Select Page

Data parsing: the hard work that makes great software

by | Jan 26, 2021

Like so much of life, the old metaphor of the iceberg perfectly illustrates the software development process. Most of what you love about your favorite software—be it Google, Dropbox or Among Us—is hidden deep below the small slice of UI users actually see and experience. Down below the surface is also where all the hard work of building great software happens.

Before we began building Pagemunk, one big unknown for our development team was what kind of data we would encounter across PDPs. In other words, how deep did the iceberg go below the surface of the water? Every Shopify theme formats content differently. On top of that, most Shopify stores are heavily customized. All that variation and customization creates a better experience for the customer, but it adds a lot of complexity for developers behind the scenes.

Text

Description automatically generated
This is what a product description looks like to a consumer
Text

Description automatically generated
This is what a product description looks like to a developer 

What is data parsing and why is it important?

When it comes to PDP content, data parsing is essentially separating the consumer-facing wheat from the coding chaff. We do this so PDP content can be displayed in a clean, easy-to-edit format within Pagemunk, unencumbered by confusing HTML and formatting.

Not only do we need to make sense of the data, we need to teach the app to make sense of it without any human intervention. And get it right every time on every site.

Beyond that, we also need the app to recognize the difference between unique content types—say a headline and a description—and then separate each type into its own cell. We do this so users can change one element of their product information, like a single product detail, across multiple products with a simple find and replace.

Within Shopify, you’re able to edit one product page at a time from top-to-bottom (what you might call vertical editing). But you can’t change a single element across multiple pages at once (what we call horizontal editing). Bulk editors allow you to change a single field, like a tag or price. But they don’t allow you to change a single element of your product description. With Pagemunk, we want to make it as easy to work across an entire collection or product line as it is to work on a single product.

This kind of deep data parsing work is the equivalent of building the foundation of a house. It’s not glamorous, it’s hard, but it’s also necessary to build a valuable app.

Our first encounter with live PDP data 

Over the last month we’ve worked with several of our beta customers to install the earliest version of Pagemunk. While the UI was barely started, it provided us with a glimpse into real-world data straight from live Shopify PDPs. We found lots of complexity but, luckily for us, no insurmountable hurdles.

PDPs come in many different shapes and sizes, from very deep storytelling on Tracksmith.com PDPs to the very data-driven PDPs of article.com. Though the layout and depth of information is different from store to store, when it comes to content types, most every ecommerce site follows a stable pattern. These shouldn’t surprise anyone running an online store. What surprised us was how stable these types were across a wide range of sites selling a wide range of products:

  1. Product overview 
  2. Product details 
  3. Social proof

Product overview: product photography, product description, product name, buy box (price, shipping, colorway, buy button).

Graphical user interface, application

Description automatically generated
(Herschel.com) 

Product Details: Fit, benefits, materials, sizing, tech specs, explainer videos

Graphical user interface

Description automatically generated
(Pelacase.com) 

Social proof: user-generated photos and videos, reviews and comments

A collage of a person

Description automatically generated with medium confidence
(doelashes.com) 

Ecommerce changes quickly. Building trust never does.

The PDP experience has stabilized into these three content types because the PDP is all about creating trust. And whether you’re shopping on a PDP or having a conversation with a stranger, these are the three types of content that build that trust. They answer the three questions we always ask when we encounter anything new:

  • What is this and is it for me? (Product overview)
  • Does it really do what it promises? (Product details)
  • Do people I trust use it and like it? (Social proof)

By the time consumers reach the PDP, they’ve picked the product off the digital shelf. The role of product page content is to close the deal by removing any lingering concerns that are keeping them from clicking the buy button.

No matter how much a brand pushes the design of their PDP, the real power to entice customers and close the deal comes down to the content. Which is why we’re doing the hard work of parsing data deep below the surface of the PDP right now. The product photography, story, details and reviews are what build trust, and trust is what creates valuable relationships between brands and their customers.