A design team discusses a software interface redesign, comparing two screens alongside analytics and sketch prototypes.
Office's Ribbon had a rough landing, and it's easy to remember it as a pure telemetry success story. The primary sources tell a more nuanced one. XDA Developers' retrospective, built on a 2008 talk by former Microsoft designer Jensen Harris, describes how the Ribbon came to be. The talk, plus Harris's own write-ups and a 2008 Bill Gates speech, shows that data was only one ingredient. Observation, design principles, prototyping and long-running usability testing mattered just as much.

A design team discusses a software interface redesign, comparing two screens alongside analytics and sketch prototypes. Why Office 2003 needed fixing​

By mid-2003, Harris said, the common view of Office was "good enough". Yet when his team talked to people who used it every day, they heard frustration. Users suspected a feature must exist but couldn't find it. Others felt the suite was powerful but that they couldn't use it well.

Harris's conclusion, as quoted by XDA, was that the interface was "actually failing our users" and that Microsoft had "vastly overestimated" how well it worked. New features arrived with each release, but hardly anyone found or used them.

The core problem was sprawl. Harris walked through the history in his MIX08 presentation:

  • Word for Windows 1.0 had a Format menu with five items.
  • By Word 2003 there were more than 250 discrete menu items, along with dozens of toolbars and task panes.
  • Harris projected 45 to 50 task panes by Office 2007 if nothing changed. That's his projection from the talk, not an audited count.

Harris didn't blame menus and toolbars themselves. He called them a strong paradigm that was designed for less feature-rich programs. When a command had no obvious home, Office just added another place to look.

What the telemetry actually showed​

Harris described opt-in data collection through the Customer Experience Improvement Program. Notes from a UX conference talk record over 3 billion anonymous sessions and nearly 6,000 individual data points. Harris said Microsoft was gathering about 2 million sessions a day. He also said the data could show which commands were used, how they were sequenced, and whether they were triggered by keyboard or mouse.

The talk described how this changed internal arguments. In the old days, feature debates were won by whoever had the loudest voice or most seniority. Data gave the team a way to check those assumptions.

The findings were sometimes counterintuitive:

  • Paste was the top command in every Office app except Outlook, where Delete led. Harris said Delete was used 14 million times, against 1.8 million for Send.
  • Save was clicked in the toolbar 80% of the time, despite the common belief that everyone uses Ctrl+S. XDA reports that Undo and Paste showed a similar mouse-heavy pattern.
  • Some commands were nearly unused. Harris cited a handful of people converting a stacked chart to a pyramid diagram, and a single shape being changed to a lightning bolt.

Treat those figures with care. Harris described the command ranking as roughly one week of Office 2003 data. That is a snapshot, not a verdict on every user or every era.

The telemetry also came with a bigger claim. Harris later wrote that telemetry showed 90% of users used less than 10% of the product's features. That is his own historical summary, so it's best read as his claim.

Data plus people: how the design came together​

Calling the Ribbon "data-driven" is fair, but incomplete. Harris laid out a four-stage process of research, design principles, prototyping and evaluation. The research side included more than 10,000 hours of video of people using Office, plus interviews and visits. Telemetry told the team what people did. Observation helped explain why.

Harris also said the team wrote design tenets. These included "straightforward is better than clever" and giving features a permanent home. The tenets were used to reject ideas, such as a proposed Clippy-style helper.

Prototyping was prolific. Harris said the team accumulated roughly 50 to 60 GB of prototypes, including 25,000 images. Those figures came from his talk and were repeated by XDA.

The Ribbon's origin was almost mundane. The team pasted the Standard and Formatting toolbars onto a blank canvas and drew tabs above them. A side-mounted Ribbon was discussed, but evaluation favored keeping controls at the top of the window.

Evaluation went well beyond short lab sessions:

  • Harris described longitudinal usability testing over weeks and months. He argued that a two-hour lab session only tells you how a design works for the first two hours.
  • He described a tool called SortIt, which tested how people expected commands to be organised.
  • He also mentioned paper prototypes, eye tracking, and feedback from 3 million beta users.

Several ideas were tried and dropped, including object commands placed next to a selected picture and an "immersive" mode where Word turned into a picture app. Testers found the latter confusing. Contextual tabs, which appear when you select an object such as a picture or chart, came out of that exploration.

More than a Ribbon​

The Ribbon is the famous part, but it was one piece of a wider Office 2007 redesign. Harris lists Galleries, Live Preview, the Mini Toolbar, Contextual Tabs, Super Tooltips and KeyTips alongside it.

  • Galleries bundle commands into visual dropdowns that show the outcome.
  • Live Preview lets you hover to see a formatting result before applying it. Harris says the aim was to end the cycle of applying formatting and then clicking Undo.
  • The Mini Toolbar puts common formatting commands near selected text.

Harris says the Ribbon unified menus, toolbars and task panes into one consistent bar across more than a dozen Office apps.

Was it "hated"?​

XDA's headline says the Ribbon was hated on arrival. The evidence here supports a narrower claim. A major change disrupted muscle memory, and the author recalls power users reacting badly. None of the sources measures how widespread or intense that backlash was.

Gates's 2008 remarks complicate the picture. He described the Fluent UI as a major risk and said Microsoft had usability tested it heavily. He added that it had received a very strong positive reaction. He also pointed to third-party applications adopting Ribbon-style interfaces.

That is Microsoft leadership assessing its own product. It isn't independent research. Harris also reported that 88% of people said Office 2007 was more fun to use, which is again Microsoft-reported data.

Both framings can be true. A design can be disorienting at first and still help people discover features once they adjust. The sources here don't prove that every user benefited, or that the Ribbon alone caused broader feature use.

What this means for IT admins and developers​

For anyone who supports Office users, the story offers some practical lessons.

  1. Don't trust assumptions about habits. The claim that "everyone uses shortcuts" failed against actual usage data. Users who seem keyboard-savvy may still reach for the mouse.
  2. Expect change friction. Even a well-tested redesign breaks muscle memory. Training and a clear explanation of where things moved matter for any UI shift.
  3. Pair telemetry with observation. Harris's team used numbers to see what happened and interviews and long-term testing to understand why.
  4. Discoverability beats feature count. Adding capabilities without making them findable doesn't help users.

The Ribbon's influence has lasted. Ribbon-style interfaces spread to other developers' applications, especially tools for Microsoft-related products. The XDA piece notes that LibreOffice offers a Ribbon-style option too. Microsoft itself has since introduced a redesigned tabbed ribbon as part of the Office 2021 UI refresh.

Bottom line​

The Ribbon's story is less "data decided everything" and more "data challenged assumptions, then people and prototypes did the rest." Microsoft's own accounts say it improved discoverability, though those accounts are self-reported. Harris's talk remains a useful case study in how telemetry, usability research and design principles can work together on software with a huge feature set.

 

References

  1. Office's ribbon was hated when it arrived in 2007, but it's an unsung hero of data-driven design XDA 2026-11-17T15:31:22+00:00
  2. Bill Gates: Office System Developers Conference 2008 - Stories news.microsoft.com
  3. Jensen Harris jurecuhalev.com