--- title: "Shopify Integration: Order to Invoice | JustAutomate" description: "We wire Shopify into invoicing, e-invoicing, CRM and a data warehouse. Variants, currencies and margin after returns, with nothing retyped by hand." url: "https://justautomate.ai/en/integrations/shopify/" lang: en --- E-commerce and finance # Shopify integration from the order to the filed invoice Shopify sells, takes the payment and keeps the customer profile. It will not ask for a VAT number where your accountant needs one, file a structured invoice, or tell you what one size and colour earn after returns. We wire the store into invoicing, CRM and a data warehouse, so invoices, stock and credit notes are created without retyping, and profit per variant is visible every day. Book a free consultation [See how the workshop works](https://justautomate.ai/en/services/process-workshop/) - 7 data streamsthe ones that usually leave a Shopify store, from order lines to discount codes, and the order we take them in gets set in the workshop - 2 to 6 weeksfrom the workshop to a working integration, depending on the number of markets, currencies and systems standing on the other side of the store - 5 to 10 secondshow long the AI agent needs to say what one product variant earns after discounts and after returns The problem ## Why does Shopify sell without effort while the paperwork stays manual? Shopify is a global platform. That is its strength in selling and its weak spot wherever local tax rules begin. The store takes the order, collects the payment, sends the confirmation and saves the customer profile. It will not force a VAT number into the field your accountant needs, it does not know your document numbering, and it does not know that an accepted return calls for a credit note. The second pile of manual work grows on variants. One model in five sizes and three colours is fifteen separate lines, each with its own SKU and its own stock figure. Shopify counts stock per variant and per location, while the ERP or the spreadsheet often counts it per model. Once a week somebody reconciles the two, and that person, not a system, becomes the source of truth about stock. The third thing is a deadline nobody will move for you. Mandatory structured e-invoicing is being introduced across EU countries, each on its own timetable, and the filing format is set by the national scheme rather than by your store. Shopify files nothing. Between the store and the tax authority there has to be invoicing software, and somebody has to decide what links an order to an invoice. - The VAT number typed onto the order by hand, after an email from a customer who wants an invoice - Line items retyped from Shopify into the invoicing system, one at a time - Stock on one size corrected in the store admin after a phone call from the warehouse - Revenue in euro and pounds converted at the rate from the day of the export, not the day of the sale Why now ## An app from the Shopify App Store, or your own integration? Shopify has an app store and for plenty of jobs that is the cheapest route. One app for invoices, a second for stock, a third for post-purchase email. If each one does its own job and they never have to talk to each other, there is no reason to build an integration, and that is what we will tell you in the consultation. The limit shows up in three places. Apps cannot see one another, so a rule that depends on data from two of them has nowhere to live. Each one carries its own subscription, and at six apps the fixed cost often beats a one-off rollout. The third one matters most: the data stays inside the app, so when you want margin per variant after returns, there is nothing to compute it from. That is when the logic moves into n8n, Make or a custom integration over the API, and Shopify goes back to what it is best at: the storefront, the cart and the customer profile. The technical layer, meaning authentication, events and rate limits, we describe separately on the page about [API integration](https://justautomate.ai/en/services/api-integration/), so this page can stay about the process. Then there is the AI agent. A rule will move data between systems, but it will not tell you whether a variant that mostly sells with a discount code still earns anything. BOND, our company AI agent, reads sales, discounts, cost of goods and returns at once and answers in a plain sentence in 5 to 10 seconds, with nobody logging into four panels. What changes ## What does the store team get, and what does accounting get? - ### The order enters the document flow on its own The store reports a paid order within a second, and the workflow turns it into an invoice in Xero, QuickBooks, sevDesk or Fakturownia. Buyer details and the VAT number get checked on the way, so the document does not come back for correction. The invoice is built from order data, with no line items retyped - ### The invoice is filed and the number comes back to the store Your invoicing software handles the filing to the e-invoicing scheme in your market. We look after what happens before and after: complete buyer data, document order and writing the invoice number back onto the order in Shopify. An accepted return starts a credit note. The invoice number visible on the order, not only in accounting - ### Variants and stock counted in one place We agree one source of truth for stock: Shopify, your ERP or the warehouse. The number travels at variant and location level, so size M does not sell twice when the last piece is sitting in the physical store. Stock per variant and per location, distributed instead of retyped - ### The shopper is the same person in your CRM The Shopify profile, marketing consents, order history and cart value land in HubSpot, Pipedrive or Salesforce as one record. Sales can see what that company bought in the store instead of asking customer service. One customer record instead of two databases, store and sales - ### Several markets and currencies in one set of reports An order in euro or in pounds is saved with the rate from the day of the sale and with the delivery country, so the report in your reporting currency agrees with the books. What counts as one market for the store and what counts as a separate one for tax gets decided once, in the workshop. The rate from the day of the sale, not the day of the export - ### Margin per variant after discounts and returns Sales, discount codes, shipping cost and returns from Shopify, plus cost of goods from accounting, land in BigQuery or in a sheet, lined up per variant and per market. The AI agent answers from that data in a plain sentence. One number per size and colour, refreshed every day Proof ## Do you work on store data like this yourselves? Yes, in a store one of our founders owns. That store does not run on Shopify, so read this as proof of the mechanism, meaning sales and returns data in one place plus an AI agent, and not as proof of the Shopify connector itself. Below is six months of measurement, together with what the number does NOT prove. Built for ourselves. These are not client deployments but implementations inside the founders' own companies. We say so plainly, because proof you cannot verify is not proof. - ![Surf Inc.](https://justautomate.ai/klienci/surf-inc.svg)![SurfBot](https://justautomate.ai/agenci/surfbot-80.webp)SurfBottheir agent Surf Inc., a premium clothing brand with an online shop and two physical stores 16 people from the company work with the agent 21 systems wired in permanently 89 processes run without a human 323 thousand lines of the agent’s code A catalogue counted in hundreds of variants, selling across several markets and in several currencies. Data moves from the order system to the ERP, the store, the data warehouse and the ad platforms. The AI agent launched in March 2026 and was fully running by May. Sixteen people use it, warehouse and store staff included. 8 of 21 systems it works in - PrestaShop - BaseLinker - Comarch Optima - BigQuery - Google Ads - Meta Ads - GA4 - Zalando Before From February to August 2025 every unit of ad spend came back as 3.53 in online revenue, and advertising absorbed 28.3 percent of that revenue. Sales and returns per product were lined up by a person, in a spreadsheet. After In the same months of 2026 the ratio is 4.19 and advertising is down to 23.9 percent of revenue. The ad budget is 6.9 percent lower with sales 10.3 percent higher. Return on ad spend from 3.53 to 4.19, 18.5 percent higher, on a budget 6.9 percent lower And in your store ## What if our Shopify is set up differently? Four sentences come up more often than any others on this platform. Here is what we answer to each of them. - We have an invoicing app from the Shopify App Store, that is enough. If it issues a compliant invoice with the buyer VAT number, files it to the e-invoicing scheme and copes with a credit note after a return, then it is enough and we will say so. We check that in the consultation, on your own documents. An integration comes in when the document has to depend on data from outside the store: a credit limit in the CRM, stock in the ERP or separate numbering for one market. - We sell from one store into three countries. Then we start by separating two things that are easy to confuse: what counts as one market for the store and what counts as a separate one for tax. We save the currency, the rate from the day of the sale and the delivery country, so the report in your reporting currency agrees with the books. The VAT treatment of cross-border sales is set by your accountant, we deliver the data in a shape that treatment can be applied to. - Our catalogue runs to thousands of variants, that cannot be connected. The number of variants is not the problem as long as one key exists. The problem is an SKU that means one thing in the store and something else in the ERP. In the workshop we check what those systems actually match on, and if there is no key, the first job is to agree one rather than to build an integration on matching by product name. - We have no IT team and nobody here has touched the technical settings. Nobody has to. What stays on your side is creating access in the store admin and deciding what data may leave. We do the rest, and after the rollout we hand over the documentation and train one person in customer service to read the error queue and see what did not go through. Store and sales Shopify, WooCommerce, PrestaShop, BaseLinker, Amazon Accounting and invoices Xero, QuickBooks, sevDesk, Fakturownia, e-invoicing filing CRM HubSpot, Pipedrive, Salesforce, Zoho Data and reporting BigQuery, Google Sheets, GA4, SQL databases Team work Slack, Google Workspace, Microsoft 365, Notion Automation layer n8n, Make, Zapier, custom API integrations ## Data held in Shopify and what is missing on the accounting side Seven streams, the ones we usually start from. The order we take them in gets set in the workshop, because it decides what can be switched over straight away and what is better run alongside the current process first. | Data in Shopify | What is missing on the accounting side | Where we connect it | | ----------------------------- | -------------------------------------------------------------------- | -------------------------------------- | | Order and line items | buyer VAT number, document numbering, the rate for that market | invoicing software, e-invoicing filing | | Variants and SKUs | a shared key with the ERP and the warehouse | ERP, warehouse, BaseLinker | | Stock per location | one source of truth when the physical store sells off the same shelf | ERP, warehouse, sales channels | | Customer profile and consents | a link to the company record, not only to an email address | HubSpot, Pipedrive, Salesforce | | Currency and amount | the rate from the day of the sale and the delivery country | accounting, data warehouse | | Return and partial return | a credit note and the stock put back | invoicing software, warehouse, CRM | | Discount codes | cost of goods, without it margin after discount is guesswork | BigQuery, spreadsheet, AI agent | A first rollout usually covers two or three streams. The rest come in down the same track once the first ones pass their measurement at 30 days. How we do it ## Five steps from a Shopify store to a connected flow Every step ends with something you can judge before you decide on the next one. We do not install tools blind. 1. ### Workshop One week. We walk the path of one of your orders: from the cart, through payment and the invoice, to the return and the credit note. We note where somebody adds a VAT number, where somebody fixes a variant stock figure and where somebody converts a currency. We also look at which apps in the store already do this and whether they are worth keeping. 2. ### Integration design A list of data streams, the direction of each one, the shared key between Shopify and the ERP, and one source of truth for stock. A decision on what stays an app from the store and what moves to n8n, Make or an API integration. Always the simplest route to the goal, not the most elaborate one. 3. ### Pilot on real data The flow runs next to your current process with customer messaging switched off. We replay situations from last season: an order from a company with no VAT number, a return of two lines out of five, a purchase in euro shipped abroad, and the last piece sold online and in the store in the same minute. 4. ### Production and training Once it checks out we switch it to production with error monitoring. An off switch stays on your side, along with an error queue that shows what did not go through and why. Somebody in customer service learns to read it and close it, without raising a ticket with us. 5. ### Measurement after 30 and 90 days We check the metrics written down in the workshop: how many orders reached an invoice without a person, how many documents came back for correction, how many times a variant stock figure drifted. Only then do we add the next stream, down the same track. Before and after ## How the day changes for the person running the store ### Before - The morning starts by adding VAT numbers to overnight orders and retyping line items into the invoicing system - The credit note after a return waits until month end, because it has to be issued by hand - Stock on size M gets corrected in the admin after a phone call from the warehouse - Asked what one colour earns, the answer is I will run an export and get back to you ### After - Overnight orders have an invoice, a filing number and a CRM record before anyone opens a laptop - The credit note starts on the day the return is accepted, together with the stock going back - Variant stock is counted in one place and travels to the store on its own - Asked what one colour earns, the AI agent answers in 5 to 10 seconds The first stream, usually the invoice from the order, runs in production around 14 days after the workshop. The effect on handling time and on margin gets judged after 30 and 90 days, on metrics written down before the start. For the person signing this off ## What the store owner gets out of it The e-invoicing obligation stops hanging on whether somebody remembered an extra step after the order. The invoice is built from order data, and gaps in the buyer details show up the same day rather than at month end. That is one less risk in the conversation with your accountant. The second thing is buying decisions. When sales, discounts, cost of goods and returns sit per variant in one place, you know which size and which colour to order for the next season. You start with one stream, so the first decision is small and reversible. - The workshop ends with a document you can execute without us - The pilot runs in parallel, so your current process works until the tests are done - The off switch and the error queue stay on your side - App, tool and model costs go straight to the vendors, with no margin from us What it costs ## The workshop has a published price, the integration is quoted after it The process workshop costs **EUR 3,500 net per week** and runs from 1 to 7 weeks. A Shopify store usually needs one week, two when it sells across several markets and currencies or when a separate ERP stands on the other side. We quote the integration after the workshop, because only then is it clear how many streams leave the store and what state the variant data is in. Your Shopify subscription, app store fees, the cost of scenarios in [Make](https://www.make.com/en/help) or [n8n](https://docs.n8n.io/) and model usage, at EUR 0.005 to 0.07 per query at provider list prices, you pay the vendors directly. We add no margin and no invoices pass through us. - ### Process workshop EUR 3,500 net per week, usually one or two for a Shopify store A self-contained piece of work. It ends with a map of the order flow, a list of streams to connect and a decision on what stays an app from the store. - A map of the order path from cart to credit note after a return - A review of the apps you already run and a decision on what stays - The shared key agreed between Shopify, the ERP and the warehouse - The invoice route to filing, with an owner for every step - The metrics to measure at 30 and 90 days Start with the workshop - Most popular ### Workshop and Shopify integration Quoted after the workshop scope follows from the number of markets, currencies and systems The option most stores choose. A working flow 2 to 6 weeks after the workshop, launched next to your current process first. - Everything in the workshop - Invoice and credit note from the order, filed through your own software - Stock per variant and per location from one source of truth - Customer profile and consents tied to the company record in the CRM - Currencies, rates and delivery country saved for reporting - Pilot on real data, error monitoring, training for your team Talk about a rollout - ### Shopify and an AI agent Quoted after the workshop for stores that ask about variant margin every day Everything in the integration plus BOND, which answers from that data in a plain sentence. The agent takes around two months to roll out. - Everything in the Shopify integration - A warehouse with sales, discounts, cost of goods and returns per variant - The AI agent wired into the store, accounting and the data warehouse - Answers in 5 to 10 seconds, in English and in Polish - Access for customer service, the warehouse and the owner Ask about the agent The workshop ends with a document that stands up without us: the flow map, the list of streams and the rollout order. If it turns out the apps you already have in the store are enough, we say so plainly and show you how to set them up. The matter ends at the workshop, not at an integration you do not need. ## Frequently asked questions What does a Shopify integration actually connect? It ties the store to what stands on both sides of it: accounting and e-invoicing, the CRM, the warehouse or ERP, and the data warehouse. New orders, stock changes and returns leave Shopify on their own, and results come back onto the order, the invoice number for example. Shopify stays the storefront and the customer profile, while decisions that depend on data from other systems are made in the automation layer. How long does a Shopify integration take? One week of workshop and 2 to 6 weeks of rollout, depending on the number of markets and how many systems sit on the other side. The first stream, usually the invoice from the order, runs in production around 14 days after the workshop. We add the next ones after the measurement, not all at once. Will Shopify issue a compliant invoice and file it for me? Not on its own. Shopify issues a sales document in its own format and knows nothing about structured invoices or your document numbering. The filing is done by invoicing software: Xero, QuickBooks, sevDesk or Fakturownia, depending on the market. Our job is to deliver a complete set of order data to it and write the number back onto the order in the store. When is an app from the Shopify App Store enough, and when do you need your own integration? An app is enough when there is one job to do and all the data sits inside the store. A custom integration makes sense in three situations: the decision depends on data from outside Shopify, stock in the ERP or a credit limit in the CRM for example; the subscriptions for several apps add up to more than a one-off rollout; or you need that data in a warehouse and the app will not let it out. How does stock sync work with product variants? Stock travels at variant and location level, so size M is a separate counter from size L, and the physical store is a separate place from the warehouse. In the workshop we agree one source of truth, the store or the ERP, and the shared key the systems match items on. Without that key an integration guesses by name and drifts apart at the first catalogue change. We sell in several currencies, how does that show up in reports? The order is saved in the currency of the sale, with the rate from the day of the sale and the delivery country. In the data warehouse the amount in your reporting currency sits next to it, so the report agrees with the books rather than with the rate on the day of the export. The VAT treatment of cross-border sales is set by your accountant, we deliver the data in a shape that treatment can be applied to. Where does the data for margin per variant come from? Sales, discount codes, shipping cost and returns from Shopify, cost of goods from accounting or from the ERP. It all lands in BigQuery or in a sheet, lined up per variant and per market. Margin is then counted after discounts and after returns rather than before them, and it is visible every day. We write more about pulling store data together on the page about e-commerce automation. Do we need a developer and access to the Shopify admin? No developer. What stays on your side is creating access in the store admin and deciding what data may leave. The technical layer, meaning authentication, events and rate limits, we describe separately on the page about API integration. After the rollout we hand over the documentation and train someone in customer service to read the error queue themselves. ## Let us find where your Shopify ends and the manual work starts The free consultation takes 30 minutes. You tell us how many markets you sell into, which apps you run in the store and who adds the VAT number to orders today. We show you what can be connected straight away and what needs the variant data tidied up first. You do not have to give us admin access at this point. We run at most two rollouts at a time, so that someone is actually sitting on each of them. [Book a free consultation](https://justautomate.ai/en/contact/) **PS.** If somebody in your company spent today retyping line items from an order into an invoice, that is exactly this problem. Once it is connected the invoice is built from the order and goes out for filing, and the profit on a single variant is visible every day. You write to Paweł, a founder, not to a shared inbox. ## Related - [E-commerce automation](https://justautomate.ai/en/services/ecommerce-automation/) - [E-commerce: margin after returns](https://justautomate.ai/en/industries/e-commerce/) - [Automatic inventory update](https://justautomate.ai/en/automations/automatic-inventory-update/) - [BaseLinker integration](https://justautomate.ai/en/integrations/baselinker/) - [API integration](https://justautomate.ai/en/services/api-integration/) - [BOND: the business AI agent](https://justautomate.ai/en/services/bond-business-ai-agent/) --- **JustAutomate** · contact: contact@justautomate.ai · form: https://justautomate.ai/en/contact/