Firefox3/Themes for Data Collection

Discussion document for identifying long-term areas of focus for market research and data collection efforts for Firefox, and short-term identification of target differentiation vectors.

Please try to adhere to the following structure if possible:

• the name of the theme, just to act as a label
• why the investigation is worth the time and effort
• what the scope of investigation should be (long term? immediate?)
• the questions that need investigating

THEME: Just what are our users doing with our product?

Why it's important: The Firefox charter states that the browser should be serving the majority audience in their online tasks, but there is no recent data (especially since the Web 2.0 explosion) showing what users are actually doing online. It's easy to make assumptions based on personal and peer group experience, but those results skew to our own user bases.

Scope: immediate & long-term

Questions to investigate:

• who are our users, and how much are they doing online?
• what's the average user:browser profile ratio?
• what sort of tasks are they doing online?
• what do they want to be able to do more of online?
• what is the relative usage per type of online task?
• what kind of websites do they visit?
• how many unique websites do they visit?
• what percentage of time is spent visiting new websites as opposed to returning to previously viewed sites?
• how are users collaborating and sharing online?

THEME: Making the browser disappear

Why it's important: The best kind of tool is one that becomes a natural extension of the individual. We should aspire to entirely disappear from the user's conscious view and just be an extension of themselves.

Scope: immediate

Questions to investigate:

• what are the most common user annoyances with web browsers?
• how do web browsers and web applications conflict in terms of UI?
• where does Firefox interrupt instead of assist common user taskflows?
• what UI/tools/features are not used or needed in the product?

THEME: What is the competitive landscape for web browsers?

Why it's important: The choice our users will have to make is, "why Firefox?", so we should understand the factors on which they will be basing that decision. Getting into feature wars is uninteresting, but understanding long term competitive strategies is a must.

Scope: immediate & long-term

Questions to answer:

• what is the emerging strategy of our competitors?
• what features are they developing?
• what is their messaging?
• who are they partnering with?
• how does the strategy of OS vendors affect the web browser and our ability to compete?

• what OS integration points can they leverage (security, desktop search)?
• what desktop application integrations can they leverage (MS Office, Apple iLife)?
• what online service integrations can they leverage (Windows Live, .mac)?
• what browser runtime platforms can they integrate with their tools offerings (WinFX + Visual Studio)?

• what "costs" must a user endure to use Firefox relative to competitive offerings?

THEME: Improving a user's security while online.

Why it's important: Identity theft and phishing are getting a lot of attention in the technical press, and last year saw several studies showing that SSL indicators weren't being properly interpreted by users.

Scope: immediate

Questions to investigate:

• what are the emerging standards for website authentication (i.e. user centric identity, sxip, infocard)?
• what are the current levels of awareness about online scams?
• what are the current trust/authentication indicators and why are they failing?
• what are the primary vectors of attack for web scams?
• what are the primary vectors of attack for malware?
• what are the rates of attack, and how are attacks trending over time?

THEME: Improving a user's privacy and control over personal information

Why it's important: Placeholder text

Scope: immediate

Questions to investigate:

• what privacy features are other browser vendors providing to keep information from being recorded locally?
• what user/groups schemes are currently in use by applications or online services?
• are there ways to allow users to track the distribution of a user's personal information?

THEME: WS-* SOAs, remixable APIs and web applications

Why it's important: Over the past year the web has become more and more about online applications and services, and recently a lot of remixing of those services has heralded the promise of the "Service Oriented Architecture". Understanding how service remixing, especially user-generated, will affect the web will keep us prepared for the next paradigm shift.

Scope: long-term

Questions to investigate:

• with more available APIs and light- and heavy-weight service oriented architectures, how popular will service remixing become?
• what APIs are being currently developed?
• what APIs are predicted to be developed in the future?
• what are the predicted effects on various Firefox stakeholders?
• what are the predicted effects on web content creators?
• One thing that makes Firefox 2 different is support for third party providers: web feed readers and phishing black lists.

THEME: Making use of rich metadata

Why it's important: More and more metadata is available (Microformats, RDF, RDF/A, CC info, attributions, etc.) yet systems aren't taking advantage of it. Yahoo! is starting to support microformats on the following web properties:

• Yahoo! Local fully supports the hCalendar, hCard, and hReview microformats on almost all business listings, search results, events, and reviews.
• Flickr supports XFN and hCard on all profile pages, and Upcoming.org supports hCalendar.
• Yahoo! Tech uses the hReview microformat for all product reviews.

Being able to react intelligently to this metadata would make the web browser a more personal and useful tool for users.

Scope: immediate and long-term

Questions to investigate:

• what are the most frequently occuring kinds of metadata available?
• what other metadata exist, though not in any formal specification?
• in what ways is the browser most frequently considered by users to be "dumb" when not reacting to metadata?
• what tools exist for leveraging metadata and acting upon it?
• why did the semantic web never take off?
• how does collecting data conflict with privacy concerns?
• what are the emerging standards for metadata handling?