Because I Carry My Cell Phone Anyway:
`Functional Location-Based Reminder Applications
`Pamela J. Ludford, Dan Frankowski, Ken Reily, Kurt Wilms, Loren Terveen
`University of Minnesota, Department of Computer Science
`4-192 EE/CS Building, 200 Union St. SE, Minneapolis, MN 55455, USA
`{ludford, dfrankow, kreily, wilms, terveen}@cs.umn.edu
`
`
`malls, stores, churches, auto-repair shops, health clubs,
`laundromats, salons, and other public places [11].
`Our research focuses on understanding these everyday tasks
`– how people manage them today, what works and what
`doesn’t – and the utility of location-based reminder systems
`(LBRs), for supporting them. LBRs are implemented on
`mobile devices equipped with position sensing technology
`(e.g., GPS). While previous research has investigated
`aspects of task management and LBR technology design,
`this earlier work leaves three major gaps that motivate our
`research.
`From PIMs to everyday tasks. Personal Information
`Management (PIM) research explores the artifacts and
`processes people use
`to manage meetings, contacts,
`documents, events , etc. [13]. This research largely focuses
`on tasks performed in the workplace and information
`managed at the workstation [1, 2, 3, 4]. Researchers
`recently have recognized, however, that many tasks require
`ubiquitous personal information – untethered from the
`desktop – and have called for research in this area [3]. By
`our focus on personal everyday tasks, we meet this
`challenge. Our research has discovered a common pattern
`for doing everyday tasks: people pre-plan at a base
`(typically home or work), create information resources
`(frequently lists), and take these resources with them to
`refer to at the place where the task is performed (e.g., a
`grocery store).
`
`Can my LBR manage your everyday tasks? LBR
`research began with proof of concept implementations [14];
`a recent study explored how people create personal
`reminders for themselves [16]. We agree reminders are
`useful, but we ask: ‘Is there anything more?’ Can we
`enable new and more efficient practices? To date, there has
`been no explicit investigation of the process for completing
`tasks upon receipt of place-based reminders. We study
`practices people use to support errands and their satisfaction
`with them. A major finding of our research is that a key
`practice – creating and using lists – has some problems
`(e.g., lists are easily lost). We show other LBRs have not
`supported lists well, and PlaceMail does.
`LBR: are we there yet? The whole idea of an LBR is that
`a reminder is delivered for a location. But what does that
`mean exactly? In practice, it has meant that a reminder is
`delivered when a user is near a location, defined as entering
`
`ABSTRACT
`location-based
`Although they have potential, to date
`information systems have not radically improved the way
`we interact with our surroundings. To study related issues,
`we developed a location-based reminder system, PlaceMail,
`and demonstrate its utility in supporting everyday tasks
`through a month-long field study. We identify current tools
`and practices people use to manage distributed tasks and
`note problems with current methods, including the common
`“to-do list”. Our field study shows that PlaceMail supports
`useful location-based reminders and functional place-based
`lists. The study also sheds rich and surprising light on a
`new issue: when and where to deliver location-based
`information. The traditional ‘geofence’ radius around a
`place proves
`insufficient.
`Instead, effective delivery
`depends on people’s movement patterns through an area
`and the geographic layout of the space. Our results both
`provide a comp elling demonstration of the utility of
`location-based
`information and raise significant new
`challenges for location-based information distribution.
`Author Keywords
`Ubiquitous computing, lists, location-based reminder, PIM,
`cell phone, location-based information delivery.
`
`ACM Classification Keywords
`H5.m. Information interfaces and presentation (e.g., HCI):
`Miscellaneous.
`INTRODUCTION
`Does it never end? In today’s busy workplace, knowledge
`workers complain “I’d be overwhelmed, but it’s just one
`more thing to do” [8]. Yet when work ends, it doesn’t.
`Evenings and weekends are packed with everyday tasks:
`taking children to school, buying the groceries, maintaining
`and fixing the house, attending social events and religious
`services, etc. While the home is the base for these tasks,
`many are performed elsewhere: a survey shows that
`Americans spend over two and a half hours each day at
`
`
`Permission to make digital or hard copies of all or part of this work for
`personal or classroom use is granted without fee provided that copies are
`not made or distributed for profit or commercial advantage and that copies
`bear this notice and the full citation on the first page. To copy otherwise,
`or republish, to post on servers or to redistribute to lists, requires prior
`specific permission and/or a fee.
`CHI 2006, April 22–27, 2006, Montréal, Québec, Canada.
`Copyright 2006 ACM 1-59593-178-3/06/0004...$5.00.
`
`
`Google, Exhibit 1007
`IPR2022-00742
`Page 1 of 10
`
`

`

`
`
`(or perhaps exiting) a “geofence” around the location. Our
`research shows that setting an effective delivery point is a
`complex process. When and where users want delivery
`depends on factors such as users’ plans, motion patterns,
`(e.g., the route they usually take to the place) and the social
`geography of the area (e.g., if there are lots of distractions
`enroute).
`We address these gaps through 3 research questions (RQs).
`RQ1. What tools and practices do people use currently to
`perform personal everyday task s, and what are
`their strengths and weaknesses?
`RQ2. How well does PlaceMail support everyday tasks?
`Does it improve existing practices and enable new
`ones?
`RQ3. What factors determine when and where a reminder
`for a location should be delivered?
`The remainder of the paper is organized as follows. We first
`survey related work, illustrating how we build on or
`advance it. We then describe the PlaceMail system,
`highlighting distinctive features that appreciably affect
`usage. The heart of the paper describes the deployment of
`PlaceMail as a support tool for personal everyday tasks; we
`focus on how our results answer our three research
`questions. Finally, we discuss the implications of our
`results for future design and research.
`RELATED WORK
`Location-Based Systems . Previous researchers pioneered
`the LBR: early proof-of-concept designs include Cybre-
`Minder [6] and comMotion [14]. This work defined the
`basic idea – virtual reminders associated with physical
`locations, and the comMotion research culminated with a
`prototype built using wired hardware assemblies available
`at the time. The E-Graffiti [5] and GeoNotes [7] systems
`were technically similar, but targeted different usage
`situations such as social messaging or community
`announcements. And other reminder systems are location-
`based in a quite different sense: for example, Gate
`Reminder [10] is installed at the doorway of a house.
`Recently, researchers have implemented
`location-based
`systems on cell phones and have studied their use in
`empirical field tests. For instance, DeDe supports location-
`and time-based social messages: a sender specifies a place
`or time when the addressee will receive a text message [9].
`Similar to PlaceMail, Place-Its supports location-based
`reminders and runs on a cell phone [16]. Our efforts are
`closely related to the Place-Its research, so it is worth
`contrasting our projects. First, PlaceMail offers several
`new features not available on Place-Its. For example,
`PlaceMail has a web-based interface and a voice input
`function; these lead to distinct system usages. In addition,
`our empirical study differs from the Place-Its study [16];
`the latter centers on opportunistic reminding. In contrast,
`we investigate existing practices for managing personal
`everyday tasks. This identifies unfulfilled needs that can be
`met by LBRs : we thus establish a baseline for comparing
`
`method and tool utility. We also examine the effectiveness
`of place-based delivery heuristics and present related design
`guidelines.
`Personal Information Management (PIM). A number of
`studies characterize PIM research and illustrate its focus on
`novel systems for the workplace [1, 2, 3, 4]. Active topics
`in the HCI literature include: methods for managing
`meetings,
`contacts, documents,
`file
`systems,
`and
`outstanding tasks. Our focus is distinct – on everyday tasks.
`Processing these tasks differs – many are inherently
`distributed, with planning often done at a base (typically
`home or work) and execution done elsewhere (e.g., a
`child’s daycare center). Therefore, concepts may transfer at
`some level – e.g., to-do lists are useful in both settings,
`however, the specific s can vary significantly. For example,
`to fulfill everyday tasks, people carry resources like to-do
`lists while mobile, and this can lead to a different set of
`issues.
`Location-Based Message Delivery. Most location-based
`information systems, from comMotion to GeoNotes, DeDe
`to Place-Its, have used a simple, intuitively appealing
`method for deciding when to deliver a reminder for a
`location: deliver it when the user enters or exits some
`distance threshold (a geofence) around the location. Just
`one study that we know of has investigated effective
`delivery design. Paay and Kjeldskov
`[15] defined
`geographic regions around static attractions in a tourist area
`using architectural design principles. Like ours, this work
`defines tactics for effectively delivering location-based
`information. However, our work is distinct in that it targets
`dynamic situations. We inform delivery for users moving
`freely in the world – and their messages deliveries can be
`for anywhere - rather than in a set region with a fixed
`number of attractions.
`
`Last, Tamminen et al. [18] reveal how people move through
`urban spaces in everyday living contexts. Their work
`describes general patterns, while we target how navigation
`affects ideal place-based information delivery .
`PLACEMAIL
`Now that we have characterized related research, we detail
`the PlaceMail application design. To begin, sending a
`PlaceMail is like sending email to yourself, but with a twist.
`Instead of receiving the message in an email browser, you
`receive it on a cell phone at a time and place of your choice.
`A short tune alerts you to incoming mail.
`
`Implementation. We implemented PlaceMail on the Mo-
`torola i88s iDEN mobile phone, with Nextel wireless
`service. We chose a mobile phone rather than a PDA
`because the cell phone has been universally adopted for
`everyday use; the PDA remains a niche product. PlaceMail
`uses the phone’s built-in assisted GPS (Global Positioning
`System) to sense location.
`Interface Design. A major design goal was to make it
`easy for people to create reminder messages, a challenge
`given the tedious process of text entry on mobile phones.
`
`Google, Exhibit 1007
`IPR2022-00742
`Page 2 of 10
`
`

`

`
`
`Thus, while we implemented a phone interface for creating,
`viewing, and editing messages, we also provided two
`additional features. First, we designed a web interface with
`the same functionality as the phone interface (see Figure 1).
`The web interface emulates the look and feel of an email
`browser.
` Second, we implemented a voice message
`function: users can optionally record voice reminders on the
`phone in lieu of texting.
`
`
`
`
`message. After saving the message, the user can view, edit,
`or delete it via either the web or phone interface.
`
`When a PlaceMail message arrives, the cell phone plays a
`short tune. Messages are delivered once, but the user can
`request redelivery at a future date and time upon receipt. If
`they choose this option, the system re-activates the message
`at the specified time, and redelivers it when the user
`subsequently nears a relevant place.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Figure 1. The PlaceMail web user interface design.
`Specifying Message Delivery. Users can specify one or
`more delivery places for a message. For example, a user
`might associate a message “Check furnace filter prices”
`with several hardware stores (see Figure 2). The message is
`delivered when the user is near any of the stores.
`Users can specify a delivery date and time instead of, or in
`addition to delivery place(s). When both place(s) and
`date/time are specified, PlaceMail “activates” the message
`at the given date/time. It then begins checking whether the
`user nears a specified place.
`
`Place acquisition. As Figure 2 shows, users select message
`delivery points from a list of personally meaningful places.
`At the onset of the study, subjects provided us with places
`they commonly
`frequent. We used Google Maps
`(maps.google.com/ ) to compute the latitude and longitude
`of their places and entered this information in the PlaceMail
`database. We note that automated place acquisition is not a
`focus of
`this study, and
`that other researchers are
`investigating this issue [19]. We believe LBRs will need to
`exploit automated acquisition methods
`in order
`to
`proliferate in the future.
`User Interactions. To create a PlaceMail message, the user
`follows a simple 3-step process. After
`logging
`into
`PlaceMail on either the web or cell phone, the user: (1)
`specifies where she wants to receive the message by
`checking delivery point(s) on her place list (see Figure 2),
`(2) enters the body of the message as text or an audio
`recording, and (3) optionally selects a delivery time for the
`
`
`Figure 2: Specifying places for a message to be delivered
`
`System Architecture. PlaceMail uses a client-server
`architecture.
` PlaceMail stores user data (places and
`messages) in a database on a server. This enables easy
`synchronization between the phone and web-based clients.
`The phone client retrieves a user’s places and current
`messages at login over a wireless HTTP connection.
`During active use, messages are stored locally on the
`phone. New or edited messages are pushed to the phone and
`web client every 60 seconds. Most important, the phone
`client takes GPS readings at frequent intervals (every
`minute) and sends this information to the server, where
`computations determine whether any messages are relevant
`to the user’s current location. If so, they are delivered.
`Computing Location. We use a three-tiered procedure for
`computing location.
` First, we use the GPS reading
`whenever available. However, since GPS uses line of sight
`to satellites, it does not work if a user is indoors, in an
`“urban canyon”, or the line of sight is otherwise obscured.
`The assisted GPS software on the i88s always knows the
`latitude-longitude of the serving cell tower; we use this to
`implement two fallback methods when GPS is unavailable.
`
`In our primary fallback method, we employ the last GPS
`reading as a surrogate for the user’s location. This works
`well, for example, if the GPS signal is lost when the user
`enters a building and she is currently situated there. We
`identified 2 instances, however, when this is unlikely to be
`the case: (1) the last GPS reading is very old (we chose >
`12 hours), or (2) the last GPS reading is distant from the
`current serving cell tower (which we define as > 2 miles). If
`neither of these circumstances is true, then the last GPS
`reading serves as the user’s location. Otherwise, we
`fallback using a second method: we employ the serving cell
`tower latitude/longitude as the location surrogate. This
`method is least accurate so we use it only when the other
`methods fail.
`
`Google, Exhibit 1007
`IPR2022-00742
`Page 3 of 10
`
`

`

`
`
`the
`location,
`to
` In addition
`Delivering messages .
`PlaceMail program acquires the user’s estimated speed
`from the cell phone software. The application uses both of
`these factors to determine when to deliver a message. First,
`if the user is moving, PlaceMail delivers messages for
`places they can reach within 2 minutes, given their current
`speed. Or if the user is stationary, the system delivers
`messages for places within a half mile. As we discuss later,
`our results show
`that
`this simple,
`intuitive delivery
`procedure does not sufficiently ensure that messages are
`delivered when and where users want them.
`EXPERIMENT: DESIGN AND METHODS
`During the summer of 2005, we conducted a field study to
`investigate the utility of PlaceMail for supporting personal
`everyday tasks. We recruited subjects through community
`newspapers and mailing lists, and used an online survey to
`find participants who regularly perform everyday tasks such
`as grocery shopping, home repair, and child-related chores.
`20 qualified subjects participated. Their backgrounds
`include advertising, marketing, chemical engineering, IT
`consulting, nursing, architecture, communications, stay-at-
`home parents, and small business owners. 12 of the
`subjects have children living at home. None of the subjects
`knew members of the research team prior to the experiment.
`Subjects used PlaceMail for 4 weeks. To get started, we
`met each participant for a detailed face-to-face interview.
`The goal of the interview was to address RQ1. To this end,
`subjects brought at least one physical artifact that they use
`to manage everyday tasks. We discussed in detail how they
`employ the artifact, as well as other management methods
`and effectiveness of current practices. Subjects also learned
`how to use PlaceMail, and we assigned them an i88s phone.
`These sessions lasted 45-60 minutes. At the end of the
`study, we conducted in-depth interviews with each subject
`about their experiences with PlaceMail; these interviews
`helped us answer RQ2. The exit interviews lasted about an
`hour.
`We told subjects to use PlaceMail just as they wanted: they
`could send any type of message to any place and use any
`system feature. We asked subjects , however, to send and
`receive at least 2 messages a week. We offered a modest
`incentive: two random subjects who participated at the
`minimum level received a $50 gift certificate. 90% of the
`subjects met the requirement. On average, subjects created
`17 messages during the study. We analyzed usage logs and
`interview responses to answer RQ2.
`We addressed RQ3 by sending subjects an online survey
`after they received a PlaceMail delivery. The survey was
`administered within 24 hours of message receipt and
`inquired about
`the message’s utility and delivery
`conditions. To lessen the subjects’ effort, we did not query
`them after every receipt. Instead, we randomly sent surveys
`for 22% of all message delivery events. We sent a total of
`77 surveys and got 67 responses for an 85% response rate.
`We administered the surveys via email to elicit user
`
`comments; this proved effective as 62% included free-form
`observations that helped us answer RQ3.
`RESULTS
`Basic Usage. Subjects created 344 messages, an average of
`17 per person (min: 4; max: 31; std: 7.42). 189 (55%) were
`created with the web interface, 79 (23%) with the phone
`voice interface, and 76 (22%) with the phone text interface.
`Text messages created from the web averaged 33 characters
`in length, those created from the phone averaged 13; the
`overall text message mean length is 28 characters.
`
`51% of the time, the recipient wanted message delivery at a
`place or places, 33% at a place and date/time, and 16% at a
`specified date/time only. Of messages that specified a
`delivery place, 92% specified one place, 8% multiple
`places. The majority of messages – 61% - were for public
`places, including retail stores, parks, libraries and post
`offices. 18% were left at home, 5% at work, and 16% of
`messages had only a delivery date/time, not a place. This
`message/place distribution differs significantly from the
`recent Place-Its study [16], where subjects left 80% of
`messages at their home or workplace.
`This difference could be due to different place acquisition
`methods. Place-Its subjects had to physically visit a place
`prior to leaving a message there. In the 2 week long Place-
`Its study, perhaps subjects did not visit many places other
`than home and work. In contrast, PlaceMail subjects could
`leave messages at any of their places starting on day 1 of
`the experiment. The PlaceMail usage pattern reflects the
`distribution of everyday tasks: recall that a previous study
`found that people spend over 2 ½ hours a day at places
`commonly associated with errands [11].
`
`With this background, we now address the core of our
`results, which are organized around our three research
`questions. We begin with RQ1 and describe subjects’
`everyday
`task management practices prior
`to using
`PlaceMail. We follow by identifying opportunities for
`improving these practices, and then address RQ2, which
`addresses how subjects employ PlaceMail. To review, here
`are the first two RQs.
`RQ 1. What tools and practices do people use currently
`to perform everyday tasks, and what are their strengths
`and weaknesses ?
`RQ2. How well does PlaceMail support everyday tasks?
`Does it improve existing practices and enable new ones?
`
`Study subjects commonly employ a basic record and refer
`strategy for managing everyday tasks. 95% of subjects
`reported that they first record task-related information on a
`list, calendar, day planner, or other tool, and later refer to
`the information as needed. Typically, subjects record
`information at a base location such as home or work and
`refer to it at the place where they carry out the task. They
`occasionally update task information when mobile.
`List Types and Strategies. The most common record and
`refer artifact is the list: 90% of the subjects regularly keep
`
`Google, Exhibit 1007
`IPR2022-00742
`Page 4 of 10
`
`

`

`
`
`both shopping and “to-do” lists, typically writing them on
`paper. Lists can be sequential. For example, three subjects
`put multiple places and related tasks on a single list. They
`run the errands during a single outing, usually ordering
`them to minimize transit time. The list serves as a point of
`reference: they check where they should go next, and may
`dynamically adjust their plan if they are running short on
`time.
`According to subjects, lists constantly evolve. For example,
`85% explained that they continuously maintain a grocery
`list: when they purchase an item, they cross it off the list.
`When they get home, they start a new list, carrying over
`unfulfilled items from the old one.
`List Benefits and Drawbacks. Paper lists, often written on
`the ubiquitous Post-It, are lightweight and thus easily
`portable. As a result, they are often used for tasks like
`grocery shopping, which require
`leaving
`the “base”.
`However, lists present problems at reference time. For
`example, subjects said that because lists are small, they are
`easy to lose (for instance, the note in Figure 3 was written
`on a 3” x 3” Post-It). Limited size also limits the number of
`items that can be recorded. As lists expand, people resort to
`crowding in new items however they can, typically writing
`later items smaller and between previous items. And as
`subjects cross out completed items, the list becomes
`messier; unfulfilled items may go unnoticed.
` When
`circumstances merit, subjects recopy partially completed
`lists onto new pieces of paper.
`
`
`Figure 3: One subject's to-do list, which illustrates: (a)
`paper lists are typically small: a 3”x3” Post-It is easily
`lost; (b) lists evolve: some items are crossed out,
`different inks indicate intermittent updates; and (c)
`paper has limits as a list-making technology: items are
`not aligned, some words are very small, there’s no room
`to add more items (although that didn’t stop the subject
`from trying!). There’s no way to reorder items.
`
`Paper lists can only be in one place at a time, and subjects
`said this is frequently a problem. One characteristic
`participant said: “it is difficult to anticipate when I will
`need my shopping list.” Shopping trips often are unplanned
`and opportunistic (“I’m out for lunch, so I’ll drop by the
`
`pharmacy on the way back to work”). If she does not have
`her list, then she either has to stop for what might turn out
`to be an inefficient shopping expedition (“oops, I forgot to
`get cold medicine!”) or else miss the opportunity and be
`forced to go out at a different time when the list is at hand.
`
`Figure 3 depicts a study subject’s list, and exemplifies
`several of the points we have made.
`Improving the list. We observe that lists are commonplace,
`yet they have a number of weaknesses . As developers, we
`saw this as an opportunity: a well-designed tool could
`reduce problems such as lost or forgotten lists, messy,
`disorganized or unreadable lists, and the need to re-copy
`copy partially completed lists. While it seems natural that
`an LBRs should support list management, in practice this
`had not been the case: in the Place-Its LBR field study,
`subjects did not create lists [16]. We show next that our
`results differ.
`Lists, Dos, and Get: PlaceMail Usage Analysis
`After the PlaceMail study was complete, we assigned each
`of the subjects’ 344 reminders messages to one of ten
`categories1. We began with the classification scheme used
`in the Place-Its study [16], but we extended it to fit our data.
`This was necessary because Place-Its users didn’t create
`lists, but PlaceMail users did.
`29% of all PlaceMail messages were lists, which we
`categorized into three types. First, shopping lists contain
`two or more items the subject wanted to get from a single
`place. Second, to-do lists record two or more tasks the
`subject wanted to perform at a single place. Third, multi-
`place lists contain tasks for more than one place.
`Moreover, another 23% of messages were reminders to get
`a single item from a place, and 26% of messages were
`reminders to do a single task at a place. In effect, these are
`single-item shopping or to-do lists. Thus, 78% of all
`messages were reminders to get one or more items or do
`one or more tasks.
`Table 1 summarizes and illustrates the most frequent
`message types. Five other categories accounted for the
`remaining 22% of messages; none of these categories
`accounted for more than 6% of messages.
`
`Our results contrast markedly with those of Sohn et al. [16],
`(Place-Its), the only comparable study. As we mentioned
`earlier, Place-Its users didn’t create lists at all, and get and
`do accounted for only 30% of messages. We think this
`difference is due largely to our web interface: over 86% of
`all lists were created with it (and of the remainder, 6% with
`the voice interface and 8% with the phone text entry
`interface). Place-Its provided only a cell phone text entry
`interface. This likely explains why Place-Its subjects didn’t
`
`
`1 Two independent coders classified each message. The
`inter-rater reliability was 86% (91% for lists).
`
`Google, Exhibit 1007
`IPR2022-00742
`Page 5 of 10
`
`

`

`
`
`make lists: the input method was tedious for longer
`messages.
`
`Proportion
`26%
`23%
`8%
`
`16%
`
`5%
`
`Shopping list
`
`To- do list
`
`Message Type Example
`Do
`“Use 40% coupon”
`Get
`“Buy raisin bran”
`Multi-place list
`“Eat at Punch before stopping at
`B&N for Harry Potter”
`“Shampoo, body wash, dryer
`sheets, downy ”
`“Cash check, extra envelopes,
`submit XLE form”
`Table 1: List, Do, and Get messages in PlaceMail
`TASK DETAILS
`Next, in initia l interviews, study subjects explained they
`commonly record task details on paper for later reference.
`For example, one participant showed us a notepad with this
`information on it:
`7/11/05 check # 7658876098, New York, NY $138.00 (cid:224)
`check court order to see if I need to reimburse.
`
`(Actual values have been changed to preserve privacy).
`The subject explained that she needed to research child
`support issues and recorded background information about
`the task on paper. This type of cognitive offloading was
`common among study participants.
`Task Details on Paper: Benefits and Drawbacks.
`According to study subjects, paper generally is a popular
`medium for recording everyday tasks for several reasons. It
`is cheap, universally available, fast, and easy to use. In
`addition, paper is lightweight and easily portable. However,
`task details written on paper suffer from the same drawback
`as paper lists: the reference is easily misplaced.
`
`Task Details in PlaceMail. 37% of the messages subjects
`created during the PlaceMail study contain specific task
`details. For example, one subject left the message, “pick up
`21 ½ inch Weber grill grate” for a hardware store. This
`behavior marks an advance: Place-Its users did not record
`task details [16]. We believe the PlaceMail web and voice
`interfaces again explain the difference. Subjects recorded
`89% of task details with the web interface, 5% with the
`voice input interface, and 6% were texted directly on the
`cell phone.
`DAY PLANNERS AND CALENDARS
`Prior to using PlaceMail, five participants showed us how
`they employ (paper-based) day planners for everyday task
`management. First, they demonstrated how they frequently
`add place-based information and other artifacts to their
`planners. For example, one participant puts post-it notes
`containing “to do lists” in her planner. She attaches the note
`to the day when she wants to complete the list, and removes
`the post-it when all of the listed tasks are fulfilled. She
`explains it is easy to move the post-it to another day if the
`tasks aren’t completed as planned. Frequently, the lists
`contain one or more place-based errands. Another subject
`stores long-term reference information in his planner. For
`
`instance, he recorded the size of his home’s furnace filter
`and refers to the planner whenever he needs to buy a
`replacement.
`Subjects who routinely use day planners reported carrying
`them everywhere, including to personal appointments, in
`their car, and into the grocery story for reference while
`shopping.
`Day Planner Benefits and Drawbacks. Like other paper
`media, subjects find it easy to record information in their
`planners. And while lists are easily lost, subjects did not
`have this problem with their day planners. We conjecture
`this is largely because of their size. The day planner serves
`as a center for several types of information (calendars, lists,
`and long-term reference information), and users find benefit
`in this affordance.
`According to study subjects, day planners have two main
`drawbacks: first, they are bulky and thus sometimes
`awkward to carry. Second, while it is easy to add
`information to a day planner, sometimes the user forgets to
`refer to it at the opportune time. We will revisit this notion
`in an upcoming section on opportunistic reminding.
`Calendars. In the pre-study interview, 18 of the 20 subjects
`reported using a paper calendar at home to coordinate
`personal and family events. The calendar is kept in a
`stationary, prominent place in the home - this affords easy
`viewing and updates. The subjects do not typically bring
`their calendars on errands. One participant explained this is
`not necessary: if she unexpectedly needs the calendar while
`away, she follows up with a phone call or email after
`arriving home.
`In addition, four subjects use the calendar feature on their
`PDA. While these subjects carried the PDA between home
`and work, they did not usually bring it on errands. This
`behavior is consistent with the paper calendar: subjects
`primarily update and refer to it at a base.
`
`Calendar, PDA Benefits and Drawbacks. Participants
`said the paper home-based calendar is easy to update,
`difficult to lose, and is often in their line of sight so they are
`constantly reminded of upcoming events. Subjects did not
`report any major drawbacks with these calendars.
`Subjects also find the PDA calendar beneficial. Those who
`use it avoid direct text entry by updating appointments on a
`desktop computer and synchronizing it with the PDA. It is
`interesting to note that many subjects understood the
`relationship between the PlaceMail web interface (where
`they created messages) and the cell phone client (where
`messages were delivered) as a
`similar
`form of
`synchronization. They observed that both methods provide
`the same affordance: they can avoid text entry on the
`mobile device if they choose.
`
`Last, we asked subjects i