The most common misconception about function-driven content is that it requires building a lot of new data infrastructure. Teams imagine they will need to commission new fields, integrate new sources, and run a long data project before they can generate content. Sometimes that is true. More often, the data they need is already in the database, already accurate, already updating, and simply never displayed on the pages where it would help.
Before you plan any data work, take inventory. This article is a walkthrough of the variables most e-commerce databases already contain, organized by where they live and what content they can drive. As you read, check each one against your own system. You will likely find that most of them already exist.
Variables that live at the category level
Category-level variables are computed across all the products in a category. Your database can produce all of these on demand because they are simple aggregations of data you already store.
- Product count. How many products are in this category. A simple count. Drives "Choose from 412 tactical bipods."
- Brand count. How many distinct brands are represented. Drives "from 38 brands."
- Lowest price. The minimum active price in the category. Drives "starting at $24.50" and "as low as."
- Highest price. The maximum active price. Together with the lowest, drives a price-range statement.
- Deepest discount. The largest percentage or dollar markdown currently active. Drives "save up to 40%."
- In-stock count. How many products are currently available. Drives availability and urgency signals.
- New-arrival flag. Whether any product launched recently. Drives freshness signals.
- Top-rated indicator. The highest or average review rating. Drives trust signals.
Every one of these is a computation over data you already store: prices, inventory, launch dates, review scores. You are not creating new data. You are aggregating existing data and exposing the result.
Variables that live at the product level
Product-level variables are attributes of individual items. These are the richest source of long-tail content, because they let single-variant product pages capture highly specific queries.
- SKU, model number, product ID. Specific identifiers that searchers use directly. Drives long-tail title tags that capture exact-match queries.
- Specifications. Dimensions, weight, capacity, material, color, compatibility. The single largest untapped content source on most sites.
- Price and sale price. The current price and any active promotion. Drives value and savings signals.
- Stock status. In stock, out of stock, backordered, limited. Drives availability and urgency.
- Launch date. When the product became available. Drives "new" badges and freshness signals.
- Brand and manufacturer. Drives brand-page eligibility and brand-plus-category content.
- Review count and rating. Drives social proof on the page.
Specifications deserve special attention. On most sites, specifications are stored in a structured table that powers the filter sidebar, and then never used for anything else. That same specification data can drive title tags, product names, and descriptive sentences. A bipod's leg material, height range, and mount type are sitting in your specifications table right now, powering a filter, and never appearing in a single SEO-relevant text element. That is wasted content.
The specification goldmine
Specification data is usually the largest underused content asset on an e-commerce site. It is structured, accurate, maintained, and unique per product. It already powers your filtering. With function-driven content, the same data can drive product names, title tags, and descriptive sentences that capture an enormous range of long-tail specification queries your competitors are not even trying for.
Variables that live at the relationship level
Some of the most valuable variables describe relationships between products rather than attributes of a single product. These power internal linking and cross-sell content.
- Related products. Items in the same category or sharing key specifications. Drives internal-link blocks.
- Frequently bought together. Products commonly purchased in the same order. Drives cross-sell and complementary-product content.
- Compatible products. Items that work with this one, like a mount compatible with a specific optic. Drives compatibility content that captures high-intent queries.
- Price-tier alternatives. Cheaper and more expensive options in the same category. Drives "good, better, best" content.
These relationship variables are sometimes already computed for merchandising purposes, the "you might also like" widgets, and never used for SEO. The same relationships that power a recommendation carousel can power an internal-linking strategy that distributes link equity intelligently across the catalog.
Taking inventory before building
The practical exercise this article is pushing you toward is simple. Sit down with whoever knows your product database, an engineer or a data analyst, and walk the list of variables above. For each one, determine three things: does this data exist, is it accurate, and is it currently displayed anywhere on the public site.
You will usually find a pattern. Most of the data exists. Most of it is accurate, or accurate enough. Almost none of it is displayed on the SEO-critical text elements of your pages. The data is doing internal work, powering filters and recommendations and inventory management, and contributing nothing to your rankings.
That gap is the opportunity. The data project most teams fear is often unnecessary. The data is already there. The work is in connecting it to the functions that will surface it on your pages.
The trap door
The inventory exercise sometimes reveals that data which appears to exist is not reliable enough to display. A price field that is accurate for in-stock items but stale for discontinued ones. A specification table that is complete for new products and sparse for older ones. Displaying unreliable data at scale broadcasts your data quality problems to every visitor and to Google. The inventory must assess not just whether data exists but whether it is reliable enough to surface publicly.
The shift in how you think about data
The mental shift this article is after is to stop thinking of your product database as inventory infrastructure and start thinking of it as a content source. Every field in that database is a potential variable. Every variable is potential content. The database your business has maintained for years, at considerable expense, is also the richest content asset you own, and most of it has never appeared on a page.
Function-driven content is, at its core, the practice of connecting that database to your pages. The variables in this article are the raw material. The functions and conditional statements from the previous Insights are the machinery. The output is pages that are specific, unique, and updatable, built from data you already have.
The final Insight in this section covers how natural language processing turns these variables into sentences that read naturally. After that, the Tactical section gets specific about which variables drive which page elements.
From the book
Sizzle: An E-Commerce Revolution contains a detailed catalog of updatable, specific, and unique variables, with guidance on which ones drive which content elements and how to assess data reliability before building.