We made this.
We’re proud of it.
Remembering ClipCard as a product and an idea
ClipCard is shutting down today. We’ve reached the end of our capital and are taking our products offline.
Before we turn out the lights and head to the bar, we want to leave behind an artifact of what we’ve made, how it worked, and some thoughts on why it matters. Today may be the last day for this project, but we believe the broader ideas we’ve been working on have a bright future. We hope our effort is a useful contribution to those who will carry that future forward.
What was ClipCard?
We built ClipCard for people like us — those who rely on dozens of apps and cloud services to do their jobs, and are constantly jumping between browser tabs and screens and folders just to find what they need. Is that file in Box or Dropbox? Are the notes in Evernote or a Google Doc? Is that contact in Office or Gmail? The JIRA task that got assigned to you this morning… is it related to this Trello card or that GitHub issue?
The fastest way to get anything from across your apps,
in context, everywhere you’re working.
We set out to make a more seamless experience across those silos. Our goal: be the fastest way to get anything from across your apps, in context, everywhere you’re working.
Users connected their apps to ClipCard to bring together notes, issues, contacts, files and more. They could see the latest from all their accounts as things updated and narrow in on a topic, place or time to get a useful summary and act on it quickly.
Here’s what it was like to use:
How did it work?
In little more than a year, we connected to a dozen major cloud services at scale and mapped their diverse content types into a single normalized index. We shipped three major products which made that content quickly findable and actionable for people on the web, on Mac, and inside Slack. We grew thousands of users from around the world and indexed well north of 40 million items on their behalf. It was a modest start to an ambitious project.
Anything from across your apps
Connecting your apps to ClipCard took just a few easy taps. With your permission, we’d summarize the full contents of your accounts — everything you’ve saved in Dropbox or added to Evernote or shared in Google Drive — without ever having to know your password. Our commitment to user privacy was strong and centered on a simple mantra: your stuff is yours.
Under the surface was a distributed system that processed streams of data from a range of APIs at scale. ClipCard indexed tens of millions of records while playing nicely with the rate limits of our partners, gracefully handling errors and idiosyncrasies as they arose.
That system existed for a reason: to summarize the stuff from your connected apps. Every contact, note, file, issue, and task was concisely displayed in a familiar format — a ClipCard. We designed a normalized index for describing diverse types of content, mapping the different organization schemas of file systems, contact managers, and productivity tools into a common tagging structure so you could quickly and easily find related concepts across apps and information types.
We made this index available via a tidy RESTful interface to a range of end-user products. This was effectively a single API on top of many disparate APIs. It was engineered to scale, not only to accommodate the additional users, sources, and apps that our team would build, but ultimately to support a community of developers contributing to and benefiting from the value of quick context and access across silos.
In context
Context had a few critical dimensions for our users. The first, and most important, was time. What’s happening right now across all my apps? Our phones and email accounts buzz all day long with a cacophony of notifications from each individual app and service, but there are desperately few tools that give us a higher-level view of what’s really going on with the projects and topics we care about.
ClipCard was that view. When you first logged in, you saw all the most recently updated things from across your apps and could quickly narrow in on any one item.
Another critical dimension of context is place. What’s happening near me? Who did I meet at that conference in San Fransisco last week? Where are the photos I took while on vacation in Buenos Aires last year? ClipCard would extract location data wherever possible — from notes in Evernote, photos in Dropbox, contacts in Google and Office 365, etc — and give users an easy interface for finding them on a map.
Most importantly, context is determined by relevance. What’s going on with this project? Where are all the design assets and notes I need for this feature? What do I know about this client? The heart of ClipCard was a search engine that would help you quickly narrow in on concepts, tags, and specific items regardless of where they were stored.
Everywhere you’re working
ClipCard didn’t aspire to be yet another app. Our goal was to get as close as possible to the places you were already working, removing the interface barriers between you and your stuff.
Our first product was a responsive web app that scaled gracefully from narrow touch screens to wide cinema displays. This meant you could jump from phone to tablet to laptop, and still have a single way to get to anything across your apps.
Shortly after the release of our web app, we introduced ClipCard for Mac. It added a critical feature: a hotkey that you could fire any time to pop up a quick search across your apps. It was like Spotlight or Alfred for all of your cloud accounts. Writing an email, and need to reference a file on Dropbox or a task on Trello? Just pop up a quick ClipCard search and grab the link.
It was around this time that Slack was taking off as an unstoppable force of venture capital and workplace productivity, and many of our early users asked for a way to integrate ClipCard with their new favorite team communication tool. Search in Slack is powerful, but only reaches items that have already been directly shared with Slack. What if you could get anything from your connected apps, even if it hadn’t already been shared, without leaving Slack?
We released a smart bot that brought most of the same functionality of our web and Mac products into a friendly text chat right inside Slack. You could message @clipcard directly to get useful summaries and deep source links to your stuff without leaving your chat experience. This gave users more comprehensive access across their connected apps than Slack’s built-in search tool, and was one of the first Slack bots to establish private authenticated connections to third-party sources for individual users.
ClipCard for Slack was also an important demonstration of the real product and vision behind ClipCard — the API and normalized index that made it possible to easily deliver access across sources and content types into a range of experiences.
Why does it matter?
We weren’t able to fully realize the vision behind ClipCard. But that vision is still important. Seamless context and connection across apps is something we all need. The more services we rely on, and the more choices that are available, the more jarring and unproductive those context switches will inevitably become for everyone. There has to be a better way.
Seamless context and connection across apps is something we all need.
We’re not the first team to pursue this problem, and we aren’t likely to be the last. Back in 2011, writing about one of the earliest efforts at cross-app search, TechCrunch co-founder Michael Arrington wondered why the major players hadn’t stepped in already. He posited that the conflicts of interest between leaders like Google and Facebook would prevent any one of them from creating genuinely useful connections across their increasingly competitive offerings:
This may be one service that just has to be done by a small startup that can play off all the politics involved. And once they’re big enough, it’ll be too late to stop them.
The startup he was writing about went on to be acquired by Apple. Fast forward several years later, however, and we’re still waiting for a genuinely connected and contextual experience that brings together our own content across devices and interfaces.
Our team still believes it’s possible, and more importantly, that it’s necessary. We look forward to the progress and success of those who will continue this work.
If we can ever help, let us know.
Who made ClipCard?
ClipCard was made in Seattle. Tim Kearns (CEO) and Byron Nutley (CFO) led our team and investors through a difficult pivot and toward an ambitious agenda, providing guidance and runway throughout. Ryan Cunningham (yours truly) served as the product lead and was responsible for our visual design and communication efforts.
Trevor Schmidt, Ben Reilly, Matt Skone and Cameron Stillion designed and built the core API, index, and connector infrastructure that made ClipCard possible. They used that infrastructure to connect to twelve major APIs in production (and several more in prototype or development), each one presenting its own unique challenges.
Matt Menzer, Trevor, and myself collaborated on the product design of our web and Mac apps, and Mr. Menzer took the lead on the front-end development that brought those products to life for users. Mr. Skone took the helm on our Slack bot project, re-inventing those same interfaces in text.
Scott McGhee and Ryan Small took on the intimidating task of deploying and continuously integrating the major distributed system that ClipCard became. Mr. Small and Cory Paul ran the extensive automated and manual testing effort that maintained quality throughout the stack and across dozens of end-user experiences. Joshua Ruybal assembled a telemetry system which analyzed gigabytes of daily logs and provided valuable insight not only into the health of our system but also the nature of cross-app usage among our users.
We owe credit and gratitude to many more people, including our investors, partners, former colleagues, community members and families. And, of course, our users. Thank you all for your support, feedback and trust along this journey.
