BLOG

What Is Software Normalization? Inside the Block 64 Product Library

October 13, 2026
•
0
min read

Software normalization is the process of turning raw software discovery data into one consistent record per product: one publisher, one product title, one version.

Without it, a software inventory counts the same application several times under different names. Every report built on that inventory is off.

It's one of the oldest problems in IT asset management. It's also the first thing the Block 64 Product Library fixes.

The library now holds 1.5 million unique software assets from more than 74,000 publishers, cataloged from inventory across nearly 10,000 customers and 3.7 million devices.

Here's what software normalization is, why your inventory needs it, and what Block 64 adds once the names are clean.

What is software normalization?

Software normalization converges raw software records into standardized publisher, product title and versions, so the same product is always counted as the same product.

Discovery tools collect whatever an installed application reports about itself. That might be the publisher's full legal name, a product title with a build number stuck on the end, or nothing in the publisher field at all.

Those raw strings are what a software inventory is built from. Normalization matches each one to a known product in a reference catalog, so the data becomes something you can count.

In practice, plenty of installed software doesn't carry clean tags. Discovery tools still see raw strings, and normalization is what turns those strings into data you can count.

Why your software inventory needs normalization

Software publishers don't standardize how they package and label their products. The same product from the same publisher can show up under two different names in the same environment.

We ran into this constantly in our own work. Two names for one product means two rows in a report, split install counts and a license position that doesn't match reality.

That breaks the simple questions IT gets asked every quarter:

  • How many installs of this product do we actually have?
  • Which publishers do we depend on most?
  • Are we covered going into the next true-up or renewal?

If the inventory can't group records correctly, none of those answers hold up.

The four kinds of messy software data

Most normalization problems fall into four buckets.

1. Publisher variants. One publisher, many spellings: full legal names, abbreviations, capitalization and punctuation differences.

2. Missing publishers. The publisher field is blank, so the record never shows up when you filter by vendor.

3. Versions and editions buried in titles. The product title carries the year, edition and version, so every release looks like a different product.

4. Title noise. Build numbers, language codes, architecture tags (x64, x86) and other strings that split one product into many.

Software normalization examples
ProblemRaw record (publisher · title)Normalized (publisher · title)
Publisher variantsMICROSOFT CORP. · Microsoft EdgeMicrosoft · Edge
Missing publisher(blank) · Google ChromeGoogle · Chrome
Version in titleMicrosoft Corporation · Microsoft Office Professional Plus 2021 - en-usMicrosoft · Office Professional Plus
Title noiseIgor Pavlov · 7-Zip 23.01 (x64)Igor Pavlov · 7-Zip
Illustrative examples of the variation the Block 64 Product Library normalizes.

How the Block 64 Product Library normalizes software

The Block 64 Product Library is our catalog of every software, SaaS, and product we discover, with hardware coming soon. Normalization is its first job: standardize publisher names and software titles so every record lands on the right product.

We do it with purpose-built AI agents working against a catalog built from real customer environments and reviewed by our team of (human) experts. A new record either matches a known product or becomes a new entry in the library.

Today, 99.9% of the software assets in the library are normalized:

  • Installed software: 99.96%
  • Operating systems: 99.98%
  • SaaS products: 97.75%

Normalization is step one. Enrichment is where the answers are.

Knowing a product is installed isn't the same as knowing what to do about it. Our customers kept asking the next question: Is it supported? How is it licensed? What does it cost?

So once a record is normalized, the library enriches it. Block 64 collects data from publicly available sources, such as publisher support documentation, lifecycle pages and press releases. Where reliable data doesn't exist, we leave the field empty instead of guessing and then send it to a human for review.

Every normalized product can carry six enrichment fields:

Block 64 Product Library coverage
SegmentAssetsNormalizedClassifiedLicense metricsVendor EOL dates
Installed software1.48M99.96%98.2%55%28%
Operating systems~8,20099.98%98.7%62%89%
SaaS products~8,00097.75%82.9%26%Not reported
Source: Block 64 Product Library data, October 2026.

What clean, enriched software data lets you do

A normalized, enriched software inventory turns a list of installs into a list of decisions. Here's where customers use it most.

Find unsupported software and operating systems

This is the question we hear first. Under Microsoft's Fixed Lifecycle Policy, once support ends Microsoft "no longer provides new security updates, DST updates, or other non-security updates." CISA includes unsupported and end-of-life software on its list of bad practices, warning that it "significantly elevates risk" in critical infrastructure.

With supportability data attached to every normalized record, you can see which installs are past their date and which are close. That's the starting point for deadlines like Office LTSC 2021 end of support on October 13, 2026, SQL Server 2016, which left support on July 14, 2026, and Windows Server 2016 in January 2027.

Prepare for license reviews and renewals

Clean install counts plus the licensing metric for each product give you a defensible starting position before a true-up or renewal conversation. It's the groundwork for software license optimization and for unified licensing visibility across SaaS and on-premises software.

Build an application inventory you can rationalize

Categories group products by what they do. When three tools in the same category do the same job, that's an application rationalization conversation with real data behind it.

Separate paid software from free software

The commercial flag and MSRP show where the money actually sits, so the review starts with the products that carry a cost.

The Block 64 Product Library by the numbers

  • 1.5 million unique software assets cataloged
  • 1 million+ distinct products
  • 74,000+ publishers
  • 3.7 million devices inventoried across nearly 10,000 customers
  • 99.9% of assets normalized
  • 98.8% of assets enriched

Lifecycle coverage is strongest where risk concentrates: 89% of operating systems carry vendor end-of-life dates. Many installed applications have no published lifecycle at all, and where a publisher hasn't published dates, the field stays empty rather than guessed.

Software normalization FAQ

What's the difference between software discovery and software normalization?

Software discovery finds what's installed or in use across your environment. Normalization cleans up what discovery finds, so each product is recorded under one publisher, one title and one version.

What is a normalized software inventory?

It's a software inventory where every record has been matched to a standard publisher and product name. Counts are accurate, products group correctly and reports can be trusted.

Why does the same software appear under different names?

Publishers don't package and label products consistently. Version numbers, editions, language codes and company name changes all create variations of the same product.

Does Block 64 normalize SaaS and operating systems too?

Yes. The Block 64 Product Library covers installed software, operating systems and SaaS products. Normalization rates are 99.96%, 99.98% and 97.75% respectively.

Where does Block 64 get end-of-life and pricing data?

From publicly available sources such as publisher support documentation, lifecycle pages and press releases. If reliable data isn't available, the field is left empty.

Does the library store customer data?

No. The library stores product-level facts like names, categories, licensing metrics and lifecycle dates. Customer metadata is deleted.

‍

See your software inventory, normalized

A Block 64 Rapid Assessment runs your environment through the Product Library and hands back a defensible plan to cut IT waste and risk in 30 days. A 12-month Block 64 platform subscription is included, so the clean data stays with you.

Explore Rapid Assessments or book a demo.

‍

‍

Get your free scan and see where your IT is hiding

14 days free. 15 min set-up. No credit card required.