Context
1Password currently has a CSV importer tool, but it has significant limitations in functionality. Users who were looking to move from another provider to 1Password have trouble importing their data. Given that various password management services employ different methods for exporting and organizing data, users invest a substantial amount of time organizing their CSV file to match with 1Password's import formatting. However, even after completing the import process, users may find that not all of their data transfers over or that specific items are missing.
If we can redesign the importer tool to address the majority of there edge cases, we have the potential to reduce hundreds of support tickets per month (all of which are related to importing passwords) and offer customer a better user experience.
Involvement
I was part of the admin console team that was tasked with reducing the friction that our customers faced during the importing process. As the lead product designer, I was responsible for spearheading major aspects of the importing data flow, from researching and planning to brainstorming and designing the MVP product, Alpha, Beta, and General Availability (GA) stages.
Throughout the development process, I consistently worked to enhance the import functionality in the admin console application and user experience, ensuring that users can seamlessly move their information and passwords over to 1Password.
Goals
Reduce number of customer support tickets for data not importing or CSV file imported with incomplete data.
Increase the reliability in using the 1Password importer tool to migrate data without being dependent on third-party integrations or non-verified importing solutions.
Increase confidence from (1–10) in using the importer tool instead of contacting customer support.
Core Responsibilities
User Experience Design (UX), User Research, User Flows, Prototyping, User Personas, Competitive Analysis, and Usability Testing, Responsive Design
Team
Product Designer (Me), Product Manager, UX Researcher, Engineers, and Content Designer
Challenge
1Password is missing an importer tool that has the capabilities for large-scale migration and did not offer customers the ability to seamlessly import their passwords from other apps or major contending password management services.
Our assumption was to improve the user experience of the importer tool, we needed reduce friction during the importing process. For customers importing from other apps, we also have to integrate the app's documentation into our importer flow that is preloaded with the steps needed to export their data from their provider and import into 1Password.
01
02
Analyzing support tickets and customer analytics
Together with customer support we identified main customer problems
To find out why so many customers were struggling with the importer tool, I collaborated with the Customer Support Team to look at support tickets from customers and past video recordings from older research archives. I took notes of immediate user experience problems that stuck out to me and prioritized them based on it's potential business impact.
I have also conducted 8 qualitative user research sessions with customers to gain a deep understanding of their pain points and challenges with importing passwords, as well as identify opportunities for improving the importer experience. These sessions involved B2C End-Users, B2B End-Users, and Dev Admins from diverse industries. Their feedback have revealed common themes that helped me form actionable insights to enhancing their overall experience.
Below are the key insights and feedback gathered from those users:
Low-discoverability
Lack of information architecture optimization causes confusion for customers to find the importer tool.
Lack of importing documentation
Customers want a guided importing process that show them how to export their passwords from the password management app or browser they use.
Poor instructions on transferring data between applications
Instructions are not clear when it comes to exporting data from their current provider and importing into 1Password.
Importer complexity problem
Customers who were unfamiliar with using a password management service didn't understand the concept of a CSV file.
Unclear what source is supported
Customers want to know how to import their data if they don’t see their password management service in the import options list.
CSV template was undiscoverable or not available
Customers spend a lot of time and labour drafting a CSV file together while having trouble finding a standardized template from 1Password to work off from.
Communicate our findings through a Jobs-to-be-done (JTBD) framework
I worked closely with a user researcher to translate the customer feedback user and research findings into user personas. Through a series of brainstorming activities with stakeholders, we were able to come up with three user personas for: B2C end-users, B2B end-users, and Dev admin.
I was also able to connect the dots between design and user research. This made it easy for us to get alignment to design an importing experience that satisfy the user problems while prevented teammates from proposing solutions based on pure assumptions.
Ideation
Identifying entry and end points of the password importer experience
I created and proposed user flows that encompassed all of the potential importer solutions our team could build. Since the timeline to ship the importer tool was only 4 months, I worked with the head of engineer and the engineering team to cut back on features that cannot be accomplished in a reasonable timeframe to only implement the most important features. This allowed us to ship an MVP importing experience that we could onboard beta users with, validate the flow, and iterate quickly.
The first flow is for users who have had selected any password manager or browser option. Our importer can detect CSV format text data files that use a comma, a semicolon, a tab, space, or a custom character as the delimiter between values (the essential CSV data structure). Consequently, much of the laborious process can be automated which eliminates the intricate and menial importing processing steps that the second flow has.
The second flow is for users who will select “Other” as their previous password management provider. Our assumption was that users can feel overwhelmed and are afraid to make mistakes if they had choose this route. I wanted us to focus on building trust and help users feel confident with data validation without compromising the larger vision of the release.
Analyzing the importing experience of our competitors
To understand what customers are familiar with already when it comes to importing passwords, I analyzed other password management services to determine what our importer is doing well or where it failed, where we could improve if we are missing something that our competitors have, and to get an in-depth look at overall importing paradigms.
01
All password managers request users to select an import source (their existing password manager applications or browser) to start importing data.
02
All applications requires that the user will need to have exported data from their current providers application. A CSV file with their items and information is then downloaded onto their device where they would add the file into the importer and upload. Dashlane users can export a DASH file to securely move all their data from one Dashlane account to another. But also offers users the option to export data to a CSV file for transferring to other services.
03
All password managers provide users with on screen instructions to make it easier on behalf of the user to export data from their current providers application or browser. Most password managers requires the user to enter their master password in order to get their data exported. Some password management services try to lock you into their service by making it difficult for users to find the export option.
04
LastPass Teams or LastPass Business accounts have some restrictions with exporting data due to policies enabled by your LastPass admin. For example, private folders will be converted to tags when importing LastPass data into 1Password. On the other hand, shared folders can only be imported by a person who has administrator privileges in the folder.
05
Most password managers instruct users to define the item categories of their items while also providing them with a ‘Do not import’ option to ignore certain rows from import.
06
If a user has LastPass Authenticator, one-time passwords on certain items, it won't be imported. But the user can manually add the non-importable data after they’ve imported their other data. Similarly with Dashlane, specific items that can’t import are matched to a specific field in the ’note’ section as a form of secure notes.
07
After users completed importing their password into a new password manager, most password managers warn users to securely dispose of their recently exported file to protect the user and prevent their data from being compromised.
Exploring and optimizing for a better importing experience.
The MVP solution redesigned the existing importer flow, introduced the most requested importing features from users, configuring what items users want to import and where, integrated instructions that are unique to every password management service that users tend to export data from, improved discoverability of the import feature and more.
Import feature discoverability
Users can now easily discover where the importer tool is located from the Admin Console in ‘Settings’ or trigger the importing process in ‘Onboarding’. From our user research, we learned that some users would leave the experience altogether if the the importer tool was not immediately accessible to them upon completing the creation of their account. Bringing the importer tool to their primary dashboard view help nudge new customers to at least begin onboarding and start importing from there.
Users can alternatively access the importer tool directly from their ‘Settings’ at later stages of their onboarding journey.
Integrate instructions on how to export data
We recognize that being able to prepare and export data to be imported into our importer solution is paramount for managing user and organizational data. Users can view instructions on how to export data from their current provider and facilitate data export themselves without contacting support for help.
While the step does not have supported images right now, the on-screen instructions alone was a massive step in the right direction. This strategy provides the surface visibility needed for customers to make quick importing decisions all while preventing any technical barriers from getting in their way.
Showing customers how to prepare and structure their CSV file
At any stage of the importing process, users can find information about exporting and importing data from a CSV file from the ‘Help’ button in the import wizard. From our user research, some users struggled to figure out how to do basic importing tasks, such as exporting data from their current service or following a CSV template to prepare their file for import. Now we could build awareness on data preparation and teach customers how to use the importer more competently.
Validating the importing experience
I conducted 8 usability testing sessions with the same individuals that I had previously interviewed prior. For these sessions, the participants I had selected were admins and end-users who are currently using a password manager that isn’t 1Password. I also interviewed users who are using 1Password but haven’t used the importer tool before.
I observed how many customers selected the import source versus selecting the ‘Other’ option, if they needed more instructions on exporting data, and if they uploaded or (drag and drop) the correct file. Customers who were coming from other password management services tend to complete the import setup with ease. Whereas, customers who selected ‘Other’ as their previous password manager provider needed more instructions to import, so the upload rate was lower but the completion rate was very much the same.
I was also tracking if users can define whether if their imported items are logins, credit cards or secure note. All users quickly defined each items in their import and were confident in its definition to proceed to the next step. In the labelling step, most users struggled to add labels to each column because they were not prompted to. While users who ignored the step would find out later that the step was necessary to ensure that there data won’t become disorganized when imported.
01
Users were unconfident in their next steps when there are no contextual import guidance to help them keep track of their progress. Some users have a lot more logins and item types to import than other users, so it was important for them to have indication of how much of the importing process has been completed and how much is left.
02
Users were unaware that they had to label the data columns to ensure that their credentials are correctly imported into 1Password.
03
Some users had assumed that the labelling step was automatically mapped and that they did not need to worry about examining each column and its contents to ensure that the appropriate label was selected.
04
Showing the on-screen instructions on how to export data from their previous service in isolation from the ‘upload file’ step makes it harder for user to reference because they don’t remember all the steps to export their data. Some users found it overwhelming to navigate back to re-access the export instructions.
Iterating upon design critique and user feedback
Customers should know where they are in the importing process
Users were unconfident when they are deep in the importing process and are unable to determine how many steps are left until they successfully complete it. For some users, the path through the import process may be quick and simple, while for others it can be duplicative and requires a lot of mapping.
A progress indicator was implemented into the import process to help users keep track of their progress during each phase of data migration. We were able to see that users were more confident in their next steps when we use indicators because expectations were managed upfront and user have an increase in understand of how much time and effort the import process will take.
Combining instructions and uploading as a single step
Showing the exporting instructions separate from the ‘upload file’ step felt peculiarly disjointed to users. In the previous version, users would have to know exactly how to export their saved password from their providers. They didn’t particularly like having to navigate back to preview the instructions.
I combined the on-screen instructions to the ‘upload file’ step of the importing process to make it easier and faster for users to export their data. We saw slight increases in upload completion, and in the number of users that were able to complete this step and successfully importing their data over to our service.
Draw attention to the importance of data labeling
In previous versions of the import process, users were unaware that they had to examine each column and its content—and then label them to make the import file compatible for import. Some users had assumed that the CSV file that they’ve uploaded will be mapped automatically based on their defined item type.
The importer requires the CSV file header row to determine how to map data. To help users acknowledge that these columns are requirements and need labeling for accurate data import, the corresponding column headers are highlighted in a red accent to indicate unmapped columns and to prompt users to map them if it was ignored or skipped. When columns are completely labelled, the headers are highlighted in a green accent reassuring the user that they can move forward to complete import.
Prompt customers to take protecting their data seriously
While experienced users intuitively understood the risks of exporting a file containing their passwords, there were a few users who would often times forget to delete their CSV file after it has been imported into 1Password. A few users were also unaware of the best security practices when importing involves an unencrypted file.
There wasn’t anything particularly wrong with using informational callout as it did prompt users to take action. The previous messaging was somewhat unclear, which made it difficult for users to commit towards deleting their unencrypted file. These findings led me to collaborate with a content designer to revisit how we want to convey the message to users. Users need to feel supported and informed after completing the import process. We provided additional information and added more context around the importance of deleting the imported unencrypted file from their computer.
Final Design
The new importer tool was iterated on and launched receiving positive reception from users. The following designs were truly a collaborative effort between myself, the engineering team, and all stakeholders who were involved throughout the design process. Working together allowed us to design the end-to-end experience of the Importer Tool, including the MVP experience of several essential importing features only in a span of 4 months.
Impact
After the MVP was launched, we found that we were able to eliminate the user barriers with importing passwords and saw an accelerated increase in the number of user completing an import, but there is still room for improvements. Base on feedback from users, we continued to optimize existing importing features and consolidated a list post-MVP additions that we could conceptually experiment running an unmoderated usability test with users in future milestone.
Deobfuscating columns, import options that included spreadsheet capabilities, and a full list of item categories to support such as (API credentials, bank account, crypto wallet, and more) were added onto the roadmap.
Fewer support tickets related to importing data
Improvement in the discoverability of the feature
Increased confidence of using the importer tool
Customers incorporating the feature into their process
Retrospective
The importer tool has garnered overwhelmingly positive feedback. The successful implementation of this feature involved extensive collaboration among the engineering team and content designers. I have also learned a lot from collaborating with a user researcher to plan and conduct both moderated and unmoderated usability tests. It has been instrumental in my continuous growth and learning throughout my career.
As a relatively small team at the time, we were committed to launch scrappy solutions to get fast learnings and iterate quickly. We were continuously tracking data and interviewing customers to see which part of the import experiment were successful, and what failed. Having a clear vision of the final experience and a path to get there was vital to our success.
Learnings and next steps:
The importer can only import passwords and items from specific password managers or browsers. With this technical limitation, it was mandatory for users moving data to 1Password to create a CSV file containing only one item category and manually perform the import one at a time.
Optionally, a user can manually add a passport, drivers licenses, social security card, or any other items in the desktop or mobile app if the option of modifying entries of a CSV file or create one is hard for them if they are not technically savvy. In a future release of an updated importer, the team is exploring ways to support other item types (eg. bank account, medical records, drivers license, and more) and to make importing other item categories less labor intensive.
Most participants were able to complete importing their data. User had expressed that the in-app contextual guidance had helped bring focus to where they should start mapping and guided them to identify how to map and label the columns to match the specified item category type.
A small percentage of users still struggled to import data because their CSV file contained unsupported item categories and that it took long for them to import a large batch of data. There are still untapped opportunities to improve the importer to support importing a CSV file with custom field labels and improve the time it takes to import.
The importer does not check whether items in the CSV file to import already exist in your vault. If the user happens to import multiple files or import a file with items that already exists in their vault, it will create duplicates of that item. Users would have to go into the vault where the duplicate item was imported into to remove any entries they might not need.
Item deduplication was a feature that many users had requested to see implemented into the importer. This was not easily solvable because it is difficult to determine whether if an item is an accidental duplicate in the file that needs to be deleted, or if it is an intentional duplicate that needs to be preserved to import.