sometimes that is reasonable, and sometimes it costs more later. Below I will go through where the gap between the price on the site and the estimate comes from, what affects it, and where the line runs between reasonable saving and risk.
Why the base price and the final estimate differ
On the services and prices page, each type of project shows a price “from”: a landing page from 15 000 ₽, a business-card site or a portfolio from 30 000 ₽, a corporate site and an online store from 50 000 ₽, a Telegram app from 30 000 ₽, updates from 15 000 ₽. This is the lower bound for a typical scope. It helps you see the order of the costs, but it does not describe a specific project.
The final cost of website development comes together after the brief, when the design, the functions, the volume of content and the state of the source materials are clear. Two landing pages with the same structure can cost different amounts: for one the texts and photographs are ready, for the other they have to be collected, selected and processed.
At the same time, basic things should not turn into extra services. Adaptation for phone and tablet, ordinary request forms, standard testing and the agreed basic settings are part of the work. Anything beyond the agreed scope I first estimate and agree with the client, and only then do I do it. That way the estimate does not grow unnoticed at the end of the project.
When people look for inexpensive website development, they usually compare the “from” prices. The comparison is useful, but it is more honest to compare estimates for the same scope of work: what is included in the design, who prepares the texts and photos, which functions are needed and what happens after launch.
What actually increases the cost of website development
The reasons for a higher cost are fairly predictable, and most of them are visible even before the start. I divide them into four groups.
- Appearance.Custom layouts, complex animation, non-standard adaptation for different screens. If the site is built on Tilda, Zero Block belongs here too: in it you can assemble almost any composition, but each such section has to be thought through separately and checked on every screen.
- Functions.Quizzes with branching questions and a cost calculation, calculators, personal accounts, catalogs with complex filters and product variants, custom code in HTML, CSS and JavaScript.
- Connections with other services.CRM, online payment, delivery services, external APIs. Each service has its own documentation, access and ways of passing data, and all of that has to be set up and checked on real scenarios.
- Content and changes along the way.Preparing texts, selecting and processing photographs, bulk upload of materials, new pages and language versions, revisiting decisions that were already agreed, and shortened deadlines.
The first three groups are usually discussed at once: it is easy to talk about the visible part of the site. The fourth is remembered later, although its volume can also be estimated before the start, if you show in advance what state the materials are in.
How the readiness of materials affects the cost
Texts, photographs, a logo, contacts, prices, product data — a site cannot be assembled without these, even if the design is ready. The question is in what state the materials reach the developer.
Take photographs. If Google Drive holds a large archive with no selection and no clear names, before design starts it has to be downloaded, unpacked, sorted into sections, renamed and matched to pages or products. Then, from many similar frames, the suitable ones have to be chosen and the choice agreed. All of this is real work, and it is counted separately in the estimate. If you choose 5–10 photographs for the cover and the gallery in advance, the design can be agreed faster, and the time spent on selection and repeat processing goes down.
Frames for the first screen, background blocks, sliders and previews stand apart. Text is often placed over them, so the frame needs free space and a calm background. Darkening a photo with a translucent layer does not always work: sometimes you need a local light correction, another frame, or an agreed neural-network treatment. I discuss such steps in advance. An image changed by a neural network must not be presented as a documentarily accurate photograph of the object.
With texts the story is similar. If information about the company is scattered across correspondence, old presentations and documents with different figures, someone has to collect it and check it. In an online store the same work applies to products: names, specifications, variants, prices and photos. A substantial part of the effort in a store is often tied to product content, so a tidy product table noticeably reduces manual work.
The client does not need to understand the technical part. Their task is to give information about the business, help choose the materials and say what must be on the site. The developer decides how to show it.
Why decisions made on time speed up the work
A site is assembled in stages: structure, prototype, design, functions, content, checking. Each next stage rests on the decisions of the previous one. If an answer to a question is delayed, the work stops or goes on under assumptions that will have to be redone later.
A separate cost item is changes to what has already been approved. A new block in an agreed layout, a different menu structure after the pages have been assembled, a replacement of photographs that were already approved — this is a new volume of work. Fixing a mistake does not belong here: if something was done differently from what was agreed, that is part of my work.
Simple habits help reduce such changes. On the client’s side it is better to name one person who approves decisions. It is more convenient to collect comments in one list, rather than send them one by one in different messengers. The structure and the prototype are worth looking at carefully before the design is approved: at this stage it is easiest to correct the order of the blocks.
There is also a direct link to the price. Shortened deadlines are one of the factors that increase the estimate. If decisions are made without long pauses, the project fits the ordinary schedule, and there is no need to pay for urgency.
What you can save on, and what it is better not to
The question of how to save on website development usually starts with a wish to remove something. What you should remove is what the site can do without and still do its job.
Not every page needs its own layout. When several service pages are built the same way, they can be assembled from one template, and the visitor will not notice. The number of unique templates directly affects the volume of design.
If the site is on Tilda, most of the pages can be assembled from standard blocks. That is how I make, for example, a landing page on Tilda: the owner later changes the price, the date, a photo or a review in the ordinary editor, without a programmer. I use Zero Block where it really cannot be done without it. The saving here is double: less work during development and fewer requests for small edits after launch.
Functions can also be split into queues. A personal account, a complex calculator or a quiz with branching is useful when the sales process does not work without it. While there are few requests and they are similar to each other, a simple form will do, and the complex function can be added after launch, when it is clear that it is needed.
It is reasonable to keep animation where it helps people read and choose: to show the order of stages or to draw attention to a button. Effects for the sake of effects increase the estimate and give the visitor almost nothing.
And here is what it is better not to save on:
- checking the site on a phone and on different screens;
- testing the forms: a request must arrive where it is expected;
- the structure and the prototype: mistakes found later are more expensive to fix;
- the quality of the source files for the first screen, because a weak frame will still have to be replaced or processed for a long time;
- the basic SEO setup: headings, meta tags and page addresses.
Another doubtful saving is to refuse preparation of the materials when you have no texts and photos of your own. The site then launches with temporary texts and random photos, and soon it has to be redone. It is more worthwhile to include the preparation in the estimate honestly from the start.
Which tasks are worth automating
Automation costs money at the start and pays off where a task repeats. A one-time upload of several pages is simpler to do by hand. Regular updates of a large catalog are already a reason to think about a script.
It usually makes sense to automate:
- uploading and updating products from a table, if the range is large or changes often;
- passing requests to a CRM, Google Sheets, email or Telegram, so that nobody copies them by hand;
- updating prices, stock or a schedule from one source, if this data changes often;
- working scenarios with AI tools where the same processing of text or data repeats.
An important condition for automation is structured data. A product table with separate columns for the name, the article number, the category, the price and the photo is loaded into the catalog by import, without filling in each card by hand. The same information, scattered across correspondence, first has to be brought into such a table by hand.
What I can take on
Not every client has the time and the people to prepare the materials. In that case I take this part on myself: I collect and organize the content, select and process the photographs, prepare the graphics, and build complex functions and integrations. This is ordinary work, only its volume needs to be included in the estimate. That is why I discuss it in advance, and do not add it along the way.
If the main development is already finished and you need to update pages, fill the catalog or add new blocks, there is a separate line of work for that — updates and content.
To see how to make your project without extra costs, tell me about it in the quiz on the site. I will reply in person, clarify the details and propose a structure and an estimate that take your materials and tasks into account.