Save For Later

Firefox and “save for later”
All browsers support two functions: searching and revisiting. My research questions whether constructs like bookmarks really are the right model to support revisiting. I worked closely with Mozilla user experience researchers and designers to rethink how Firefox can better offer “save for later” in the browser. Firefox front-end developers and product managers for mobile and desktop were also involved throughout the process.

Timeline for this nine-week exploratory study

The request for this study initially came from the Firefox for Android team, who wanted to know why bookmarks usage was minimal on mobile devices and if users would be more likely to bookmark if the feature was designed differently.

Screenshots from one build of Firefox in 07/2012

I evaluated all instances of bookmarks in Firefox. I found that the bookmark metaphor and its associated visual elements were being used in inconsistent and overly complex ways. Instead of focusing on how to enhance the bookmarks feature, I wanted to know why we bookmark and what bookmarking means to users today.

Central research question

Ultimately, I determined that users bookmark as a means to save for later, so this became the focus of my research.

Research phase
Out of 60 profiles, I carefully selected ten people to participate in this study. These participants represented a wide demographic of users.

A set of three diaries was mailed to each participant

I asked my participants to record their habits for three days, with each day focusing on a different theme: saving content, revisiting content, and bookmarks deprivation.

Day 1 – Saving: What are you trying to save? How are you saving this? Why are you saving this?

Day 2 – Revisiting: What are you revisiting? Describe how you got there. What triggered you to revisit this?

Day 3 – Bookmarks Deprivation: What do you want to do (create a bookmark / access a bookmark)? Try to do this without using a bookmark, and describe your experience.

One participant shows me an application he uses to manage his bookmarks

After participating in the diary study, my participants came to Mozilla HQ with all of their devices for an hour-long followup interview with me. During the interview, I asked them to think aloud while performing tasks on their devices. These tasks gave me insight into how behavior varies across different devices as well as what the interplay between these devices is like.

Participants sorted their saves into self-defined categories

I took apart the diary entries from the three days and asked the participants to sort their saves. The card sorting activity provided a high level view of what types of information people perceive their saves.

I watched the recordings of all of my interviews and pulled out important quotes from them.

I sorted hundreds of interview quotes into categories

This affinity diagram shows the different ways people save for later. Colors are associated with different users. The diagram allows me to see commonalities across different user and device types.

Different save for later services are grouped into columns

Some notable save for later methods

In the browser, leaving windows and tabs open was one of the most frequently used save for later methods. Users enjoy the convenience of picking up where they left off.

The save for later method that surprised me the most was email. Almost everyone I interviewed described instances where they needed to email a link or content to themselves for later use, which made this the most popular method across all user types. When participants needed to save content from one device and view it on a different device, email offered the most reliable way of saving. In retrospect, this makes a lot of sense since email is the one tool we consistently take care of and it is accompanied by powerful features like search and tagging.

Various applications and services offer save for later functionality

Many participants cited using one of the above services, which I collectively group as save for later services. All of these applications are essentially places where we save for later. Some are focused on specific forms of content: Dropbox saves files for later and Evernote saves notes for later. Others, like Pocket, Springpad, and Kippt, are more general.

Six things people do with their saved stuff

So what happens “later”? I consolidated the types of things people do with their saves into six categories. The emotional quality of saves varies based on the save for later mechanism. For instance, a user may save a series of photos to a Pinterest board, and in contrast, these same photos can be saved in a bookmark folder. The two methods serve the same purpose but vary on the emotional spectrum.

(1) Consume: Once you’ve seen it, watched it, read it, heard it, or interacted with it, the save loses its value (ie. funny cat video, news article about iPhone unveiling).

(2) Reuse: You use and reuse this save regularly, and the content is often dynamic (ie. Caltrain schedule, Google calendar).

(3) Share: You save it in a way that allows it to be used by someone else (ie. Delicious bookmarks, Tumblr favorites).

(4) Nothing: You save it because you think it may be useful later or for archival purposes (ie. Gmail starred items, Twitter favorites).

(5) Organize: You save it and group it with other saves based on relationships (ie. Pinterest boards, bookmark folders).

(6) Clean Up: You save it and delete it.

Analogy for types of things people do with saved stuff

The typical work desk is a great real-world analogy for what saving looks like online.

“Later” varies across devices

All six activities are fairly common on desktop devices. Phones and tablets are most commonly used for specific activities. Users commonly save reusable and shareable content on phones, and tablet content is consumed and often cleaned up afterwards.

Quantitative research

About:Home is essentially a dedicated bookmark

Based on Test Pilot data, we can infer that people like having a set homepage, but what Firefox offers as default, About:Home, might not be meeting those needs.

Ways we can bookmark in Firefox

Survey of over 5,000 Firefox users

Although 73% of users knew how to bookmark from the “Bookmarks” menu, only 36% of users actually use this method. Clicking the star in the URL bar was the most favored technique.

I leveraged data from existing Firefox chrome heatmaps, which show that although there are many access points for bookmarks, there is no clear solution.

Design brainstorm

Based on my findings, I synthesized four design principles to guide our work and explore how Firefox can better support saving for later. These principles are:

(1) Ready For You, Wherever You Go: People want their saves to be readily accessible on their device of choice, wherever they are, whenever they need it. Saves need to be offered in a form that is instantly usable.

(2) Embraces All Inputs: Saves are diverse, from articles to videos to feeds. The system should respect the natural shape of content and easily accept exactly what people want to save—nothing more, nothing less.

(3) Builds Confidence: The more information people save to a system, the more valuable the system becomes. People need to feel assured that their saves will always exist in one place. They want to be certain that they’ll be able to effortlessly find what they’re looking for if the need arises.

(4) Use It However Or Never: People want to keep their saves in a place that is flexible enough to support a multitude of things they might want to do later. It ultimately allows people to use their information however they want to, if at all.

UX designers, here and remote, sketch concepts

I led a structured brainstorming session with our UX Design Team to generate many concepts as rapidly as possible, each one focusing on a particular design principle. I limited ideation to around 7 minutes per design principle with breaks and discussion in between. Afterwards, I helped organize the concepts based on themes that emerged.

Sketches for Dropzilla, an integrated save for later space in your browser

One of the concepts I came up with, Dropzilla, aligns with many of my design principles. It allows users to drag and drop a favicon to a dedicated save for later space in their browser. The interaction consolidates the saving bookmarks and organizing bookmarks, since the user sees exactly which pile their save is going. I envision this space being tightly integrated with other save for later services. Dropzilla presents saves in a highly visual format, and users can hover over their saves to see a drilled-down and content-specific view of their save.

Medium fidelity mockup showing the core interactions

Mozilla is uniquely positioned to deliver an integrated save for later solution like this that preserves choice on the web.

Dropzilla is simply one concept that came out of the brainstorm—it is not the perfect solution. Mozilla UX will continue to ideate based on the design principles that came out of this research. One of my biggest takeaways from this process is that our current definition of the browser should not dictate or limit future possibilities of what a browser can be.  Find my talk on save for later here.

59 comments on “Save For Later”

  1. Paul Neuhaus wrote on

    Great study and concept!

    I’m gonna save this for later to go through it with more time. ;)

  2. Ken Saunders wrote on

    What fascinating and quite awesome work.

    The only save it for later services that I use are Firefox bookmarks and shortcuts saved to my PC.
    I simply don’t trust my data with any cloud or other service primarily because I don’t want to be tagged, grouped, indexed, profiled for marketing or any thing else.
    So, I’ll be glad when y’all get Dropzilla (or whatever it will be) going.

  3. Death Star Canteen Staff wrote on

    I save for later with 3 purposes, archive (things like GIFs that might be useful in several ocassions, or videos that cheer me up) that I keep in my bookmars, or otherwise in Evernote to access them offline on mobile (such as recipes). I use Instapaper for articles, because I rarely read at the computer, reading a certical screen seems awkward. And the 3rd is what you observed very well, tabs that are left open because I think they’re too valuable to just close them, yet not enough to read them now, I almost did this with this article, but I’m happy I didn’t because it’s nice to read Mozilla is working on this.

  4. Ping from Ideating without Borders at the Firefox UX Concept Week | Mozilla UX on

    […] Firefox features. Our user research team has conducted a number of great studies this year (such as Save for Later and Identity on the Internet), and we wanted to devote time to translate their insights into […]

  5. Unpersuaded wrote on

    Nice workflow! But I have 1500+ bookmarks… No new system will capture me unless it gives me a way to maintain those somehow.

  6. Manuel Strehl wrote on

    Good to see, that the old “bookmarks” metaphor gets an overhaul. I find the concepts quite interesting.

    What I’d love to see is a marriage between Panorama and bookmarks. That is, having bookmarks (or folders of bookmarks) as a persistence mechanism for Panorama groups and restoring groups in the Panorama view seemlessly from bookmarks. I reckon, that would also give the far too little respected Panorama a better boost. Since I discovered it, I won’t go to a browser without.

    That dream might be a bit off-topic here, but on the other hand it could serve as a starting point for re-accessing groups of saved websites.

    1. Avner Shanan wrote on

      Yes! Integration with Panorama was the first thing I thought of when I saw that last slide. I was surprised that it wasn’t mentioned as a way that people save things for later since that’s the main way I do. (Maybe the Panorama numbers are rolled into the “leaves tabs open” group.)

    2. Tempura wrote on

      I would more like to see an overwork of panorama itself. Actually it’s a pain to manage all those overlapping groups of icons which always change their own size at every opportunity :(

      Personally I use panorama as some kind of bookmark-holder, simply because it’s faster to move tabs create/delete groups, than handleing thos stupid bookmarks. And thanks to crashproofed sessionmanagment there is always some kind of persitence. But well, just if you work on a single machine. A bookmark-like persistance, which will automaticly synced an reopend at other machines would indeed be a dream.

    3. David Regev wrote on

      Yes! Some sort of conceptual merging of bookmarks and Panorama is what we sorely need. Imagine if Panorama were zoomable, so you could hold a virtually infinite number of items in it. Bookmarks would be obsolete, because you wouldn’t need to bookmark items or put them in folders, or even to learn how to manage bookmarks; all you need to do is leave the page around wherever you want in Panorama. Because Panorama is so visual, “bookmarking” would have a better visual design then currently, be almost immediately graspable, and become dead simple.

      I created a whole concept with a couple of mockups based on these ideas. Anyone interested should read it and comment.

  7. Leif wrote on

    Really awesome research and interesting results! I’m curious what you’ll come up with in the end.

    But *please* make sure it’s asynchronous: I use Pocket, but when I have bad connectivity (subway, etc.), email so far is the only reliable way for me to save URLs for later.

  8. Zaid wrote on

    I tend to use the panorama feature of firefox as an adhoc “Read Later” in combination with Pocket. What you are proposing would most likely make both obsolete for me! :)

  9. Rick Measham wrote on

    I love that you’re doing something with the horribly broken concept of bookmarking!

    Many years ago I began work on a plugin that I called “Glancemarks”.The plan was to create a new interface for “Glancemarks” — things that I thought I’d want to come back to, but not things that I MUST keep forever.

    It never got beyond the concept.

    The intention was to take everything you ‘glanced’ at and index it, then provide a search engine for finding glancemarks later. Essentially creating a private search engine.

    TO that end I’m quite surprised I haven’t seen Google come out with a “private search” that implements this sort of idea. Google could make an awesome [bookmarking] service that let you search your thousands of highlighted pages, using their existing technology.

    1. Rob Hammond wrote on

      Interesting – I had the same idea and came up with a proof of concept on; essentially delicious with a full search engine layered on top; I also had the idea of something sitting in the browser that would keep track & sandbox your browsing history for search, but that’s technically challenging to pull off.

      I think search is a big missing piece in bookmarking; ubiquity is the other – when you tweet/like/star stuff regardless of the service (or browser), you should be able to access them anywhere on any device. I’ve got Twitter, Facebook & Github integration on bkmrx but ideally all such services would hook into it.

      This article has some great research on how to contextualize the information once you’ve got it. One example of how this is done on is by adding ‘to-‘ to a tag will automatically place it in a special list such as ‘to-read’, ‘to-watch’ etc for ease of access depending on which content ‘mode’ you’re in. Not perfect by any measure as there would ideally be some automation there, but at least there’s some control over lists that might be more ephemeral than standard tags.

  10. Ray Barnhart wrote on

    This was a great read, thanks for posting! I love seeing UX research. Can’t wait to see what features come from this!

  11. Randall wrote on

    I would echo what Manuel Strehl said. I use Panorama (apparently most people don’t, though!) and the Pocket add-on. The pocket add-on adds a pocket icon to the edge of Panaroma and lets you drop whole tab groups into it for reading later.

    One problem I have, though, is that my typical workflow involves processing my Pocket queue only when mobile. Since I don’t have a dataplan, the selection of Pocket as a tool was specifically designed to facilitate getting through a backlog of reading while traveling. As a result, I also have a “Later” tab group where I keep videos, artist pages for musicians, websites, etc that I want to go back and watch, listen, play with.

    Bookmarking interactions that centered around Panorama would make great sense to me. Exposing my Pocket *as a tab group* would be ideal, rather than as a place I move tabs out of groups and into.

    Thanks for the interesting write-up.

  12. karl wrote on

    In the buckets metaphor (last image), I would love to be able to export one bucket as a list of title and links (and maybe comments) to be able to publish a blog post with the content either in markdown format either in html.

    I would love also to be able to use that features as a weekly dump links for a blog post.

    The issue I have the buckets is their unique dimensions. The same tab can belong to many categories. Be in multiple categories. The bucket could be views the same way we can manage emails and have views depending on search criterias.

    I would love also to have a very subtle information about when a tab has been added to a bucket on a tab: “42 days”.

    The buckets are cool and kind of work when not many tabs, but there is a category of users (which I belong to) who have around 200-250 opened tabs at the same time. When there is a lot of information some UI paradigms just explode, specifically those relying on real estate in the UI.

  13. Jon Loldrup wrote on

    What I miss about bookmarks is this:

    1. No frills marking. I just want to click the star and have the site remembered. As it is now such behavior will create the need for follow up cleanup work, i.e. moving bookmarks into suitable folders or apply suitable labels to them. If I don’t do this work I will get a crazy long bookmark menu in which I have no chance to discover what I’m looking for.

    2. Being able to Google-search the subset of the internet that I have at some point in time marked. Current methods of searching bookmarks in browsers only searches their titles!! This is profoundly misguided.

    Thus I don’t need a fancy interface. I just need a star button and a slightly customized Google search. But I guess this solution is just too simple.

  14. josé moreira wrote on

    Pure browser feature bookmarking features (meaning not extension based) have evolved much slower than other features ( or to be exact, concepts) and what i take in the end regarding the article are UX impromevents on the existing features but not much about concepts, namedly searching and perhaps organization.

    By search i mean the ability to, on the bookmarks page, search not only by the bookmark

  15. josé moreira wrote on

    Pure browser feature bookmarking features (meaning not extension based) have evolved much slower than other features ( or to be exact, concepts) and what i take in the end regarding the article are UX impromevents on the existing features but not much about concepts, namedly searching and perhaps organization.

    By search i mean the ability to, on the bookmarks page, search not only by the bookmark title but by the URL content itself, which would mean adding a page search indexer to the browser. When a usee bookmarked a page, it’s contents would be indexed (taking into account the security implications)

    I believe people already do it today on a sense, it’s most of the time easier to search Google than browse dozens or more bookmarks.

    As for organization, any machine learning (if im addressing the correct concept) would probably be better than my current method, because a correct bookmarking process “pain” increases exponentially with the number of bookmarks a user already has.

  16. Ray wrote on

    Agreed with Manuel. Tab Groups (or Panorama) is almost exactly the same as what is noted in the Bookmarks roundup.

    Tab Groups just needs UI-refining as it’s currently hard to easily revisit groups. Some Firefox addons like Pano do a good job at trying to add the extra functionality needed.

    I use Tab Groups for things I’d like to revisit sooner rather than later. I star something when I want to revisit it some time further into the future.

    So try improving Tab Groups first and merging Bookmarks into it.

    Thanks for reading!

  17. Aaron Zinman wrote on

    We actually built this about year ago, also taking over the new tab page into these theme’d boxes. Rather than dragging you probably want to add buttons next to the groups that say “Add current tab.” That will be much more clear for ordinary people.

    One thing we noticed was that we quickly forgot what was in those boxes and what tasks we needed to resume. This is particularly true for low-priority tasks which is exactly the kind that you’re saving for later in the first place. Bookmarking already works well for something you know you’ll use tomorrow… it’s that pesky “in the future” that creates a cognitive disconnect between current and past tasks.

    To push this further we need to think about how to engage the user to be more active and less lazy. Not everyone is a “spring cleaner” that frequently goes through their growing collection of low and medium-priority tasks. One way is to enforce constraints on how many groups may be created, or how many items can be put into them. But even then we found we wouldn’t remember to return to old tasks or to schedule time for them. It’s a very difficult problem (user attention and task prioritization) and the UX to solve it needs to be much deeper than what’s covered here.

  18. Mauricio Angulo wrote on

    It’s nice to know that people are paying attention to the ‘save for later’ issue in the browser. I believe that the label ‘bookmark’ has fallen short to describe the carefull curation an user does to the web information that is relevant for him. Obviously in a multiscreen/multidevice enviroment the user experience has to be fluid and consistent in all devices and platforms, and has to be easy to access and organize. Keep up the good work!

  19. Frank Colcord wrote on

    Please look at Zotero, which is what I’ve been using for a long time to save material for later. It lets me take a snapshot of a page, tag it and search for it later. It also keeps the data in a format I can use for citations.

    I agree that the user interface for Zotero leaves much to be desired, but the power is tremendous.

    If you are going to do your own work on this, please allow for some kind of data transfer/sync between the two.

    I also currently use the google/ig personal page to hold my book marks in a clear format.

  20. Mike X wrote on

    I have thousands of bookmarks and hundreds of open tabs, too many to keep well organized. What I’d find most useful is the ability to search on the page content of my history. Think gmail. The awesome bar works well as a first approximation, but searching page titles and tags is inadequate. Creating folder hierarchies and tags is too much work. To keep the index small and clean, maybe the bookmark star in the URL bar should be used to mark a page for inclusion in this local search index.

  21. GS wrote on

    I’m actually quite interested in this work. I tend to bookmark things I want to visit properly when I get the time. This may be months or years in the future. They build up over time, numbering easily in the thousands and it’s hard to isolate duplicates and even search for specific bookmarks.

  22. Dror Harari wrote on

    One thing that bothers me with bookmarks and their ‘safety’ is that it happens in many cases that the web breaks and the bookmark I kept (for some reason or another) loses its meaning. This is also relevant for sites that do not capture the specific page in the URL so while as a user I probably wanted to bookmark what I see, I ended up bookmarking nothing.

    Something that might help is a capture of the bookmarked page (either as an image or an archived page) that would be there for me when I revisit even if the site is no longer or if the site has changed (e.g., bookmarking pages in many newspaper online sites feels like trying to bookmark sand grains). Such a facility could also make for a great way to pre-render a bookmarked page until it gets loaded.

  23. Plano wrote on

    Congratulations for this awesome research into an almost forgotten aspect of browsers.
    One thing to consider is an easy way for users to back up, restore and merge the saved bookmarks.
    I look forward to see Dropzilla ready!

  24. Rafael Muñoz wrote on

    Fascinating stuff how all the process was done …

    Jumping to the concept: Dropzilla seems to align very well with the already present Firefox Panorama adding syncing and some UI tweaks. Is that your idea?

  25. Satyajit Rai wrote on

    Why only drag and drop?

    Did you consider presenting the bookmark organizer screen when the user clicks on the star in URL bar and provide a way to select a bookmark group without having to drag the favicon?

  26. Tom Carver wrote on

    Very interesting work, looking forward to seeing these ideas implemented in the future! One thought, the visual metaphor (while awesome) has the side effect that the user *has* to think about where to put the favicon; sometimes I just want to “fire and forget” without having to think about link organisation. All I’d add to enable this would be a “Misc”/”Default” pile where bookmarks saved with the star could go; then the user can organise them later with the same interface when they’re ready.

  27. Funka! wrote on

    I’m currently using firefox built-in Tab Groups in almost the exact same manner as the Dropzilla screenshot seems to indicate. At any moment i may have up to 200 tabs open across a dozen or more groups, roughly categorized based on a project i might be working on, a particular topic or technology I’m researching, and definitely the “save for later reading” groups. The only drawback—as this article suggests—is that these tab groups exist only per device, at least i think they are, so that my tab groups on my home computer are totally different than the tab groups on my work machine. Not that this is a bad thing, since I always have email to sync a link here or there as needed! :-)

    One interesting thing I’ve found however is that NONE of my peers seem to know about this new feature. (Since what, version 13 or 14?) I only knew about it because I read blogs like this. But I think it could benefit from more obvious positioning/discovery.

  28. mustefa wrote on

    It was great to learn about the content of your research.

    I enjoyed even more the process of the research and how you outlined and presented your findings.

    Very cool, Brian.

    1. mustefa wrote on

      At the same time, I have a hard time from pulling away from a native dedicating browsing / consumption experience on OSX and iOS.

      Safari let’s me add a reading list, and save bookmarks. Saving bookmarks seems to me like they’d get lost in the ether.

      And I can synch this seamlessly between devices. That’s the biggest blocker from jumping onto doing things in FF.

  29. R. Kyle Murphy wrote on

    I must confess I use both Pocket, and the “open a bunch of tabs and never close them” methods of saving stuff. In either case I tend to get buried in clutter after a while, so something that makes organizing these things simpler would be awesome.

    I’d love it if there was some integration with a centralized system for synching these psuedo-bookmarks, but I’d also want such a system to be both optional (I.E. you can have local only bookmarks), and also support partial synching (E.G. I’ve got a work space and a personal space, I want my work space on most of my devices, and my personal space on everything but my work devices). Dropbox for example has a really great system for handling this where you can select which folders you want to synch per-device which would be ideal, although it doesn’t support local only content because in the instance of Dropbox you already get that with your normal filesystem.

    Ideally the ability to synch a particular space to a centralized repository shouldn’t require creating a new space, but a simple toggle on the existing space.

  30. Jon Loldrup wrote on

    What I miss about bookmarks is this:

    1. No frills marking. I just want to click the star and have the site remembered. As it is now such behavior will create the need for follow up cleanup work, i.e. moving bookmarks into suitable folders or apply suitable labels to them. If I don’t do this work I will get a crazy long bookmark menu in which I have no chance to discover what I’m looking for.

    2. Being able to Google-search the subset of the internet that I have at some point in time marked. Current methods of searching bookmarks in browsers only searches their titles!! This is profoundly misguided.

    Thus I don’t need a fancy interface. I just need a star button and a slightly customized Google search.

  31. Tanner wrote on

    I am actually currently working on something to solve this. It’s too much effort to use the current bookmarking system to save temporary files, and this leaves me with a ton of tabs open. There are other apps and plugins that try to solve this problem, but either they are poorly made or too powerful/complex for what most users want.

    Although, my solution would be through a plugin, which would be much less convenient than a native browser solution. I hope browsers start working on this.

  32. David Barker wrote on

    Excellent piece of work! I have tried various methods of managing “links to look at later”. I used to email them to myself. Currently I use the speed dial to manage these links. I find the speed dial particularly useful as it gives me a visual representation of the links, but it is obviously not really designed for this. I use Diigo for my actual bookmarking but don’t use it particularly effectively.

  33. Jon wrote on

    Love it. I’ve been meaning to download Firefox for Android (I made a mental bookmark), but after reading this post I’m going to d/l right now. Excellent work.

  34. Gordon wrote on

    Wonderful research and definitely interested in the results.
    What were the differences between bookmarking on smartphones versus laptops?

  35. Caoimhin wrote on

    Interesting process. I wonder though, how a ‘Dropzilla’ solution would follow me on my phone, tablet etc? Maybe it could have a weekly email shoot off to yourself with a summary of the weeks ‘Save-For-Laters’. This would create a generic, user-owned archive. There could be a link in each mail to the master archive. Accessible in any browser, on any device. …Just a thought.

  36. ImaCrea wrote on

    Truly fascinating, thanks for sharing the whole process behind the scene. Inspiring!

  37. James wrote on

    Great to see thinking outside of the status quo. Traditional bookmarking just doesn’t work for me, and a I would love a more visual system like this. An awesome idea (although something I’m sure other browsers will quickly clone…)

  38. Alexandre wrote on

    Awesome work guys!

  39. Dustin Rodriguez wrote on

    The answer seems to simple to me, I don’t understand why it is not obvious to everyone else. Perhaps I am missing something.

    As you found, many people leave open many tabs or windows. They leave many windows open because the design of tabs is incompetent. It’s better in Firefox, where you can fix it with the TabTree extension, but it’s downright hostile in Chrome where they HAD a vertical tabbing solution but tore it out in order to hurt the users who enabled its experimental support before.

    So why do tabs have to close? Ever? Why can’t I, as a user, simply scroll backwards through a list of tabs as long as my entire browsing history? And when a prior tab is accessed, it bubbles immediately to the top, treating the list of tabs like a modified stack. If I kinda remember I was using the page yesterday, I’m going to remember it in the context of the other pages I was using, making it much easier to browse through and find things.

    This would obviously be a lot of work to do right. Firefox already has many problems with memory usage, and implementing this feature would be impossible if they continued with the same ‘once a tab is loaded into memory it stays there forever until closed’ practice. And loading the page again off the network really should be an absolute last resort. If I click a tab from a page I had open 3 weeks ago, it should pop up as quickly as streaming the content from disk permits to happen (which is quite fast).

  40. Srap wrote on

    SFL (Save for later) tabs may borrow the look of Panorama, but unlike some of the commenters, I am against the complete merge of Panorama, the Bookmarks and the (future) SFL tabs. Their main purpose is just too different for this.

  41. Ken Saunders wrote on

    Great start


    Functional prototype for a different take on Firefox’s Places (bookmarks and history) browsing features, implemented as a Firefox add-on. Also displays tab groups, like Firefox’s Panorama feature.

    The philosophy behind Mozaic is to decouple content and view. There are three types of content: bookmarks, current tabs, and history, which could possible be extended (perhaps as modules) in the future. There are two types of views: list and thumbnail grid. Each type of content is viewable with each type of view.”

  42. Alex wrote on

    Aside from the main point, what would really make bookmarks far more usable is if it had an offline snapshot of the page attached to it. This would solve one particular usage scenario, which is bookmarking the page as a reference design.

    Also, if it’s Save for later, then it makes sense to actually *save* something but the URL.

  43. Russ Brown wrote on

    Brilliant read, great work. I use so many different things to save for later, a stable tool that keeps everything in one place but is still flexible for all devices is well overdue. Look forward to seeing this feature soon.

  44. Kristina wrote on

    Thanks for sharing the process and outcomes so far of your study. Taking a step back and looking at what you are actually doing with bookmarks is great because some new, unforeseen insight might emerge. I had a quick think about for what I use bookmarks primarily and came up with 4 scenarios: 1) archiving interesting content, 2) sharing pages with others, 3) reading later, 4) quickly accessing often used sites.

    Diigo is the service I use for the first three scenarios because I have everything in one place and can share any bookmark with others via a group or list. I can simply send them off to either a group or even to my bookmarks and let them explore without having to dig in my bookmarks myself and try to make a good selection. The fourth scenario is a minor one because usually I know the URLs of sites that I frequently visit and just start typing the URL instead of trying to find them in a bookmark menu. However, I do use keywords for some sites to simply type that and come up with the page I want, or I use the bookmark toolbar. I cannot / do not want to save these bookmarks on Diigo because they are partly to internal company pages and retrieving them from Diigo is a bit time consuming.

    Thus, for me, it is important not just that my bookmarks are accessible across devices but that I can give others easy access to them as well.

  45. Ping from ElectroMake-Electronic Gadgets | Mozilla looking to redesign browser bookmarking with Dropzilla on

    […] Save For Later […]

  46. Symac wrote on

    Is this possible to download the slides somewhere ?

  47. Moonlight wrote on

    Great Job!! Brilliant ! Wonderful research and definitely interested in the results . Congratulations.

  48. Seleko wrote on

    What about to return “Send link/page” to context menu?

  49. snjflame wrote on

    I use “Smartest Bookmarks Bar” ( and I simply drag tabs to it,
    and also use “Firefox – Bookmark icon only hover selector” – userstyle

  50. Calvin Spealman ( wrote on

    I am really delighted to see this deep effort made in improving the bookmark workflow and, more importantly, meeting the needs of the reason bookmarks exist in the first place. I think, originally bookmarks existed because searching on the web was not a viable tool. Today, bookmarks needs to fill different shoes. Thank you for the work you’re doing in figuring out what that means!

More comments:1 2