Answer in brief
Two Shopify catalogue entries answer two different questions: can you sell at all, and can your team change the shop without a developer. Plus the SEO shape of categories, filters and product cards.
Start from the list of edits you will want in March, not from the demo store
Storefront decisions usually begin in the wrong place: a demo, a scroll, a feeling about somebody else's hero image. Begin instead with a sheet of paper and a year of your own calendar. What will you have to publish between now and next summer? A collection page for the seasonal drop. A size guide written to end the five emails that arrive every week. A bundle block before the holidays. A campaign landing page that has to be live on a Friday afternoon rather than the following Tuesday. Ten lines are enough to change the conversation.
Now put a name against each line — the person who would actually publish it. Most lines take a name easily, because the change is typing: a product, a price, a photograph, one more menu entry. A smaller group takes a name too, because the change is configuration: reordering the home page, swapping a banner, switching a section off for a week. What decides your budget is the remainder — the lines that take no name at all, because the block in question is not in the theme's settings and will not appear there by itself.
The catalogue publishes two answers to that remainder. The first is Turnkey online store on Shopify, listed under E-commerce at $800 – $2 500 or 40 000 – 120 000 ₽, with a stated band of 25-40 hours and a difficulty of Experienced. The second is Custom Shopify theme for your brand, in the same category, at $2 000 – $6 000 or 80 000 – 250 000 ₽, with 50-90 hours and a difficulty of Advanced, the highest grade the catalogue uses.
They answer different questions, which is why comparing the two prices head to head misleads. The turnkey entry answers whether you can sell at all: its description promises a working shop with catalogue, payment, shipping and analytics, taking its first orders on the day it is handed over. The custom theme entry answers whether your own team can keep changing the shop afterwards, because it promises custom sections and a page builder inside the admin panel. Buy the answer to the question you actually have.
What the turnkey build actually hands over on launch day
The turnkey description is best read as an inventory. Catalogue, payment, shipping and analytics, wired together rather than merely present, with the shop taking real orders on the day it changes hands. Then read the clause most buyers skim: no hosting, no layout contractors, no endless engine updates. Those are three recurring costs that a hosted platform absorbs on your behalf, and their absence from the second-year budget is as much a part of what you are buying as anything on the storefront.
The published band of 25-40 hours is a band for setup, and setup has an edge you should be able to point at. On one side of it: importing products and their variants, giving collections a structure, configuring checkout, drawing shipping zones that match how parcels really leave your building, tax and currency, a payment provider wired up and proven with one live transaction, analytics whose purchase events genuinely fire, and a bought theme tuned as far as its own settings reach. On the other side: everything made of words and photographs. Product copy, a shoot, migrating old reviews — each is its own brief and its own invoice, and the argument that surfaces at handover almost always lives exactly on that line.
Experienced, one grade below the Advanced the custom theme carries, is an honest label for the risk profile. The route is well travelled and the technology is predictable; what makes this work expensive when it goes wrong is that it goes wrong silently. A variant model too narrow for the range you actually sell has to be rebuilt later at more than the cost of the original build. Shipping zones drawn on guesswork lose money quietly, parcel after parcel. Analytics installed without purchase events leaves a year of marketing decisions resting on figures nobody can verify. None of the three is visible on the storefront the day it is handed over.
Against that, the freedom a bought theme leaves you is real and worth naming before anybody offers to expand it. Products, collections, prices, photographs, navigation, the copy inside sections the theme already provides, the order those sections appear in on the home page, and whatever colours and typefaces the settings expose. One clear range and an ordinary promotional calendar can live inside that set for a year or two without strain, and there is no prize for buying headroom you have not yet learned to use.
The wall inside a purchased theme is structural, not visual
A bought theme does not fail on the day you stop liking it. It fails the first time a campaign brief arrives that the settings panel cannot express: a side-by-side comparison of four models, a size-and-care module reading from product metafields, a gift finder that groups the range by price instead of by collection, a bundle whose price is not the sum of its parts. Nothing on that list is exotic. Every item on it is a build rather than a setting, and that distinction is where the money starts moving.
From there the options narrow to three, and each carries a price tag written in a different currency. Dropping the idea costs nothing you can see, which is exactly what makes it the most expensive habit of the three. Reaching for an app costs a subscription that renews itself, an extra script on every page and a new dependency that can break at the next theme update. Commissioning the block costs money and, more painfully, a place in somebody else's queue: from that moment your campaign moves at the speed of a developer's calendar.
Totalled across a year, the invoice is the smallest part of the bill. The rest is the launch that slipped a week, the seasonal page nobody built because the estimate came back on Thursday and the drop was Monday, the promotion that went out in the layout that happened to be available rather than the one that was designed. None of it reconciles against a budget line, and a shop can carry that cost for years while every number it does track looks unremarkable.
Read the custom theme entry with that pattern in mind and it stops looking like an aesthetics upgrade. Its description names three things — its own sections, its own page builder in the admin panel, speed above standard themes — and then states the outcome as a staffing fact: the client's marketer assembles promotional pages himself, without a developer. Whether $2 000 – $6 000 and 50-90 hours are worth it is therefore not a question about design at all, but a question about how many of those briefs you expect to receive next year.
Where the extra hours in a custom theme are spent
Set the two bands beside each other — 25-40 hours against 50-90 hours — and the gap has a specific shape. The largest single item inside it is turning every section into a configurable block with its own settings schema, so the theme editor shows it with genuine fields: a headline, an image, a link, a choice of layout, a count of items to display. A section without that schema is markup, and markup means a ticket every time somebody wants a different headline on it.
The second item is everything the word builder quietly assumes. Sections have to be available on more than one template, survive being dragged into an order nobody sketched, remember settings per page rather than globally, and hold their layout when a colleague stacks three of them in a row late on a Friday. A large share of the hours goes there, and it is the one part of the deliverable that no screenshot in a sales deck can demonstrate to you.
Speed is the third item and the only one you can put your own number on. A commercial theme has to serve every kind of shop that might buy it, so it ships with the machinery for features you will never switch on and frequently loads that machinery regardless. A theme written for one brand carries one brand's cases. That is what the catalogue means by speed above standard themes, and unlike a claim about taste it can be checked on your own pages the week after handover.
There is a sanity check here that costs a minute. Whatever hourly rate you already pay for development work, run it against 25 and 40 hours, then against 50 and 90, and lay the four results next to the published ranges of $800 – $2 500 and $2 000 – $6 000. This is an example of arithmetic to repeat with your own numbers rather than a measurement of your project, but the results usually tell you which conversation to have with a supplier — one about price, or one about scope.
Category pages are your landing pages, whether you treat them that way or not
People searching in order to buy rarely type a product's exact name. They type a class and a qualifier — a material, a colour, a use, a price bracket — and the page that answers is a category, not the home page and not a single product card. So when a shop spends everything on product photography and leaves its category pages as bare grids, that entire class of demand arrives at pages with nothing to read and nothing to rank on, and nothing on the site looks broken while it happens.
The catalogue entry for this work is SEO for an online store: categories, filters and product cards, listed under E-commerce at $600 – $3 000 or 40 000 – 150 000 ₽, with the published note that a deep audit of a large shop reaches 300 000 ₽ at market rates. Its stated duration is 3–7 days for a catalogue of up to 5000 SKU, its difficulty is Advanced, and it is the only one of these four filed outside web development, in the SEO and traffic group.
The description states the deliverable plainly: categories and product cards stop competing with each other, receive proper texts and correct markup, and filter indexation becomes controlled, so that traffic grows on the commercial pages that bring orders. Every clause in that sentence is a separate piece of work, and the first one, stopping your own pages from competing, is the one merchants almost never realise they are paying for.
Turned into a specification, a category page owes you an editable introduction, a second text block beneath the grid, its own title tag and meta description, an H1 that is more than the menu label, breadcrumbs, and a URL built to survive the next re-merchandising. Two items on that list decide who owns the page afterwards. If the introduction refuses headings and links, or if the title tag turns out to be the agency's field rather than yours, you have located a recurring cost that no build estimate will ever show you.
How filters multiply a small catalogue into far more URLs than products
Open your own filter panel and count the values in each facet, then multiply those counts together: colour by size by brand by price bracket by availability. Remember while you do it that most storefronts let a shopper tick several values inside a single facet at once, which pushes the total up again. Work it through with your real figures and the number of reachable addresses generally overtakes the number of products you sell. Treat that as an example of arithmetic you can redo in a minute with your own catalogue rather than a statistic about your shop; it is there to show the shape of the problem, not to measure it.
A small minority of those combinations has earned a page. Black leather sneakers is a phrase people type and then buy from, so it deserves a URL, a heading, a paragraph written for it and a line in the sitemap. The rest — three colours ticked, two sizes ticked, the price slider dragged somewhere near the middle — has earned nothing, and a crawler let loose among them spends its budget on pages that differ from each other by almost nothing while the pages that actually sell wait their turn.
Controlled indexation, in the catalogue's phrasing, means a written ruleset rather than a queue of individual judgements. Which parameters may produce an indexable address. Which fold back to the clean category by canonical. Which are blocked outright. What order parameters are written in, so that a single selection can never resolve to two different URLs. How the sitemap reports the result. And the rule teams routinely forget to ask for: the promotion path, describing what happens when a filter combination starts selling and deserves to become a curated collection with a heading and text of its own.
Notice that the duration in the catalogue is expressed per catalogue size rather than per page count: 3–7 days for a catalogue of up to 5000 SKU. Facet work scales with how many products and attributes exist, not with how the shop looks, which is why this entry is priced independently of both the build and the theme, and why the same shop can need it again after a large range expansion.
When a product card and its own category fight over the same query
Cannibalisation is undramatic and expensive. Two of your pages target the same query, so neither accumulates the signals it needs, the search engine alternates between them, and the position that would have justified the entire catalogue never arrives. It happens most often when product titles repeat the category name word for word, and when supplier feed text lands unedited on every card in a range, leaving a whole range of pages that differ only by a model number.
Deciding the split takes an afternoon; applying it takes considerably longer. Exact model names, article numbers and questions about one specific unit belong to the card. Class plus attribute — the shape of a query typed by somebody who has not chosen anything yet — belongs to the category. Written down, that division becomes checkable, and checkable is the only condition under which product titles, headings and internal links stop contradicting it a few hundred pages at a time.
Correct markup appears in the catalogue description as two words, and on a shop it resolves into a short, checkable list: product structured data carrying price, availability and reviews; breadcrumbs that agree with the hierarchy a visitor actually walks; item lists on the category pages. Machines read all of it before any person sees the page, and it is what feeds the product surfaces where a listing either appears or does not. A mistake there renders perfectly on screen, which is exactly why it can sit undetected while the traffic reports get worse.
The editability test applies here too. Ask, in the admin, to change one card's title tag, its meta description, the text blocks below the gallery and the alt text on its images, and to add a short answer block to a card that receives the same question every week. If any of that requires a developer, you have just found a recurring cost that will never appear in the build estimate you are comparing.
The order form and the pay button are billed as their own job
Application forms and payment acceptance on the website is the fourth entry, and the only one filed under Integrations rather than E-commerce: $250 – $800 or 10 000 – 35 000 ₽, a band of 8-20 hours, difficulty Experienced. Every figure on that line is the smallest of the four, which is a poor guide to how much rests on it. This is the work that decides whether an enquiry you have already paid an advertising platform to generate ever reaches somebody able to answer it.
Checkout arrives with the platform, which is why this entry is the one most often struck off a budget. Then the exceptions turn up. Wholesale enquiries. Made-to-order requests. Pre-orders and B2B quotes. Repair and service bookings. Any landing page published outside the storefront. None of those passes through the cart, and every one of them needs a form that lands somewhere reliable. A fair number of merchants also need a pay button that works on a page which is not the checkout, and that is a separate integration with its own ways of failing.
Three destinations are named in the description — CRM, Telegram and a spreadsheet — and each earns its place for a different reason. Ownership is the CRM's job: an enquiry that lands there has a person attached to it and a state it can move through. Speed is Telegram's job, because a lead that reads your reply within the hour behaves differently from one that reads it the next morning. Evidence is the spreadsheet's job, and you will want it the first time somebody insists that nothing came through. The description also promises that the letter arrives, which on a real domain is a set of sender settings rather than an act of optimism.
Protection from spam bots closes the description, and it is what keeps every clause before it true. The failure follows a predictable sequence: automated submissions arrive, the alert channel fills with them, the team quietly stops reading the channel, and a genuine enquiry then dies inside a system that is technically working perfectly. At 8-20 hours this is the shortest piece of work in the set of four and the one whose absence leaves the faintest trace, because no analytics report ever mentions an enquiry that no human read.
If the budget only covers two of the four, buy them in this order
Add the published low ends and you get a usable planning figure. The turnkey store at $800 plus forms and payments at $250 is $1 050. Add category, filter and card work at $600 and you are at $1 650. All four entries at their lowest published figures come to $3 650; all four at their highest come to $12 300. That is arithmetic on published bands rather than a quote for your shop, and any real estimate lands inside those bands for reasons specific to your catalogue.
Sequence matters more than the total. The turnkey build goes first for a reason nobody argues with: a shop that cannot take money is not a shop. Forms and payments follow, because they catch the demand checkout was never designed to see — the wholesale message, the made-to-order request, the question that becomes an invoice the following week. Search work is third, and it is wasted any earlier, because it needs categories that already exist and already deserve to rank. The custom theme is last, unless how the brand looks is literally what the brand sells.
One exception is legitimate, and it is common. If the shop already trades and the marketing calendar is visibly stuck behind a developer queue, the theme moves up. Every month spent waiting is paid for in campaigns that never ran, and that bill has been arriving quietly for several quarters already. Bought in that situation, the 50-90 hour band is neither an upgrade nor vanity: it is the removal of a bottleneck you have been renting.
Recognise the moment by counting rather than by mood. Across a single quarter, tally the developer requests raised per campaign, the recurring app subscriptions bought to imitate a section the theme does not have, and the promotions that went out in a compromise layout. Three small numbers are an argument for staying where you are. Three numbers that have stopped being small mean the more expensive entry in the catalogue has already become the cheaper one for your business.
Five admin screens to open before you sign the handover
First, a collection page's text editor. Add a heading, a paragraph and a link inside it, then confirm the change appears on the live page where a reader and a crawler will both find it. Second, the title tag and meta description fields on both a collection and a product, edited and saved by you rather than by the person who built the shop, because that is the difference between owning the pages and renting them.
Third, the theme editor. Add a section to a page, move it above another, change its settings and remove it again. If a stock theme is what you are buying, note precisely which sections exist, because that list is your ceiling until you commission something else. If a custom theme is what you are buying, this screen is the deliverable you are paying 50-90 hours for, and it should feel like a tool rather than a demo.
Fourth, the form. Submit a test enquiry and watch it arrive in the CRM, in Telegram, in the spreadsheet and in the mailbox, then send a second one with an obviously robotic payload and confirm the spam protection catches it. Fifth, the filter rules and the sitemap: which parameter combinations are indexable, which are canonicalised, and by what procedure a combination that starts selling gets promoted into a real page later.
None of those five screens is about design, and that is the point. A handover is not a set of screenshots; it is the list of things your team can do alone on a Friday afternoon without asking anyone's permission. Write that list during the sales conversation, verify it in the admin panel before the final payment, and the choice between the two Shopify entries stops being a matter of taste.
Questions and answers
How much does it cost to build a Shopify store?
The catalogue lists a turnkey Shopify store at $800 – $2 500, or 40 000 – 120 000 ₽, with a stated band of 25-40 hours and a difficulty of Experienced. That covers catalogue, payment, shipping and analytics on a bought theme, with first orders taken on the day of handover. A custom theme is a separate entry at $2 000 – $6 000 and 50-90 hours.
Do I really need a custom Shopify theme, or is a purchased one enough?
Decide by what your team must change without a developer. A bought theme covers products, prices, menus and the sections it already ships with. The catalogue's custom theme entry, at $2 000 – $6 000 and 50-90 hours with a difficulty of Advanced, promises custom sections and a page builder inside the admin, so that a marketer assembles promo pages alone.
How long does e-commerce SEO for categories, filters and product cards take?
The catalogue states 3–7 days for a catalogue of up to 5000 SKU, priced at $600 – $3 000 or 40 000 – 150 000 ₽, with a published note that a deep audit of a large shop reaches 300 000 ₽ at market rates. The deliverable is texts, correct markup, controlled filter indexation and an end to categories and cards competing with each other.
Isn't the order form and payment already included in a store build?
It is a separate catalogue entry, Application forms and payment acceptance on the website, under Integrations at $250 – $800 or 10 000 – 35 000 ₽ and 8-20 hours. It covers a form that writes to CRM, Telegram and a spreadsheet, email that actually arrives, a pay button that really takes money, and protection against spam bots.