November 16th, 2016 Brian Herzog
Earlier this year, when the woman who used to be in charge of all our email newsletter retired, I became the library's primary contact for our Constant Contact account. This meant I started getting the weekly emails about how many new subscribers we had, confirmation that messages were sent out, and I was the one to get the billing notices.
All of that is fine, except that after just a couple weeks, our bill jump. A lot. And I had no idea why.
After a little digging and a few phone calls to Constant Contact, it turned out that the number of contacts in our account had grown so much that we crossed into the next higher pricing tier. We were now over 10,000 contacts - that sounds great, but the price different was tremendous, and wasn't something we had budgeted for nor could we afford.
After the dust settled, the obvious occurred to me: Chelmsford's population is about 33,000 - there's no way that enough people in town could have signed up for our email newsletter to grow our contacts list that much.
So I started looking at the weekly new subscribers notification more closely, and notice something - see if you see the same thing I did:
To me, all of these looked like fake email addresses. Not just the .ru, .pl, .top, .site, etc, but even the yahoo.com accounts looked bogus. And we were getting 100-200 new subscriptions a week, so yes, it makes sense then how we could grow from a sensible subscriber base to over 10,000.
So now of course, I needed to figure out how to delete all these fake addresses out of our account to bring the total number of contacts down so Constant Contact would drop us down to the lower pricing tier again. Their Customer Service people I spoke with were friendly, helpful, and understanding, and gave us a grace period to get things under control.
I hadn't used Constant Contact much before this, but I quickly became familiar with their Contact Management area and deleting contacts. My first strategy was just to search for things like "*.ru" and "*.pl" and delete any address that came up, because I sincerely doubt that we have any patrons in Russia or Poland. I built myself quite a list of top-level domains to search for and delete every week, and even though it helped, it still required a lot of of my time.
The first day I started working on this, I deleted something like 2,000 contacts. That was a good start and gave us some breathing room to stay below 10,000, but I knew there were lots of other junk addresses in there that I needed to figure out how to eliminate.
I also wanted to stop the flow of new signups. I talked to Constant Contact again, but they said there was no way to block signups by country or domain. They said all I could really do was require First Name and Last Name during signup, and also use "confirmation opt-in" (where patrons must click a link in an email sent to them after they filled out our web form before they are actually subscribed to the list). I did turn on the first name/last name requirement, but didn't want to use "confirmation opt-in." That extra step annoys me, and it bugged me that real patrons would have to suffer (and possibly not get signed up) because of the jerks out there. Not to mention, there is no guarantee that this would keep the jerks out.
So I kept looking in the Contact Management section for something that might help. At one point I tried figuring out where these fake signups were coming from - we only have one signup form on our website, and that's it. No Facebook form, no other apps, nothing, but I figured these fake accounts must be coming from some kind of script somewhere.
On the Advanced Search screen, I saw one of the options was "Contact Source:"
That made me think that if I could just figure out the source, I could easily search for those and delete them. Nice.
Oh, and then on list of contacts, I noticed there was a way to change the view, and the second option included the source. Ha - it's all coming together now.
After skimming through pages of our contacts, I noticed something: the source for the real contacts were either "Added by you" or "Website sign-up form." The source for everything that looked like a fake address was "Embedded JMML."
I had no idea what "Embedded JMML" or where it was to be abused like this, but at least there was some commonality. Now all I had to do was an Advanced Search for Source=Embedded JMML and everything is fine.
Except: Embedded JMML was not one of the options in the Source dropdown box in Advanced Search. Arrgh.
But, I think I solved it anyway. Using a combination of Advanced Search fields, I was able to filter out all good records, and so the results were only records with the Embedded JMML as a Source:
And there were THOUSANDS of them. I skimmed through pages of the contacts to make sure no real addresses slipped through, but they all looked fake.
So I selected them all, braced myself, and deleted 4,000 contact from our account. Whew. At least now we're back down to a realistic number, and they all seem like legitimate addresses.
I contacted Constant Contact once again, asking if they can block the "Embedded JMML" as a source for signups, or at least tell me where these signups were originating. At first all answers were no, but the support person who got my ticket really stayed with it, and convinced the development team to look at adding this as a feature. Which is great. For awhile I was beginning to think it was a big Constant Contact conspiracy to make it impossible to manage our contact list, because their pricing structure is set up to charge us more for high numbers of contacts - but that seems crazy even to me.
I know this is kind of dumb, but it really was a problem. And oddly, I didn't see anything on the internet about other people experiencing this same issue, so maybe we're just unlucky. Still, I thought it was kind of interesting and wanted to share.
So, the end result of all of this is that I still have to spend five minutes each week setting up this Advanced Search and deleting all the JMML signups, but at least it's a functioning method to get what I want, it keeps our contact list accurate and clean, and patrons don't have to jump through unnecessary hoops to get the library's email newsletter. Maybe that's the best I can hope for.
Tags: bot, constant contact, delete, embedded jmml, fake, libraries, Library, newsletter, public, signups, spam
November 9th, 2016 Brian Herzog
I was reading book reviews while doing selection for our non-fiction collection, and came across one for Age of Jihad: Islamic State and the Great War for the Middle East, by Patrick Cockburn.
After what I read I still couldn't decide whether or not to buy it for the library, so I went out to Amazon to see more information.
The bits of information on Amazon records I always look for are publication date, format, number of review stars, publisher, best seller rank, and also whatever is in the editorial review section.
Often these are blurbs written by (I'm guessing) paid reviewers, but sometimes I recognize names as someone I'd expect to be an authority on the subject. Sometimes they're even full-blow Publisher's Weekly, Kirkus, or other review journal reviews.
And sometimes they completely surprise me.
When I looked at this section for Age of Jihad, I found this:
So is this a thing now? Being mentioned flatteringly in leaked email correspondence is a source of promotional material? Huh - that doesn't seem like the best use of the technology tools we have, but I guess it's the world we live in.
It hadn't occurred to me until I saw this, but I bet a whole bunch of people immediately did Ctrl+F for their name as soon as the text of her emails were released.
June 2nd, 2016 Brian Herzog
Recently one of our patrons submitted a purchase suggestion for a book she saw reviewed in the Boston Globe: Old Bags Taking a Stand, by Faith Baum and Lori Petchers
The patron had also clipped and stapled a little part of the Globe's review, which made it sound interesting enough to investigate further. In cases like this, my first stop is Amazon to see if anyone else has reviewed it.
Now, one of the little games I play when searching - for anything - is to try to type in as few keywords as possible to get the result I want. Known item searches are of course the easiest, and in this case, I just typed "old bags taking" into Amazon's book search - and laughed out loud at the result:
Yes, the correct book was listed first, but the second result was what struck me. Little Women? Really, that's the exact opposite of "old bag" in this sense. Amazon, what in your search algorithm matched these two books?
I've showed a couple coworkers this and their search results varied slightly, but Little Women was always on the list. It seems weird, but maybe this is one of those reader preferences computers can identify that people wouldn't - maybe Little Women fans really are the target audience for Old Bags. Hmm.
May 27th, 2016 Brian Herzog
I mentioned this in passing at a meeting not too long ago, and it got enough interest that I thought it'd be worth posting here.
The meeting topic was library marketing, especially for programs. I mentioned that my library occasionally boosted Facebook posts to great success - and it seemed like no one else in the room had done this for their library.
Not that we're experts - above, "occasionally" means twice in the last three or so years. We've only done it for huge events (a major author coming to speak kind of events) where, with just about a week to go before the event, the number of tickets we'd given out was frightening low. We'd be scrambling trying to push awareness of the event however we could, and so we'd boost that post on Facebook too.
In general, our posts go out to an average of 400 people. That varies wildly, but that's probably a pretty good average. When we boosted the posts though, those would reach 5,000 people.
I think we allocated about $20 for the boosts, which we paid for on the library's credit card. Later, the Friends reimbursed us from the programming budget, since it was program advertising. We'd choose a new targeted audience by location, and just use Chelmsford's zip codes.
And our events were successful, so I'm inclined to say this was worth it. Tougher if your town only lets you pay for things by check, but still worth looking into because it isn't a whole lot of money and does seem to help us reach people on Facebook (although, I do hate having to play Facebook's game).
Does anyone else do this successfully, or regularly? Has it ever backfired on you? Please share your experiences in the comments. Thanks.
April 23rd, 2016 Brian Herzog
Since I am helping patrons less than before, the relative number of reference and tech support questions I get from staff have gone up. This is one of those.
One day right as I was getting ready to leave for the day, someone working at the Circ Desk called down to say that Evergreen (our ILS) wouldn't open. That's weird, but important enough for me to stick around a few minutes to try to fix it.
When I got upstairs and asked for details, my coworker said,
Well, something bad happened to the computer and it froze, so I forced it to shut down and then restarted it. When it came back up and I clicked on Evergreen on the desktop, it wouldn't open.
Now that is odd. I had no idea what "something bad" might have entailed, but at least the computer was on and working, so that's a good sign. I double-clicked the desktop shortcut and sure enough, it didn't work - I got that "cannot find target" error. Thinking just that shortcut got changed somehow, I tried the icon in the Start Bar, and then the one in the Start Menu, but kept getting the same error.
Hmm. So I looked at the target of the three, and all of them were the same. Also odd. I browse out to that directory, expecting to see the evergreen.exe file they were pointing at, but it's not there. I check look at that directory on a different computer that is working, and sure enough, there is an evergreen.exe file.
Now that is weird. This was far enough down into the directory structure that I didn't think any staff would have accidentally deleted it, but I couldn't think how else something would have happened to this single file and left everything else in the directory.
The immediate fix that comes to mind is to completely reinstall Evergreen, which is a pain, and I'm still trying to get out the door to go home. So, I figure what the heck, I'll just copy/paste the evergreen.exe from the good computer into that same directory on the problem computer and see what happens. This is like Windows 3.x stuff, and figure it's an incredible long shot.
But holy smokes, it actually works! I copied that file to the network and then pasted it in from there, and when I clicked the desktop shortcut on the problem computer, it opened right up as if nothing ever happened. I don't really understand it, but I'll take it - at least as a temporary fix to get Circ through the evening.
In the morning I asked our IT guy to reinstall Evergreen on that computer for real, because I figure what I did was fragile and didn't address whatever the "something bad" was that started this whole thing. Before he did that, he did some checking on the computer and then got back to me:
Symantec classified Evergreen as a virus yesterday. I didn’t check but I presume that Evergreen.exe was moved to the Quarantine area. When you copied it back to the original location you resolved the issue. There should be no need to reinstall.
Sort of an unusual thing. I modified our Symantec policy to exclude this file. It shouldn't happen again.
Wow. I have no idea why Symantec suddenly took an intense dislike to the most important application we use every day, but there you go.
Tags: antivirus, catalog, evergreen, ils, it, libraries, Library, public, Reference Question, symantec, tech support
April 16th, 2016 Brian Herzog
As Assistant Director, I'm definitely spending less time at a public desk than I expected. As such, far more of my patron interactions are actually either referrals of difficult patrons from the front-line staff, or else just stories heard second-hand. The latter is today's question.
We're doing some construction next door to my library, which means staff temporarily needs to park across the street. It's a longer walk, which is fine, but it also means we park in the lot of a strip mall that houses restaurants, a gym, drug store, bank, etc.
As I was crossing from that lot to the library one day, coming the other direction was a coworker from the reference desk who was leaving for lunch with her friend. They were going to the restaurant in the strip mall across from the library, since it was so close.
Later that day I asked her how her lunch was, and learned that the restaurant was closed - but only temporarily. Apparently there was a sign on the door saying the owners were on vacation, and they'd closed the restaurant for a week while they were gone.
She was annoyed by this and shocked that any business owner would do such a thing, but I think it's kind of cool that small businesses like this still exist, and only operate when the people who run them are available.
My coworker ended her lunch story by telling me to wait a week before I went there. However, this a Japanese restaurant and thus not on my "wheel of lunch," so I didn't think her news would be useful to me.
But then, two days later, that same coworker told me this:
Brian, you're never going to believe the call I just got. The patron said,
Hi, this might be a crazy question, but I called Town Hall and they referred me to you. Do you know if that Japanese restaurant across the street from the library is open?
Ha. She and I might have been the only people in the library able to answer that question. You never know what information will be helpful to you as a reference librarian.