APS1: Using WordPress to Power Your Institution's Entire News Presence

Lacy Tite 
Web Developer, Vanderbilt University

The audio for this podcast can be downloaded at http://2011.highedweb.org/presentations/APS1.mp3

Announcer: This is one in a series of podcasts from the HighEdWeb Conference in Austin 2011.

Lacy Tite: I'm Lacy, and I'm the lead Web developer in University Web Communications at Vanderbilt. I want to give you kind of a 411 on Vanderbilt because it explains how we do news and how we prioritize things.

We're a private university in Nashville, Tennessee that is consisting of 10 different colleges. We have 12,000 students. We have 23,000 employees because we have a medical center.

We are a research institution, so everything that we do has something to do with research, and that is our main strategic priority for everything that goes on at the university and the medical center.

My office, Web Communications, manages all of the top-level websites for Vanderbilt, all of the social media presences, and all of the content management systems.

Before we got our news presence under control, this is what it felt like when we were trying to do anything with any of our content. We knew it was in there. It was all over the place, but we couldn't really get to it or get it to do anything.


So about 16 months ago, we decided we wanted to try and get our stories and our news under control. Our chancellor had asked us to make research a huge priority, and we had all of these research stories all across various different websites and we wanted to centralize things. So that was the impetus that we needed to get started on merging all of our news sources.

What I'm going to be talking about today is where we were, how we got to where we are today, and how we're using WordPress to make a lot of this happen.

So we wanted to simplify things. Before, what would happen is we would have a story and it would be posted minimum of probably four times on four different websites. The same exact story.

We'd have a user posting it on our general news site, on our employee news site, on a magazine, and on a research site. So anytime anything needed to be changed in that story, you had to contact four different people, get it changed. Who knows when it would be changed, if it would be done correctly.


The formatting was completely different across all four websites, so we were like, 'This is really not an awesome system.'

Now the system is, we post a story one time, and then we push it out to all the different various presences that it needs to go to. So we only have to format it once. Any changes, only done once. We're good to go.

Within the past 16 months, we've merged these news sources so far. Our main news site we merged, our VUCast, which is all of our video and other media presences, we merged, myVU, which is our employee communications, which, again, we have 23,000 employees so we have a lot of internal communication with our employees, Exploration, which was a university research website, we merged, Lens, which was our medical center research news site merged in, Vanderbilt View, which is a monthly magazine that goes out to all employees, we merged that in, and then we have an Experts site that we use for mass media because we're constantly having our faculty and doctors on the mass media networks, NBC and ABC, talking about their research. So we've merged all of these into one.


Coming within the fourth quarter is we're merging in the medical center news site, and with that merge, that will then mean we have one centralized news source for the whole entire university and medical center, which everyone is very excited about.

And we haven't had to convince anyone to do this. We've had people ask, "Can we be merged into this new system?" So people are really happy with it.

Merging all of your internal content sources will make all of your content more manageable. Manageable, I mean, that's the easiest one because you're only posting at one place, you don't have so many people involved in posting your content.


It makes it more useful, because having a large content source available means all of our colleges and departments and organizations, they don't have to maintain their own news source, either. A lot of our departments just use this as their news source and they just pull a feed of their department news. Our cancer center can pull a feed of all cancer news.

We'll talk a little bit later about how having it all in one system increases the related content that you can then pull. Someone's reading a story, they have so many more opportunities for seeing related content, which is why it makes it more interesting. Because if someone comes to your site, they're reading a story about something, they're already interested in that story, so let's show them some other things they might be interested in. And when you've got all of that in one system, that just increases the number of stories that you have to lead your reader onto.


It makes it more interactive. We use Disqus for our comments. By having only one story, all of the discussion that's going on in social media and elsewhere is in one place, whereas before, we had the stories posted, sometimes stories would be on seven or eight different websites, and there might have been a really interesting discussion going on on the university research site, but on the medical center site there might have been none, but maybe if that discussion had all been in one place, it would've gotten a little bit more interactivity.

So our content strategy is to provide multiple front doors for our content. What I mean by 'front doors' is having that homepage for all of these different audiences. Our main front page is the Vanderbilt news page, so that's news.vanderbilt.edu.


On this page, what we're pulling through, again, everything that I'm showing you is from one WordPress install, one content source, we're pulling in featured stories, the latest stories, our latest featured audio and video stories, our latest featured research stories, our featured experts, and then we pull in magazine stories from across the campus. So this is one front door.

Another front door is myVU. This is our employee news website. On this one, we're pulling featured stories for the myVU audience, the latest stories for the myVU audience. We pull events from our university calendar that have been tagged for our employee audience. We pull stories from our employee magazine onto this page. We pull Flickr photos and more magazine photos.

Another front door is Researchnews@VU. This is yet another way that we can just pull only research center stories. All of this is being powered by tags and categories, etcetera.


And then another one is a subsite of myVU, which is the Vanderbilt View. That's the little logo over there on the right. In this one, we're in myVU because this is an employee publication, so we're pulling featured stories and then subpages related to the View.

So these are all front doors to the same content source.

How you create front doors in WordPress is through page templates. When you create a page template, that's just in your theme file. For each one of our front doors, we have a page template assigned to that, and that's where we can go in and say, 'We want you to pull only research stories,' 'We want you to only pull myVU stories that are featured,' etcetera.

So when you create a page in WordPress, then you just create your page as a normal page, and then over on the right, in page attributes, that template is going to show up in that drop-down and you just assign it. So then when people go to news.vanderbilt.edu/myvu, it pulls that myVU page.


This is one of the little queries that's on the research page. First of all, on all of our pages, we set up a 'no duplication' array because we don't want a story that shows up in one section to then reappear in another section on that page. So we say, 'We don't want any of these stories to show up again.' So then we're going to set up a new query and we're going to say, OK, here we're saying, 'Pull six posts that are tagged featured research.' And then we add all those post IDs to the 'no duplication' array.

And then here in the middle, you can style each post however you want to, whatever you're using it for. So we're saying, 'It's a permalink,' and 'Pull through the image,' 'Pull through the slider text.'

Here is how we're pulling a research subsection. We've got our main research page, and now we're going to pull only engineering and technology posts. Or actually this is Med Life.


So we say, 'Pull everything in this certain category.' So we have a Medical and Life Sciences category. 'Pull those. We only want to pull three. And be sure you don't pull anything that's been duplicated in the slideshow.' So that's where the 'no dupes' comes in. And then, again, we just cycle through however we want that content to display on each page.

So we do this for all of our front doors. We create these page templates.

How the user on the back end decides how they want a story to show up and which sections they want it to show up in is just in their categories. We have categories that help them determine which sites it shows up.

We use categories for organizational purposes. We use tags for topics and for more specific things, but the categories are very broad-based groupings of content. And you can just see a couple here.


So on this story, they've said, 'I want this to be a release.' So that means it's going to go on our regular news site. 'We want it to be on the research site and we want it to be in the engineering and technology subsection.' That's what the user sees.

On the back end, we've done all the work to say, 'OK, they've said they wanted these three places, so now we're going to make it show up on all of those three front doors.'

The most important thing, some of you might have already realized this, this is all one content source. So it's very important that you establish your brand hierarchy for the individual stories so that you can determine what that story individually is going to look like when you go to it because this is all one single source. So something might be part of your myVU website, your research site and your news site, but the individual story can really only look like one of those.


So you have to think about that strategically. And again, for us, research is the most important thing, so if something is on our research site, it's going to look like a research story.

To determine what you want your individual stories to look like, that's done in your single .php template. Here up at the top I'm saying, 'If it's in the research category, include the research header. Otherwise, if it's in the myVU category, include the myVU header. Otherwise, just include the regular Vanderbilt news header.'

So you might have as many of these as you want. How many front doors you have is how many 'if else, then' statements. And then you do the same for the footer, you can do the same for your sidebars. However much you want to customize your template, you're basically just deciding whatever that brand hierarchy is is the order you're going to do your 'if else' statements in.


If we wanted to have myVU be the most important thing, we would make that one first. But for us, if something is research, it's always going to look like a research story. I'll show you what that basically means.

Based on what type of story this is, we've got various different headers, different sidebars, different subscribe options. Sometimes there's a big video player that shows up.

Here is a research story that's also in the View. It's using the research header, but it's in our View magazine so that logo shows up over on the right. So we know it's from the Vanderbilt View magazine, but it's a research story.

Here's a story that's only Vanderbilt news, so it's not on myVU, it's not on the research site. It's only there so it's using our default Vanderbilt news header, sidebar, etcetera.

Here's a video story that's just part of our news site. When you say you want it to be a video story, then we have it show a large video player at the top, and then it shows the different related videos down at the bottom and it pulls a different look and feel. Here's a story that's only myVU. It's saying it's myVU, so it's going to use the myVU header.


That's what I'm talking about when I'm talking about the brand hierarchy, is a story can only look one way.

Sometimes you're going to run into some political issues if you try to merge these things because people want things to look like their site. That's when we start talking to people about why it's so powerful to have all your content in this system.

Usually within the first month when their traffic doubles, they are totally cool with the fact that they have a myVU header in their Vanderbilt View logo on the right, because they're getting so much more traffic. Because they're in this larger content source that's being pushed out everywhere, their content is appearing as related stories on the research site because it's all in this one big content well.


So here is how we've got various different programs pulling in news onto their sites. We've not only merged existing news sources but we have a lot of smaller sites that were maintaining really tiny news presences, where they might only post one story every couple of weeks. So a lot of these groups have decided to forgo that, and now they're just pulling a feed from the news site that fits their office's news.

Here we've got the Neuroscience program. What they're doing over on the right, where they used to just post their own stories and maintain it, is they're pulling an RSS feed of stories from the Vanderbilt news site that we've tagged 'neuroscience'. If they've got something that we haven't posted, they can just send it to us and we'll post it, tag it 'neuroscience', and then we might want to post it on some of our other news presences.

The Human Resources site over here on the bottom right, they used to have their own news section. Now they pull all of their stuff, things that are tagged 'HR', and then it's Human Resources, so they thought, 'Well, we're just going to pull all the news that's tagged for employees as well.' So they pull HR announcements and then they pull staff news as well.


Really cool is, because we have all of our video in our news site as well, we post it through YouTube, but we also have the files on our streaming server because we can build video playlists for events.

So we have a video playlist for commencement. When commencement activities start going on, we switch over their homepage and we have a video player that automatically updates as soon as any videos are added for commencement.

And, again, this is all automatic. We just do the coding on the back end and all the user does is say, 'Oh, yeah, we have a big party the night before commencement. That video should go on that playlist.' And it automatically shows up.

We've also built into all of our content management systems. We use WordPress and Omni of Vanderbilt half and half and we build into our templates the ability to pull through news feeds based on tags, based on categories, etcetera.


So someone on any of our websites can say, 'I want to pull a feed onto my page of all the news related to breast cancer.' All they do is put in the URL for the feed and it pulls it through with thumbnails and looks really cool, and it looks like they have their own news section that's nicely formatted.

The other thing that made us want to do this was we wanted to connect with our users and connect our users to more content. So how do you keep users on your site? You show them more of what they want.

We have lots of really cool things going on with research, and people will click on them and read them. You know they're already on your site, they're reading about neuroscience or whatever, so how do you keep them on your site? You give them related stories.


Because we have this now really large well of content, currently we have about 7,500 different articles within the system, and more and more when Medical Center joins in. That will just get enormous

All of these are tagged. Every single story is tagged about what the story is about. They're tagged with any faculty member's name, any doctor's name, any research agency's name, any funding agency's name. So if we wanted to pull all of our stories related to NIH funding, we can do that. We could have a page about, 'Here are NIH-funded projects.' We love tagging all of our content, which increases the reliability of what our related stories are.

On all of the stories, no matter what kind of big brand they belong to, over in the right column, these are related stories.

We're actually using a plugin for this. We don't use very many plugins for WordPress because I like to write things so I know what every single line means. So whenever WordPress upgrades, I don't have to worry about a plugin breaking things.


We are using yet another related post. It works really awesome. And we can tell it to do only during a certain period, so we only have it pull related things from the last 10 months. Because with research, a lot of things, even 10 months is iffy. A lot of things can change within 10 months. So we wouldn't want something that's been disproven showing up as a related story. We can go in and tweak that and we can tell it to ignore certain stories as well.

There's always going to be about four related stories over there. We can follow through our analytics that this actually does keep people on our site. Because they're already interested in the topic, then you're just leading them along. And of course, when they go to another story, then it's showing four more related stories. So it's kind of a never-ending thing that if they're interested in a topic, they could spend an hour on your site just going through all the different stories about the research.


Or the user can actually click on any of those topic tags and see a listing of all the stories that are specifically related to that. Again, we tag everything with all the topics and the faculty members' names.

So this is how we keep people on the site and help people find what they're looking for.

Another way, just randomly, we had a flood in Nashville about a year and a half ago. A lot of stuff was going on with the flood at Vanderbilt and in Nashville, so we were easily able to create our flood archive page because we just tagged everything 'flood 2010'. So then we automatically had a feed of all of our news items related to the flood.

And it's on the fly. You don't have to think about it. You don't have to get a developer involved. Our news service just did that, and then I didn't even know about it. My Web Comm office, they just create their own tag and then they started linking to that archive for all the flood news.

And that's where it's really cool, when you can build things into the back end to where your developers do not have to get involved. I love that, because we support a lot of clients. Almost all university web offices are overwhelmed with work, so anything that you can do to automate things is better.


One way that we really, really, really love WordPress for our news is we send out a lot of emails at Vanderbilt. On Mondays and Thursdays, we send out an email to all of our faculty and staff with the news from the last four days.

Previously, it would take one of our staff members about half a day at minimum and 75 phone calls to my office to make this email happen. So we decided, 'Let's just take that out of the equation.' Now, we have the email automatically generated, she goes to the page, and it takes like two seconds.


You can do this, again, with a page template. So create a page in WordPress, call it 'myVU email' or whatever, and then create this page template in the back end that pulls what you want to have in that email.

Here is our myVU preview email. This is what goes out to our faculty and staff twice a week. Over here on the top right, we're saying, 'Pull all of the stories in myVU that have been tagged 'featured' that have been posted within the last four days.'

Then over on the right, we'd say, 'Pull all of the upcoming events from our calendar,' not WordPress, 'the next eight events.' Underneath we've just got, 'Here are the stories on myVU that weren't tagged 'featured' but we want to put them in here anyway.' Over on the right, we say, 'Pull any video that's been tagged 'for myVU' and 'featured'.'

Again, automatically-generated, and they just send it through our internal list serve. And now someone has six hours extra of their day on Monday and Thursday to do something else.


We send out a research email every Friday with all the stories that have been posted within the last seven days. This one wasn't actually being built before, but when we did the myVU template, suddenly everyone was like, 'Wow! We can do these emails really easily!' So they got really excited about creating these little sublists. Again, this is automatically-generated, so it pulls all research stories within the last seven days and all the videos tagged 'featured research media'.

Another email automatically-generated, the Vanderbilt View magazine, goes out once a month. Here we just pull all the stories in Vanderbilt View that have been posted within the last 28 days. It shows large thumbnails for ones they feature and it shows little thumbnails for ones that were just regular stories.

Audience 1: Excuse me.

Lacy Tite: Yes.

Audience 1: Do you create email from the template?

Lacy Tite: Yes. It does the nice little tables and inline styles and all that really gorgeous email code.


Yes. It's really awesome.

I just don't want anybody to view source on those, because you can actually. We've got these email templates, so actually the research email, they actually tweet the link to the email. They send out the email, and then they tweet, 'Hey, do you want to see last week's research?' so they can go right to it. So people actually could view source, which drives me crazy because it's so horrible.

So email template basics, to go to your question. There's lots of really cool ways you can do queries in WordPress. This one, we're setting a political function where we're saying, 'Look at the post date and subtract four or five days from it and pull all those posts.' And then we're telling it, 'Do that, but only get ones that are in myVU and only ones that are featured myVU.' Then if you have the posts, do all that lovely HTML code, and then end that query.


Here's some of the lovely HTML code, which I hope you can see, where we've got our table and our inline style, padding, and even our font family. So awesome to have that inline. However you want your email template to look, this is what you're going to build your template to do, so yes, you do have to do it for HTML.

We're lucky in-house with our internal email. Everyone internally is using Outlook, so we don't have to do as much with that email because we know they're using Outlook or they're using Outlook Web Access. We still do everything inline, but at least we have our unknown audience. We know exactly what clients they're using to access our email.

Our research email and all the other emails are who knows you could be in them, so we have to be a lot more diligent about making sure that it's going to be compatible with all those email clients.

Another really neat thing, because you're automatically generating this, we could never have gotten our content publishers to do this before, is we can add Google Analytics Campaign Tracking to every single link within the template, so then we can track how effective that email is internally.


Because the myVU email is an internal email sent through our internal email system, we're not using My Emma or ConstantContact or MailChimp, our internal mail system doesn't really have any way for us to track things. So we can just do campaign tracking on these links, and that way we can go through and see how effective are our internal emails.

Again, we're just adding a little var at the very top of the template and you can set whatever campaign tracking that you prefer to use. In here, we're just saying, 'We want to track the myVU preview.' The medium is email and the campaign is myVU preview, plus the date.

So then we can go in Google Analytics and see how effective the myVU preview was on Monday versus Thursday, if anyone reading this email at all.


Audience 2: Lacy?

Lacy Tite: Yes.

Audience 2: Are you sending those to some external mail list there or are you sending it to the actual users that?

Lacy Tite: The email? We have an internal list serve. All of our faculty, students and staff are in it, and we can send HTML email through that. So we take the code from this and just send it through that. But we can also take the code from this and put it in My Emma or MailChimp if we wanted to if we were going to send it out to an external audience.

Another thing that you can do automatically in WordPress that saves a lot of time is schedule stories to post and then go do something else. We have a lot of news releases that will be embargoed for various reasons because of funding agencies wanting to make the announcement first or whatever.

Sometimes things have to go live at really bizarre times like 4am. No one really wants to be sitting there to just have a post go live at 4am. But if you were to say, 'We're going to have to wait until someone gets into office,' you would not have happy people. So we can just say, 'Post a story at 4am, whenever you have the story ready,' and it will go ahead and do it.


If you don't want to create a customized email, hook your WordPress up to Feedburner Emails and it will do a lot of this for you if you just put your feed into Feedburner. Turn on the email option, add the subscription form to your website, and you can set it to how often you want it to send the emails.

We actually have this set up so that if someone really, really, really cares about every single news story that we post, then this will send them an email every single day with everything that's been posted that day. So we just let Feedburner do that because all it will do is show the title, a little excerpt, and it automatically handles all that for us.

Another automatically fun thing you can do. We've got all these different front ends that use images all over the place and we've got slideshows and thumbnails and little video things, but we have people who aren't versed in Photoshop. They aren't really well-versed in proportions and how images should look.


So what we tell our users is, 'You need to upload an image that is at least this big.' We tell them how big the image needs to be for the largest image we're ever going to use.

For us, our research site has a slideshow that's about this big. You can tell how many pixels that is. We say, 'We need an image that's at least this big,' and then WordPress will do everything else for us, because across the system, we've made the decision we're going to pretty much go with horizontal or square images for all of our slideshows and thumbnails and all that stuff.

And you will have to do that if you're wanting it to automate this is you're going to have to say, 'We're either going to be a horizontal chop or we're going to be a vertical chop,' because if you're wanting to automate it, it's got to know which direction to go in and crop things.


So again, our users go in and they upload an image that for us, we tell them, is about 700 by 400. So they upload that one image, and then it automatically generates all these other images that are proportionate to that, or a cropped square, because it goes in the middle of the image.

To do that, in your functions.php file... And this is native to WordPress. This isn't something that we've magically discovered. WordPress just rocks. If you go into your functions file, you add theme support, which, if you've downloaded or used WordPress within the last several versions, this is already on, and then you add in all the different sizes that you need and it will automatically generate it.

So whenever a user uploads an image for us, it automatically generates something that's 685 pixels by 350. And then it says, 'Oh, I need one that's 300 by 175, I need another one that's 170 by 95, and I need another one that's 215 by 110.' So it generates these four because I've told it to.


But it also generates all the other ones. If you've looked in WordPress and you insert an image and it's got large medium, it also generates all those. But these are really specific sizes that we need for certain areas of the site.

So our users are always going to be able to use that medium image. If they want to have a vertical image on a press release or a story, they can, but this is what we're going to use for the slideshows and for the thumbnails and all that stuff.

Audience 3: Question.

Lacy Tite: Yes.

Audience 3: How do you guys manage all of these media?

Lacy Tite: For video or audio or images? Ask me at the end. I'm about to clear up, and then we can answer your questions about how we're staffed.

Shorten. And this isn't necessarily WordPress, but it is related to how we brand our content. We want to always put forth 'This is Vanderbilt content' no matter where we're posting it.


So we set up our own URL shortener, vanderbi.lt. It's really fun talking about that in meetings because I always say it 'Vanderblt.' But it took us a while to find an extension that would let us register it, because a lot of these are owned by little countries. We tried something with .vu and it wouldn't let us. But vanderbi.lt's really cool.

So we registered this domain, and then we actually set up YOURLS, if you've heard of that, the open source shortening. We set that up and customized it so that it's tied into LDAP for us, so we can have users log in and manage only their own link. So we have one big URL shortening service.

And super admins, we can log in and manage all the links that have been created. But we have a lot of little departments who wanted to be able to shorten their own links, so we hooked it into LDAP. Now, Joe from the College of Education can log in and manage all of his links, but he can't change or edit any of the links that the School of Engineering has created.


And again, this is open source. So find a cool URL for your school and then just install it and hack and bash it to get it to do what you want it to do. It's really powerful.

This allows us when we're tweeting our stories that we can have a neat little shortened URL in Twitter. We also use these in print so that if we've got something in some of our View books or whatever, we can use this little short Vanderbilt URL instead of TinyURL or bitly or all the other different ones. This is like, 'Yeah, this is Vanderbilt content.'

We've got this hooked into WordPress via a plugin so that anytime a story is published, it automatically generates this URL. So they don't have to go out and do it. It does it from within the system.

They can also tweet it within the system. There's this little Tweet box and it will pre-populate with whatever the title of your post is and put the little shortened URL in the end there. You can change that text if you want.


And it knows if it's been tweeted, so if it's already been tweeted, it doesn't show you the Tweet box. So you don't have to worry about someone tweeting a story twice.

And again, all open source. Just Google 'YOURLS'.

One of the reasons we really like that version of doing our own shortened URLs is we can have really simple statistics for those, so we can see how many people use that shortened URL and where they're from, and all this is in a database and you can pull whatever data you want to pull. But it does really basic charts for you about the highest hits of all time.

Here's a story about a bionic leg, which is awesome. On one day it got 886 clicks on Twitter. Whoo! But it tracks all of this stuff so we can... For a print piece, this is really cool, because you can see who actually went to a link from a print piece. Because if you put a specific URL in there, then the only people that use that are the ones that saw it in print.


So again, not related to WordPress, but it's tied into our WordPress, and it helps us, again, brand everything as a central news source.

One of the last things I want to talk about is help. Our office, again, is a central Web office, and until about four months ago I was the only Web developer in the office. Now we finally have another person, which is really, really exciting.

So being the only one in there and supporting pretty much everything on campus, providing help for people is so key so that you don't get a phone call or an email every time someone is like, 'How do I do a link again?'


You can build in help menus into WordPress for your users, and you can go over really, really generic things on those help pages or you can go over really, really specific things.

Basically my thing with doing the help pages is, whenever I get an email, I look and see if it's in the help pages already. If it's not, then I'll add a help page in there so that in the future, people will look there first. And they really are trained to look at these help pages first before emailing when they have a question.

Here, we've got users, they're logged into WordPress. Over there, on the left column, they see this menu. So it's like, 'How do I post a story? How do I post a video story? How do I do images? How do I do tags? How do I do live video? How do I schedule a story? How do I post an external story?'

The external stories is, yes, they're still content that is not in the system. There is always going to be content all over that is not in the system, but we want to pull it into our system so that it can start showing up as related stories.


For stories that are in some of our alumni magazines, for instance, we post them in WordPress. I should've put this on a slide earlier. We post them in there with the title, we tag them and put them in categories, but the page template for that story redirects it to the original location.

What that means is those stories can show up in our archives, they can show up on any of our front doors, they can show up as related stories, but then when you go to that story, it's going to take you to the original source.

And we did that because, again, there's always going to be content that we're not going to be able to get in our system but we want it to show up because there's still awesome content out there that's not in our system, and we realize that.

Again, we've got, 'How do you embed a video? How do you embed YouTube? How do you use shortcodes?' We have various shortcodes set up so that, for instance, we have a live video studio for use when the networks interview faculty and staff.


So we have a shortcode where if they put 'view cast blurb' shortcode in the post, it spits out this whole paragraph about contact zone, for 24/7 video services, blah blah blah, so I don't have to retype that every single story where they're talking about a faculty member. We also have one that's like maps. If they do that, then it provides 'view the campus map' and then 'directions to campus,' blah blah blah.

So if you've got content like that that's constantly being posted over and over and over again, then you can create shortcodes so that your users don't have to constantly be typing out that stuff.

'How do I resize images?' We have people who use picnic.com, which is so easy and anyone can use it and it's free. So that's how, if we're wanting them to resize their images before uploading, we're like, 'Go to picnic.com, resize some of your images to our big image size that we need, and then bring it over to WordPress.'


'How do we do a gallery slideshow?' And then there's various different things for the individual front doors. This will save you so much time.

Here is a couple of them using images and stories. This is where we're talking about, 'Here's the biggest image that you'll ever need.' And then we go through 'How do you set the featured image? How do you insert an image into a story? How do you embed YouTube in a post?'

This takes time to set up, but it saves time. It might have taken me 20 minutes to set up the 'using images and stories' page, but it's probably saved me three months of emails and phone calls already, because now people will just go here and look, and then they don't email me, which is awesome.

We'll do questions now. This presentation is available at lacytite.com/heweb11. Later today, I will actually have some code samples up there. I tried to get it up before the conference but, 'Only Web Developer on Campus' and it just wasn't going to happen.



So that will be there later today with code samples of how you do the email templates, how you do the page templates, how you do the different single templates, how we do the redirect files, and all that.

So question back here of staffing. We have a news and communications office at Vanderbilt, which is across both University and Medical Center.

We have staff photographers, and the staff photographers take pictures for all of our different publications. So all of the different magazines, the photographers work for them. They take photos for the websites, etcetera. So that's where we get photos. We also have a video department which is our view cast. We have three full-time videographers and two producers that produce all of our videos.

Again, so much of this is tied to the fact that we're a research institution and we are constantly promoting our research out to everywhere, so that's why we have a lot of video staff is because we're videoing a lot of stuff with the medical center and with various bits of our schools.


Does that answer your immediate question? He's writing.


Audience 4: How do you do the custom help? Is that a plugin?

Lacy Tite: It is a plugin, and I'll put that on this page.

Yes. Yes.

Audience 5: As far as WordPress itself, what kind of daily visit are you supporting with it? Is it set up with just one server?

And also, the other question I had was related to upgrades. How long do you wait between upgrading when a new WordPress vulnerability is shown?

Lacy Tite: Oh, wow. We are across lots of virtual servers. On Mondays and Thursdays, that's our largest spikes because we get probably about 30,000 hits on those two days because of the email.


We can sustain that much traffic. The server admins are on watch on days that we send out the emails, and then when we know where really big stories coming out because they can add more virtual machines in to handle more connections.

And then upgrades, we do wait a while because there is so much going on in here that normally what I'll do is create this, back up the site, put it somewhere else, run the upgrade, make sure everything is still going to work, and then run the upgrade after hours.


Audience 6: When you got started with this, did you have a specific customer that you were doing this for? Or did you just have this idea and start creating it in Vanderbilt?

Lacy Tite: Web Communications is part of the News and Communications Office. News And Communications wanted to redo their new site. Because, again, it was the whole 'We know it's in a computer but we can't get it to do anything,' we had been trying to figure out what we wanted to do. I was really well-versed in WordPress, and WordPress is built to do dated stories, news like that, so that's how we kind of got started.

And then, the news site launched, and then within about six to seven weeks, they were like, 'Hey, why don't we move myVU into this?' And then literally it was just a cascade of, 'Hey, why don't we create a research site? Hey, why don't we put this in there? Oh, I know, why don't we do this?' So that's how it came down from that.



Audience 7: How long has your site been up?

Lacy Tite: We had a previous news site before this, so we migrated about 4,000 articles from our old system into the new system. And that was super fun, to migrate that. That was a multi-step XML, XSLT, 'trying to get it in the right format, shoot-me-now'.

Now a lot of times when we're merging people in, I try to convince them to do a 'before and after' scenario, unless they just really want to migrate all their content.


What was your other question?

Audience 7: I was curious, [indiscernible]. Right now we're just bringing in our external and our internal, and they were built in our system. Some of them, we're going to be using some of the same. It's nice that you have all this related.

Lacy Tite: Well, what we did when we migrated everything over, because the old system didn't really have a lot of tags, is we had student workers go through. And we did about two to three months. Student workers are awesome. Grunt work. We had them go through about two to three months of content and go through and back-tag it.

Another issue related to migrating content is, as I said, one of the reasons we did this whole thing in the first place was we were double-posting stories all over the place, so it's been a problem merging some of these sites because the content is already in the system.


So there is some cleanup that has to be done when you merge a site that you were previously double-posting with. Then those stories are now in your system twice. Another student worker project is going through and getting rid of the second one, or the third one, or the seventh copy of that story.

Audience 7: So related to that, when you [indiscernible]?

Lacy Tite: Yes.

Audience 7: [indiscernible]

Lacy Tite: Tweet me, and I'll send you how we made that happen, the two different styles.


Audience 8: Are you using [indiscernible]?


Lacy Tite: Yes. Not on this one.

I was just telling them, we launched a multi-site for our faculty and staff to have websites about five or six weeks ago and sent out an email to the staff saying, 'This is available,' and it's got our WordPress themes on there. Within 90 minutes, we had 87 faculty sign up for sites. So now we have about 420 sites in this within six weeks.

But it's multi-site so it's kind of self-serve. It's tied to LDAP. I don't even know who all is in it, what they're putting on their sites, because it's all automatic. We only get emails when something goes wrong.

Audience 8: [indiscernible]

Lacy Tite: We do. Like all of our, yes, and we do that in pages. We've got post and pages. The news site itself has probably about 125 pages, part of it, and then the 7,000 articles or so.


So we do have lots of content-rich, the pages, especially with our Experts section, where we're basically pimping out our faculty to the news networks. [Laughter] So we've got their bios.

And then, again, because it's WordPress, we can say, 'Here we've got one of our Poli Sci professors who rocks,' and we can have it pull in all of his news mentions based on his tag. It's really, really cool when you start doing stuff like that because now you've got a tag for every faculty member.


Audience 9: What streaming server are you using? Does it connect into YouTube automatically?

Lacy Tite: We have a Flash-streaming server. All of our content automatically goes on our Flash-streaming server. And we put that on there for historical archive, basically. But everything that we post on our site is YouTube.


But we include the file name that's on our streaming server within the post, so if YouTube ever blows up, all we have to do in the template is say, 'Now use the streaming server.' Because we don't really trust that things aren't going to blow up one day, so we automatically go in and put in that file name.

But we just use it for an archive purpose. We use it for live events. I'm trying to think of anything else about the livestreaming. Now, we really like YouTube because, again, it aggregates all of our views as well, and it's more shareable and all that stuff.

Moderator: One more question.

Lacy Tite: Yes.

Audience 10: I really like how it's set up. How do you put the video there? How do you put the data in there? Is it the field or is it just search out the first video?

Lacy Tite: For any video stories, there's a custom field that's YouTube ID. What the template says is, 'If a YouTube ID exists...' And it makes that sound, too.



Lacy Tite: Cool.


Lacy Tite: Thank you!

If you have other questions, tweet me. Or, again, later on today I'll have loads of code and stuff up there.

Thank you! Oh yeah, it's crazy.