Flurish

Gimlet: Now with the Power to Forget

The Internet never forgets. Starting now, Gimlet can.

In Gimlet, we’ve always encouraged our clients to keep their data free from sensitive or identifying information. But from time to time, we’ve noticed personal data slip in to the database. And once it’s in, finding it and cleaning it out by hand is nearly impossible. So! To help you keep your (and your patrons’) digital lives free of radioactive sludge, we’ve added a new feature: the Gimlet Privacy Guard.

With this feature, you can choose an amount of time (between 30 and 365 days), and after that time, we’ll automatically clear the Question and Answer fields from your entries. After 28 days, our backups are automatically purged, and that radioactive data will be gone. Completely.

Other fields (your stats categories, tags, and timestamps) won’t be touched — so you’ll still be able to see reporting trends dating back to the dawn of time. The main thing you’ll notice is that, since the question and answer text has been cleared, the search feature won’t find these questions anymore.

A bit more background

As we go through our online lives, we leave behind a trail of data about our personal lives. In our Gmail inboxes. Our Amazon purchase history. Our Facebook profiles. The catalogs of the libraries we visit.

Tech companies are terrified of losing this data, because losing customers’ data is bad for business — but also because this data is valuable. Data lets companies train machine learning programs, so they can tell which emails are spam, make better guesses at what you want to buy, and know that you always want to read the status updates from your secret Facebook crush.1 And a large segment of the tech industry is built around selling data about your behavior to the highest bidder. (We here at Gimlet will never, ever do this. Your data is yours.)

All of this data piles up endlessly. One person aptly compared this data to the radioactive waste generated by nuclear power plants. As long as it stays safely contained and no one uses it for evil, it’s safe enough. At some point, however, private data is going to escape its confines and bad things will ensue.

For example: Did you know all of the data in Gimlet is available to the United States government, without a warrant, through a National Security Letter? If we were to receive one of these letters, we wouldn’t even be able to tell anyone except our lawyer. These days, trusting your online data will stay private means trusting the United States government — whoever may be in charge.

We take the custody of your data very seriously, and we know that giving clients this option means someone might accidentally delete something important. At the same time, part of taking care of your data means giving you the power to clean up the sensitive patron data that might slip into your database. Every online service provider should give you this ability. It’s the right thing to do.

See the documentation for more details on how to turn your Data Cleaner on and add a bit of ice from the river Lethe to your Gimlet.

If you have any questions or comments, let us know.

  1. Yes, Facebook knows.

April 18th, 2017  |  Published in Uncategorized

Flurish

Gimlet v2 – Development Update

Gimlet #2

Hi everyone,

For the better part of a year, we’ve been hard at work building the next version of Gimlet. Internally, the project has a lot of nicknames — “G2” or “#2” or “Version 2” or “the next Gimlet” — and it is pretty much all I work on outside of customer service and support these days.

Last spring at the Public Library Association conference and last summer on this blog we shared some demonstration screenshots of G2 and we were extremely excited about a design prototype we had created. Since that time we’ve worked to write the actual application code that will power our new design. We are quite a ways toward that goal now, with the hope we can begin testing an early-release of the application with a few select clients very soon (more below).

My ten favorite things Gimlet #2 actually can do today, that Gimlet #1 cannot:

1) Hide the free-hand, tag-entry form
2) Support bulk tag-editing
3) Edit text using Markdown
4) Perform faceted searches
5) Run a date comparison report
6) Download embedded charts as images
7) Perform a three-level, pivot-table report
8) Mobile interface
9) Separate time entry from date entry
10) Highlight search result text matches

SO! Our next steps for Gimlet #2 are to lock down a preview branch of the project and start collecting feedback from current clients. That step will happen later this month (April 2017).

We have already identified a few clients that are interested in helping with this next stage of the process, and we’re extremely grateful for their offer to help. If your library uses Gimlet #1 and you’d like to help us complete Gimlet #2 by previewing your data in the new interface, and providing us feedback, please contact us. We unfortunately don’t have the capacity to offer every current client a preview, but we’ll try hard to help the first handful of clients that reach out.

We’ll be back in touch with additional Gimlet v2 details each week this month. Stay tuned.

Cheers!
– Eric for the Gimlet Team

April 13th, 2017  |  Published in Gimlet #2

Flurish

New Report Option: Promoted Tags

When we launched Gimlet, our tagging option was completely free-form. “Power to the people!” we thought, “Let the best cataloging scheme arise spontaneously from each individual’s work!”

Chaos, of course, ensued.

As a salve for the unfettered free-tagging going on in everyone’s Gimlet accounts, we introduced “Promoted” tags: Gimlet administrators can define some “approved” tags, and people using Gimlet are strongly encouraged to stick to that list.

And thus, the chaos decreased.

The problem, though, was: What can people do about the past chaos — especially in reporting data? Even after setting up your list of promoted tags, reports would still show the effects of the previous free-wheeling days. And we really wanted people to be able to leave those days behind them without the work of cleaning up all the original data.

So! As of now, you can limit your tags in the Detail report to only show the promoted tags. Select the “Tags (promoted only)” option for either Rows or Columns and you’re set. It’ll change your reports from looking like this:

Uncontrolled tags screenshot

to this:

Promoted tags screenshot

September 13th, 2016  |  Published in Uncategorized

Flurish

Gimlet – Improvements Coming!

Today we’re excited to announce a series of feature improvements coming to Gimlet. I’m thrilled about the future of Gimlet and you should be, too.

We’ve spent a great deal of time analyzing customer service requests and we’ve talked directly with several clients to build a solid vision for how we can add the most value back to the software.  This post is going to provide an overview of the work we’re undertaking, with more posts to follow providing detailed information on each of these major new features.

Okay! Here’s the list of coming improvements:

Improved Search

One of our principal design goals in this round of updates is to give you better access to your data in Gimlet.  To that end, we’re adding a faceting search engine to allow you complete control over how you query your data.

Features

  • Facets for limiting your results to particular category values
  • Full support for complex, nested Boolean queries
  • Improved phrase and proximity searching

Search Results


Improved Reports

Our new search engine will also power our next generation of reporting tools.  Doubling down on this new technology will allow you to query and facet directly within reports, providing you all the means to analyze your data in great detail with rich, ad-hoc control entirely at your command.

Features

  • New downloadable charts and graphs
  • Compare data between date ranges
  • Facet or query directly within reports

Comparison Report


Awesome Mobile Support

We know a lot of Gimlet use occurs at a traditional service desk, using a standard desktop computer. We also know your staff use Gimlet throughout your library on a wide variety of devices.  We’ve designed the next version of Gimlet to be mobile-first, and fully responsive, to create a user experience that’s natural and optimized for any device.

Features

  • Use Gimlet on any device, anywhere.
  • All the power desktop Gimlet on your phone
  • Speech-to-text entry on mobile devices
Mobile question form

Mobile question form


Speed Entry

New clients are always worried about how much time it takes to use Gimlet.  It’s dreadful for them to imagine typing each question verbatim into the question form. PLEASE DON’T.

It’s always been our recommendation that people only write question and answer text into Gimlet if it’s worth searching for again in the future.  Many times (perhaps most even?) it’s not worth adding that text, but you still need to record the transaction.  Enter speed entry.

Here’s how it’ll work:

  1. Choose a question type
  2. Add tags (optional)
  3. Save! 🙂
Speed entry

Speed Entry – Choose a question type, optionally add tags.

That’s as quick as entering data can be.  You’ll be able to switch between Traditional Entry and Speed Entry on-the-fly, so select whichever makes the most sense for you at the time of need.


Okok, that all sounds wonderful! When are these updates coming?

That’s a really good question.  The new search engine is a large and important update for us.  We’ll be moving that tool into production first and before we release any other changes. In fact, when the new search engine upgrade occurs you won’t even see anything different in Gimlet.

All of the user-interface improvements described here are next. These upgrades are significant, so we’re stilling working on the details for how we’ll release them. More information will follow in the coming weeks, but for now, we felt it was high-time we shared a glimpse into the future of Gimlet.

Do you have any thoughts, concerns, or feedback? Please drop a comment below! Thanks.

July 19th, 2016  |  Published in Gimlet #2  |  2 Comments

Flurish

Top 10 tags used in Gimlet

  1. printing
  2. holds
  3. books
  4. dvds
  5. computer_help
  6. libcard
  7. renewals
  8. checkout
  9. book
  10. supplies

Okok, this isn’t very interesting, but it’s pretty true of any reference shift I ever had.

January 20th, 2016  |  Published in Uncategorized

Flurish

2025

I just set a reminder in my calendar for 2025. Ten years in the future. My oldest daughter just started kindergarten; she’ll be learning to drive when Google emails me about this.

Recently, we took the plunge, spent the money, and registered a trademark for Gimlet. (In retrospect, we probably should have done this sooner, but no harm has come from waiting, either.) With our fancy trademark certificate came a reminder that our registration needs to be renewed after ten years. Into the calendar it went.

It’s a little strange to think that whoever wins the American presidency in 2016, we’re planning to still be around when they’re long out of office.

September 24th, 2015  |  Published in Uncategorized

Flurish

Fifteen Million

Just a little shy of our sixth birthday, our database has crossed the “fifteen million questions” mark. 15,000,000. That’s a 15 followed by six zeros. A bit more than the population of Istanbul. At the rate we’re going, we’ll pass the population of Shanghai in less than two years. It’s a big number and it keeps getting bigger.

I thought I would have something big — something profound — to say about this, and how it feels, but I don’t. More than anything, it’s a reminder that libraries are vital, living places with librarians helping thousands and thousands of people every day.

Here’s to another fifteen million.

Also! As a small reminder, next week we’re headed out to ALA. We’re set up at booth 3243. We’d love to see you.

June 18th, 2015  |  Published in Gimlet

Flurish

Come see us at ALA!

That’s right: After missing out on Las Vegas, we’re headed to San Francisco for ALA 2015. Come visit us at booth 3243.

If you’d like to set up a time to talk in a more in-depth way about how you’re using Gimlet and how we could make it work better for you, send a note to support@gimlet.us and we’ll set up a time with you.

May 14th, 2015  |  Published in Gimlet

Flurish

Running a Startup: Myths and Realities

It’s hard to get straight talk on what it’s like to run an independent software shop. The media focuses on extremes: Facebook buys Instagram for $1 billion. Snapchat raises $468 million in venture capital. Everpix goes out of business despite raising $2.3 million and having $254,000 in annual revenue. And that’s just photo apps. It looks like there are two paths to walk: one of immeasurable wealth and one of bankruptcy and failure. Both paths are paved with endless 80-hour weeks, burnout, and broken families.

Part of the reason it’s hard to see a different path is that small software shops don’t talk about money very often. Money is a often taboo conversation topic in the United States (especially the Midwest), and it’s easy to sound like a jerk — either boastful or whiny. As a business, talking about revenue might make people think we’re taking advantage of our clients and living too well, or we’re not making enough money and are unsustainable. Or give competitors some kind of advantage. It’s all downside, no upside. Right?

But prospective business owners would be well-served with better information. And over the last few months, the developers of some really fantastic apps have shared their experiences… and revenue data. Some of these are quite positive and some are downright depressing. In one case, it’s simply beautiful.

We’re a different sort of business to the ones linked above: Our first product, Gimlet, is aimed at solving a specific set of problems in a limited niche. In particular, it’s designed to help libraries quantify the value of their public services (for example, reference) for their sponsors (for example, university deans and city councils). Our target market is smaller, and we’re selling something designed to provide business value rather than pleasure. Though our experiences aren’t directly comparable, I hope our perspective is useful, too. So: Here are some perceptions of running a startup, and our experiences:

Myth: You need to work a million hours right away.

Gimlet started as a side project for Eric and I. We both had (and still have!) jobs with universities that provide stable income and health insurance. Looking around, having an alternate source of income is incredibly common for independent creators. As Gimlet has grown, we’ve been able to switch from full-time to part-time day jobs. This hasn’t been trivial; it’s taken negotiation and influenced our career choices. And it’s meant spending some nights and weekends and vacation time working on software, rather than doing something more fun. And, honestly, taking a lot longer to launch than we hoped.

But, having stable income has meant we can work a sustainable number of hours on Gimlet and still spend time with our wives and kids. We built and launched Gimlet on our schedule, while we gradually learned the ropes of running a business.

Myth: You need fast, exponential growth to survive.

We’re a subscription-based service, charging based on the number of library branches using our software. Here’s our growth curve:

Gimlet: Billed Branches Over Time

It’s pretty much a straight line. It has its fast and slow times, but in general, it’s been pretty steady and predictable. This growth rate is something we can sustain and plan around; it’s a lot easier to make decisions about our steady jobs when we have a fair idea what the year’s income will be.

One thing that’s helped us thrive through our slow growth has been keeping expenses low. Not collecting much money from a business is one thing; continuing to pour money in would be hard. About half of our total expenses are in server hosting, which is $160 per month. And we launched on a server that cost us $20 per month. We were breaking even within months of launch, and made back our initial investment within a year. We’ve never taken on debt or looked for venture capital.

Myth: You need to be perfect right away.

When we launched Gimlet, there was a lot of stuff we didn’t know. How much time would tech support take? How about billing? What bugs hadn’t we found? Would they be big deals? And the biggest: How much should we charge? Because we didn’t know the answers, it was pretty unlikely that we would do everything perfectly even with infinite time, which no one has.

So: We launched small, and decided to learn on the way and be flexible. As it turns out, tech support takes about as much time as we imagined. Billing support takes way more. And because of that, we realized weren’t charging enough for Gimlet. So we decided to raise our prices, leading to the jump in our annual subscription rate:

gimlet_subscription_time

Was raising our prices a hard decision? You bet. Does the added time give us more flexibility to to work on Gimlet while keeping on top of billing? Absolutely. Do we have a ton ideas about how to make Gimlet better? Of course. Do we have a better idea of where to direct our work now that we’ve been in business for a while? Totally.

Final Thoughts

We couldn’t be happier about how building and running Gimlet has turned out so far. We aren’t going revenue gangbusters, but we’re steadily climbing towards decent full-time developer salaries while building a product we love for clients we love. Even more importantly, we’re helping keep our families fed and healthy and loved while we do it.

If you’re an aspiring developer, I can’t guarantee this path will work for you. It’s worked great for us, though.

If this isn’t success, I don’t know what is.

February 9th, 2015  |  Published in business, Gimlet

Flurish

Our New Server, and Internet Services

Last Sunday, we migrated Gimlet to a big(!) shiny new server. All of the associated hiccups seem to have been solved. We always think the new server is so huge, and then outgrow it surprisingly fast.

In this migration, we moved from our own dedicated server at Rimuhosting to a cloud host at Digital Ocean. We were very happy with Rimuhosting’s service, but the flexibility of the cloud servers is a huge advantage for us. It’ll be a lot easier for us to grow here. The price is rather lower, too, which doesn’t hurt.

Doing the migration was a bunch of work, but since the service is a better fit for us, it was worth the time spent.

Of course, as a service provider ourselves, this is a somewhat sobering realization — but a valuable one to keep in our minds.

December 12th, 2014  |  Published in Gimlet