Automating Product Builds for a Shopify Store

Client: withheld – Psychedelic Apparel Brand for Festival-Goers

Industry: E-commerce (Fashion & Print-on-Demand)

Platform: Make.com, Printify API, Google Sheets, DeepAI, Dropbox

Time to Build: 40+ hours

Time Saved: 10+ hours per week

Consulting Partner: Lumio Consulting

Project Overview / Problem

A client came to us asking for help optimizing their product builds and publishing process for a Shopify-based print-on-demand (POD) store. They requested to remain anonymous due to the potential negative perception of a POD-based store for their brand. Still, they agreed to let me publish the details of their automations, subject to their review and approval.

The initial goal was to simplify the product build phase, which involved manually uploading designs to their Printify platform, associating those images to specific product ‘blueprints’ (sections of an article of clothing), and then reviewing the image placement to ensure the final product had no errors, or gaps, in the layout.

Utilizing Make.com and the Printify API, we began with a single, simple product and expanded the workflow to automate the creation of multiple products (using a single design) within a single workflow.

The final automation (one of many iterations) has a Make.com automation that takes a single image from Dropbox, generates a name for the design, generates a product description (based on each product’s details), uploads the design to Printify, associates the image to each product’s blueprint, and preps the product for review, approval, and publishing by the client.

Although we could have automated the entire process, from image selection to immediate publishing to their Shopify store, we (and they) were adamant about having a “human-in-the-loop” to ensure there were no mistakes and that the final published product met their standards for display with their store.

Goals of the Automation

The initial goal of the automation was to save time associating design images to Shopify product blueprints; it eventually expanded into image upscaling (for clarity), name generation, and product description generation, in addition to going from a single product build to multiple product builds in a single automation.

Attached is one of the more complex automations, which takes a single product design and builds out 14 products, ready for review and publishing.

Here is the complete automation workflow:

Solution Overview / Step-by-Step Breakdown

Step 1. Pull a list of available design files from a local Dropbox directory, select a file (at random), and create a shareable link further processing.

Steps 2 & 3. Upscale the image (for clarity), using a 3rd party tool ‘DeepAI.org’ with a custom HTTP request to their API. Pull from a list of pre-vetted product names and select one at random.

Step 4. Set the product name, then pull product details from Google Sheets, and, using ChatGPT (gpt4), generate a product description.

Step 5. Upload the image to Printify, and fan out to create multiple products from the same image file

Step 6. Once each product has been created (not published) in Printify, consolidate the automation path into a single final workstream that moves the design file from a “pending upload” folder to a “completed” folder.

Challenges

We ran into several challenges while building this automation:

Image formatting and placement - Printify’s documentation offered limited guidance on exact design placement. It required multiple iterations and manual testing to align the design files to each garment’s blueprint.
Image upscaling - The initial design files for some (but not all) products were of low resolution, which required us to integrate a third-party app for image upscaling.
Collapsing multiple automation paths into one - We created a custom JSON bundle to manage metadata across product types.

Results/Outcome / Time Saved

Before implementing this automation, our client was spending 10+ hours per week manually uploading designs, associating those designs with specific products, generating product descriptions (ChatGPT was being used, but it was a manual process), and publishing those products to their store.

After implementation, our client had daily product design builds that required a simple review process and one-click publishing, allowing them to shift their focus from product publishing to product design.

We (and they) estimate their time-saving at hundreds of hours, as they can scale their publishing up or down as seasonality requires. Their new bottleneck (per later discussion) was the design process itself (which we later helped with as well)!

Since starting our work with [company], we’ve built over 50 automations for their Shopify store, including customer management tracking, product review analysis, inventory tracking, and image design automation.

Client Feedback

““Before this automation, building products was the biggest time sink in our workflow. It took hours just to prep one design across all our products - it was frustrating and completely unsustainable. Now, we drop a new design into Dropbox and the system does the rest. All we have to do is review and publish. It’s saved us hundreds of hours over the past six months and completely changed how we think about scaling.””

— Store Owner

Future Improvements

Although we’re looking at one of our original automations for this client, we later came back and made several improvements to the process:

Custom product descriptions for each product line, pulling in target audience and integrating keywords relevant to their industry.
In-workflow image analysis of the design using ChatGPT Vision (ChatGPT’s image analyzer tool), which allowed us to generate unique and accurate product descriptions for each product (versus pulling from a pool of names and generic descriptions).
Unique product names for each product, utilizing the same ChatGPT Vision module.
Theme-based product generation for Collection-style products.