Monday, July 27, 2015

New Webclient Auto-Logout Feature

Big news!  You can now set a duration for your webclient login[1].  This way, if you forget to logout at the end of a shift, you won't leave your chat service falsely online for a long period of time.

The auto-logout option appears at the bottom of the "Choose Staffing Assignments" dialog that appears when you log into the webclient for staffing.



After the login duration is reached, you'll be alerted and given a chance to extend and/or modify the timeout.  If you don't respond within a minute (because you are actually gone for the day or maybe you've wandered away while helping a patron), the webclient will log you out.



What if you don't enable auto-logout and accidentally leave yourself signed in?


If you realize you've left yourself logged into the webclient on a computer you cannot physically access, you can opt yourself out of your queues. Sign into the webclient from another location and opt yourself out using the regular queue selection box. This will opt you out of your queues from all locations no matter what software client you are using. Phew! Then, logout as usual. You'll be able to opt yourself back in when it's time for your next shift.

An administrator can also opt users out of queues. In the users management page (US, CA, EU, SG), select the user in question and uncheck the boxes next to the queue assignments to temporarily unassign users as operators on a queue.

What if you use another software client to staff your LibraryH3lp service?  


If you staff your service through another XMPP (Jabber) client such as Pidgin or Adium, no worries. Locally installed desktop clients usually have great support for automatic setting of Away status based on keyboard/mouse idle time. So with these clients, you can prevent yourself from falsely holding the service online easily.

But if you happen to not have auto-away enabled, or if the trigger is a long way off, you can still use the admin dashboard or webclient to opt out of your queues. Just follow the steps in the section right above this one.

Concierge auto-replies


One related note. You can also have automated replies based on a lengthy wait time for your guests. This helps cover things not only if your service has been left online accidentally, but also in case your operators can't respond as quickly as usual for any other reason.

[1] The auto-logout feature is not yet available on our regional Canadian service.

Wednesday, June 10, 2015

Maximize Your Ask a Librarian Coverage

In the last post, we shared a checklist designed to whip your Ask a Librarian service into shape.  We included tips and strategies for maximizing and invigorating chat, texting, knowledge base, and email services.  In this post, we wanted to focus on personal interactions between librarians and patrons and share some ideas on how you can most strategically maximize Ask a Librarian coverage.

When you boil it down, expanding real-time Ask a Librarian coverage means people.  And once you start talking about adding people, everyone starts talking about budget.  The good news is that there are ways to maximize coverage without breaking the bank, and a little bit of ingenuity can go a long way.

Make sure the busiest hours are covered

Before spending a dime on more, why not make sure you are squeezing every last bit of benefit from what you have now? Sometimes you can assist more patrons just by experimenting with schedule.

Let me introduce you to the LibraryH3lp Chats per Hour report.  It contains everything you need to make the perfect Ask a Librarian staffing schedule.  Here's a sample Chats per Hour report using the LibraryH3lp support chat queue as an example:


The report covers each day of the week and shows which hours are busiest.  Either Pam or I are usually available on our customer support chat line from 8am to 5pm on weekdays (and, yes, sometimes later and on weekends).  But we'd still get pretty good bang for the buck by only covering 11am to 4pm Monday through Friday.  This makes sense, because our customers are library staff and primarily work on managing their LH3 services during regular business hours.

The situation with students and other patrons using library resources is a different matter, of course! With library patrons, you certainly have much longer useful service hours since libraries are all about research and education, and students of all sorts often work late into the night.

Also remember that this Chats Per Hour analysis can only show you traffic trends for when you service is actually online and open for business.  Well, then, how are you supposed to tell which hours you are NOT currently covering would be appreciated by your patrons?  We have you covered. It turns out that this data is pretty consistent.

We looked at some aggregate Chats Per Hour data for a few true 24/7 services on LibraryH3lp, corrected for time zone, and you can see the general trends in this chart:

Chats-per-hour for several academic libraries with 24/7 LH3 services.

This basic pattern is typical for per-hour chat activity when the service is available at all times.  You can look at this and see if you'd like to extend your hours beyond those that you are able to cover in-house.  For example, if you close up shop at 5pm, you will certainly make a lot of people happy by adding a few more hours and staying available until 8, 9, 10, 11pm, etc...  And Sunday coverage is a huge win.

To run the Chats per Hour report on your own service: just pick your server's region - US, CA, EU, SG. You might find that resources currently being used for chat are better applied at different times or even elsewhere. If you are already dealing with limited chat hours, try keeping your busiest hours staffed and experiment by reinvesting the lower-traffic hours at new time slots.

Hire backup staffing

Let's assume you want to extend your chat service hours, but your existing staff are stretched to the max already, and it's not worth it to hire dedicated after-hours staff just for your own library. What now? One excellent option is Chatstaff. You can be up and running quickly, plus the service quality is extremely high, with wait times currently averaging less than 15 seconds (yes, you read that right!). All of the Chatstaff librarians have MLS degrees.

And, about that budget: with Chatstaff, you can be brutally efficient by homing in on the specific hours of coverage you'd like to have.  So if you only want a few hours hours here and there, Chatstaff can do that.  Or if you wanted the whole 24/7 enchilada, Chatstaff can do that too.  And pricing is fair: those very low-traffic times in the wee hours of the morning are not priced the same as busier hours.  So you will probably be pleasantly surprised at your return on investment for substantially increasing your service hours.

If you'd like, Chatstaff can even be available as your secret weapon during the day, to cover those times when all of your own librarians are busy, in meetings, a whole ton of chats just came in at once, a lot of people are out sick, etc... You can email Chatstaff directly for a quote.

Find a partner

Collaboration is another great way to expand Ask a Librarian staffing. Partner libraries can work together to provide longer combined coverage than each could provide individually. LibraryH3lp makes collaboration easy on the budget since there is no additional software cost beyond your current subscription.  And all the tools you'd use for your in-house service are also available for the collaboration, including transfer between partner libraries.  If you are interested in learning a bit more, check out our earlier blog post on grassroots collaboration and our latest doc showing how to join separate LibraryH3lp accounts.








Thursday, May 28, 2015

Whip Your Ask a Librarian Service into Shape

Summer is almost upon us!  It's time for vacations and recharging your batteries. It's also a great opportunity to find some time to finally tackle that ever-growing "to do" list. To help with all those virtual service tweaks that you've been meaning to get to, we've created a checklist that is sure to recharge and maximize your virtual services.  Also available in PDF format.

Ask a Librarian Check List

  • checkboxIs help available anywhere a patron might need assistance?  
    • Action item: Check all the usual places: website, subject guides, catalogs, databases, mobile apps.
    • Action item: Search for web pages off the beaten path.  For example, the 404 (not found) page is a great spot to include a chat widget or FAQ link to help lost patrons.
  • checkboxCan patrons easily locate chat and FAQs on your web pages with just a glance?
    • Pro tip: Place advertisements for these 'above the fold.'  That is, the portion of a web page that is visible without scrolling.  Page headers or 'Click to chat' tabs are a super place for attractive chat images.  Aim for a consistent, highly visible place where patrons can always find help.
    • Action item: Catch the patron's eye with a modernized chat widget or a shiny new Ask a Librarian image.  Need inspiration? Check out these great examples from your peers: Ask a Librarian images, Ask a Librarian chat widgets
  • checkboxDo you proactively invite patrons to chat?  
    • Pro tip: Many LibraryH3lp clients offering timer-based chat invitations to patrons report significantly increased meaningful chat stats.  
    • Action item: Update your LibraryH3lp services to include a customizable proactive chat invitation. The average time among LibraryH3lp clients for the invitation's time trigger is 40 seconds.
  • checkboxDo staff consistently and quickly greet chat patrons?  
    • Pro tip: One way to evaluate this is to grab your chat metadata over a date range and then click the wait time column header to sort chats by wait time.
    • Action item: Create canned messages for operators to use when responding to patrons.
    • Action item: Set up one or more concierge auto-responders which trigger a simple greeting after a second or two, and/or send alternative contact information if patrons go unanswered for a lengthier period of time.
  • checkboxCan patrons get help after hours?
    • Action item: When chat is offline, point patrons to your searchable FAQ.  Need to import an existing FAQ into LibraryH3lp?  Email us and we'll make it happen.
    • Action item: Interested in extending your chat service hours, but your staff are stretched too thin? Consider incorporating backup staffing through Chatstaff.  Chatstaff provides responsive, high quality service, and you can specify your service hours to maximize the bang for your buck.

Text a Librarian Check List

  • checkboxDo patrons know about your Text a Librarian service?
    • Action item: Include your Text a Librarian number on your website and make it easy for mobile patrons to click.  Here's an example of some handy HTML that does the trick:
      <a href="tel:15555551234">Text us at 1-555-555-1234!</a>
    • Action item: Include your Text a Librarian number on all printed promotional materials.
    • Action item: Promote your Text a Librarian number during tours and instruction sessions.
  • checkboxIs your offline auto-response appropriate?
    • Pro tip: If you include a (shortened) URL, patrons on smartphones can click it to launch a browser.  Point patrons to your searchable FAQ or email contact form.
    • Action item: Pretend you are a patron and text your service when it is offline.  If the offline auto-response misses the mark, email us with a new auto-response 140 characters or fewer.

Email a Librarian Check List

  • checkboxIs email reference integrated with the rest of your virtual services? 
    • Action item: Activate the 3mail module to integrate email into the staffing workflow. 3mail is a simple email system that makes the LibraryH3lp webclient one stop shopping for answering chat and email questions. 
    • Action item: Train staff to use Tag for Follow-up.  When chats can't be answered right away and you need to get back to the patron, tag for follow-up automatically sends email to pre-designated addresses with transcript attached.

Want a printable version of this checklist?  Download the PDF version!

Tuesday, May 12, 2015

LibraryH3lp Concierge: Auto-responder for incoming chats

Looking for a consistent, friendly greeting for each new chat?  Canned messages are one way to go if an available operator provides a timely response.  Having a real person respond quickly is the best customer service.  But what if a new chat goes unanswered for an unusually long period of time?

Enter the new LibraryH3lp Concierge.  With the Concierge auto-responder, you can automatically provide customized replies with flexible time triggers.  For example, if a chat languishes for 45 seconds, perhaps it is time for apologetic message and an offer of alternative contact possibilities.

Concierge auto-responses can be set up on one, some, or all of your queues. Each queue can have unique concierge settings with auto-responses triggered after a customizable time delay.

In order to set up your concierge services, please email support with the following details:
  • The name of the queue or queues.
  • How long in minutes or seconds to wait before triggering the auto-response.
  • Exact wording of the auto-response.

What happens on the staff side?

Nothing changes from the staff's point of view.  And, the staff can still answer even after an auto-response that apologizes for a delay and offers alternative contact methods.

What about text messaging (SMS)?

While you can certainly have auto-replies for your text messaging queue, you may or may not want to provide them.  Text messages already have the offline auto-reply, which is a different kind of auto-response.  If you also wish to provide an automatic greeting, you will want to make sure the overall effect is coherent.

Does Concierge require use of the webclient on the part of the operator?

No. Concierge works regardless of the staffing software the operator is using.

Can we have more than one auto-reply on a single queue: one as an initial greeting and another as a fail-safe if no one is really there?

Yes!  However, in the interest of good customer service, it is good to be a bit cautious about what the first of the two messages says.  It is often better to have a human send an initial greeting, and this can be done quickly and easily using canned messages.  This way, your guests are rightfully informed that a person is there and paying attention to their message.  If you have an automated initial greeting, and no real person can actually respond quickly, that can leave a bad impression even if an apologetic auto-reply comes along later.

Monday, April 13, 2015

Updated Webclient: auto-busy, simpler workflow, and more

We've introduced some updates to the webclient chat workflow designed to simplify chat management, especially for operators of busy services.  The basic layout of the webclient is unchanged.  However, you'll find that the chat area better uses the available real estate, making it easier to juggle lots of simultaneous chats with minimal fuss.  Plus there are other goodies too, so let's dig in!


New goodies: tiled layout; highlighted chat; greyed chat indicating guest has left the conversation.



Use auto-busy to control incoming chats

You might already know about busy (do not disturb) status, which allows an operator to continue with all of their active chats but not receive any brand new ones.  But it can be difficult for an overwhelmed operator to remember to change their own status.  Enter auto-busy!  When an operator reaches the auto-busy threshold configured in their settings, the webclient will automatically change their status to busy until they close a chat and again are under their auto-busy threshold.

When an operator changes to Busy status, other operators available on a queue will continue to receive the new chats, or if no other operators are available, the queue itself will change to Busy until at least one operator is Available again.   The auto-busy default is 9, which is deliberately set high so that it won't catch anyone by surprise, but can be changed to any number the operator desires.  Only chats on queues count towards an operator's auto-busy threshold; private IMs from colleagues are not counted.  

Pinning replaced by auto-split

The old webclient chat layout supported two chat areas, a large space for a primary chat of interest and a right sidebar for pinning up to three additional chats.  Some of our users with busier services found this old layout limiting, so we went back to the drawing board and came up with a Hollywood Squares workflow that auto-partitions a single chat area into tiles.  

You can control the number of tiles that are auto-opened in the chat area via an auto-open setting.  Auto-open only controls the number of tiles that are automatically displayed and doesn't represent a hard limit on the number of tiles you can have open at any time.  You can always open additional tiles by clicking the desired contact in your contacts list.

Check out this screencast with Pam to see the auto-split feature in action...



Chats grey when the guest leaves the conversation

Chats already include textual notifications to the operator when a guest leaves the conversation.  But we've also added a new, more visual cue to the chat.  When the guest navigates away, the chat background will turn grey and should the guest return the background will return to its original color, signaling to the operator that they can send messages to the guest again.  This improvement to the webclient goes along with some recent improvements to the chat widget's ability to detect when a guest leaves the conversation.  

Highlighting a chat with the star icon

Operators can highlight chats of interest using a new star icon at the top right-hand corner of a chat.  The highlight feature has no functional impact on the chat, but instead serves as a visual cue to draw the operator's eye.  

'Press Enter to send' is now part of your settings

Previously, you could toggle between using the Enter key to send messages and using a Send button to send message directly within each chat and within the conference room area.  In order to dedicate more real estate to chatting, we've removed the per chat toggle and added it to the Settings area.  The default setting is to use the Enter key to send messages to the guest.

Wednesday, April 8, 2015

Updated widget

We recently released a few noteworthy privacy and usability updates for the LibraryH3lp guest's chat widget.  Let's take a look!

The Clear Chat History button appears automatically after the guest sends their first message. 
Obliviate!
The only visible change to the guest widget is a new "clear chat history" button, which appears after the guest sends their first message.  The text for the button is localized in languages for which we already have widget translations.

By clicking the button at any time, guests can wipe out the widget's displayed chat history.  This is especially handy when guests use shared computers.  Clicking the button does not remove the operator's view of the chat history, but it will tell the operator that the guest has left the conversation.  

Wait, what did you say before I clicked?
We've improved embedded widget usability by preserving the chat history when the user navigates away and then returns.  Embedded widgets are simple and highly visible, which make them a popular choice for web pages.  But sometimes guests accidentally navigate away from embedded widgets when clicking links on the containing web page.  Previously, the guest would lose their prior chat history even if they navigated back immediately.  Now, that chat history is preserved, and of course the guest can always clear it with the Clear Chat History button.

Elvis has left the building
Finally, we've included a few extra tricks within the updated widget to better detect when a guest leaves the chat.  If the guest navigates away, closes the page containing the widget, closes a pop-up widget, or uses the "clear chat history" button, the operator is notified consistently.  

Thursday, March 26, 2015

Introducing the ACLator: fine-grained control over all the things

We are pleased to announce the release of the Permissions and Access Control List Editor, which we've pet-named the ACLator.  Think "Ack-ell-a-tor" using an Arnold Schwarzenegger voice.

The ACLator places the power of fine-grained access control squarely in your hands.  By default, users have limited, read-only privileges on queues of which they are members - just the minimum for staffing.  From those humble beginnings, users are commonly elevated to full administrative access across an account or to administrative access within a subset of the account.  But it doesn't stop there!  You can grant or deny read and/or write privileges to most anything to anyone - and yes, even to LH3 users outside your account.  That last part has super-exciting implications and really deserves it's own blog post.  But we'll get to that in a bit.  And as always, there is no extra charge for any of this new goodness.

Groups & Folders -  The Building Blocks of Access Control


If you've ever used the admin dashboard to manage users or queues, then you've probably noticed folders.  Folders organize users, queues, conference rooms, canned messages, chat snippets, chat skins, and knowledge bases (FAQs) into a collection of resources.  Folders can also have child folders, allowing you to build hierarchies of resource collections.

A group is a collection of users with the same access privileges to a set of folders.  There are two special groups that every LibraryH3lp account has - Everyone and Administrators.  The Everyone group automatically includes every user within the LibraryH3lp account and provides the basic read-only privileges that make staffing possible.  The Administrators group contains all users that have full administrative access to the LibraryH3lp account.  From there, you can create your own groups with customized access to the resources within your LibraryH3lp-powered service.

As an administrator if you navigate to the ACLator (US, CA, EU, SG), you'll see a list of all groups for your account.  You can manage groups (add, rename, delete) using the button toolbar above the list.

Example list of groups in the ACLator.  The management toolbar is located above the list.

Managing Membership


When you click on a group from the list, you'll see a Members tab that lists the members of a group.  The default view is a flat list, but you can switch to a tree view that reveals the folder hierarchy.  Beside the view toolbar is a button that lets you add or remove members.

Members for an Administrators group.  The membership management buttons are located above the member list.

The ACLator makes granting users full administrative privileges easy.  Click on the Administrators group as shown in the screenshot above.  Then click the add/remove members button to make your changes.  Yep.  That's it.

Using the ACLator to Isolate Departments


How about extending your LibraryH3lp subscription to include other departments, but not giving the departmental administrator the full keys to your entire kingdom?   We call these users mini-admins.  Mini-admins can manage their own resources but cannot see any data (including transcripts) for the rest of the account.  Their view of the world is capped at their own folder.

Let's set up a mini-admin in 5 easy steps!

  1. Visit the Users page of the admin dashboard and create a new folder.  
  2. Add a new user within the folder.  You can use drag and drop if needed to get the user situated. 
  3. Head over to the ACLator and create a new group.
  4. Add the new user as a member of the new group.
  5. Click the Permissions tab and select your newly created folder containing the mini-admin.  Click Grant for read and write access.
The screenshots below shows how the ACLator Members and Permissions tabs look when it is all said and done.

The new user is part of the mini-admin group. 

Granting mini-admin privileges on the "mini-admin" folder.

When a new mini-admin user logs into the admin dashboard, they'll be able to create their own users, queues, knowledge bases (FAQ sites), conference rooms, canned messages, chat snippets, chat skins, and administrative hierarchies within their own restricted subdomain.  Since a mini-admin's view is capped at their own folder, they won't see their parent folder since that is beyond their administrative access.

Building Collaborative Services


Within the ACLator, you probably noticed the Collaborators tab.  This tab let's you manage access control for LibraryH3lp users outside your account.  Awhile back, we talked about how you can build ad-hoc collaboratives with LibraryH3lp.  The Collaborators tab lets you build collaborative services on-the-fly!  The ability to self-organize is unprecedented in the world of virtual reference software, removing the need for any middle man to form collaboratives.  In an upcoming post, we'll walk you creating and managing an ad-hoc collaborative.  Stay tuned.