Session Transcript:
New phone, who dis: Building intimate audience relationships without the creep factor

Session facilitator(s): Sam Ward, Hannah Young

Day & Time: Friday, 2:30-3:45pm

Room: Thomas Swain

SAM: This is about intimate relationships so let’s get into groups. You might need computers.

HANNAH: Unless you want to split up because you became together.

AUDIENCE: Yeah, we should split up. I’m sick of this guy.

HANNAH: All right. Maybe we’ll give folks a minute or two to trickle in… all right so do I need this? Do I project? Is this good? Can you hear? Awesome. So thanks for joining us on a Friday after-lunch session. But we’re going to talk about SMS, we love it. And I think you love it, too if you’re in this session. I’m Hannah, and I work for Reveal, Center for Investigative Reporting. We’ve been 40 years in the field.

SAM: Four, zero.

HANNAH: And this is Sam And we wanted to pitch a panel and hear from you guys, SMS, what you’ve tried, because we’ve kind of dove into it in the last year, too, and love it. And there’s just a lot of things to consider and so we wanted to take some really concentrated nerd time to talk about about what we thought worked well, and more selfishly, hear from you, things that you might have tried, or things that you haven’t tried that you’re curious to try. And so we’ll be kind of running through some stuff in this presentation but spending some time on an activity and taking some time to talk about what worked, and what didn’t in that context. So if you haven’t done SMS before, you’ll have all the context after this to get started, I hope. So to that point, who in the room has done things with SMS? Like, tried to do like a bot in your newsroom? Have you done it in your job? Is it something that you have done professionally? Who has been on the other side — a user of SMS, you’ve texted a number that you’ve seen on a bus or a baseball game. Who’s had a really good experience with SMS? Who’s had a really creepy experience with SMS. Okay. That’s good. And any other surveying questions? And in terms of, like, who’s in the room, too, who identifies as like a journalist, reporter, editor, I’m on the editorial side. Product? Someone who’s…? Audience and engagement?

SAM: Kind of a third, and a third and a third?

HANNAH: Do we have any development folks, membership folks in the room? Cool because that’s one of the areas that I want to open this up. Great. So like I said, I’m just going to run through some things quickly before we begin this. So how do we get to the…? So if you have your phone on you, this will be a time that you are actually encouraged to take out your phone. We wanted to just do a round of show-and-tell. We wanted to show take away that we’re using SMS but these are the ways in which we have used it and been successful in the last year. So you can give it a list.

So from the Center for Investigative reporting, I’m Alex. It was more than 50 years ago.

Dr. Martin Luther King has been shot to death and met with devastation.

There was civil unrest from Washington D.C. to Detroit.

Of course, all America is outraged at the assassination of an outstanding negro leader.

Lyndon B Johnson, the president at the time was outraged.

And indeed this bill has had a long and stawny trip.

Congress held up the Fair Housing Act for two years, and even after King’s death, before law makers wanted to keep up segregation.

No legislation should be passed as a memorial to anyone.

But King’s death finally gave Johnson the leverage that he needed to get it passed.

I do not exaggerate when I say that the proudest moments of my presidency have been times such as this. When I have signed, into the law, the promises of a century.

Housing discrimination was totally legal until the ’60s. Foyer Housing Act outlawed it, but that law didn’t solve the whole problem.

HANNAH: So later in the show, we just did a realization that we didn’t put this at the top of the show. So you heard at the top of the show, Al is introducing this concept of what we’re going to be talking about in a radio show podcast, and it’s around modern-day Redlining and around that segment, probably ten minutes in, you’ll hear a prompt and Al will read in text HOME to that number, and as a listener, you would get that prompt from Al. So if you want to go back to the presentation. And you guys can try it out if you want and you’ll see what happens after that. So a listener is listening to the show, they hear a prompt from Al, or one of our reporters, and then they text that prompt to the number, and we start sending them things.

AUDIENCE: Can you put the number back up for a second?

HANNAH: It’s actually (202)-873-8325. And the number you word you text is “HOME.” All right. So as you could kind of hear, you hear Al setting it up in the show, and you’re kind of getting drawn into the story of modern-day Redlining and disparities and then you get this prompt: to see what it looks like where you live, text this word to this number. So one of the ways that we’ve used SMS — and we’ve primarily used SMS as a companion to our audio. We have a really successful podcast, a radio show that goes out on 450 stations across the country and we were really curious about how we can’t connect with that audience in a secondary way. So we have listeners but, you know, with listeners — does anyone here do audio journalism or work in a radio station or Apple Podcasts? In those mediums, that’s your primary connection to folks, and you want to know more about them and things like that, you thought Nielsen data, that’s like a delay by six months. With podcast data, you don’t really get anything, you only really get download. And so we were looking to build a relationship with an audience and we thought of SMS as a way to do that because maybe you’re already podcasting, you’re maybe already on your phone, as you’re listening, you’re able to text, so we thought this might be a good fit, and as we started doing user testing and user research, it seems like it really is. So these are some of the examples that we think that it has been really effective. And to and so we do things that we wouldn’t do on other channels and other stories of Al. That’s Al. People will text, hey, Al, and people send selfies back and it’s been really fun what people kind of unprompted and so they’ll send back a photo of themselves and say like, “Hey, Al — immaterial in the airport, I’m a lawyer.” You wouldn’t see that on Facebook, you wouldn’t see that on Twitter. And you certainly wouldn’t be able to do that on any other we be able to or platform. You probably got a text if you texted him that looks like the second example that’s starting to give you context around lending disparities in your neighborhood. So you text an address and you get a snapshot, hopefully, of what lending looks like where you live. So personalized information.

AUDIENCE: Sorry. I came in late if I missed this, but how conversational is the tool you’re using, when they send back unprompted stuff, is the reporter able to write back free form?

HANNAH: Pretty janky at this point, real talk.

SAM: Basically we have a catchall in place. So it’s like if you receive a string that you haven’t planned for, you can just give a canned response. But the tool we have, you can also plug in to some natural-language processing tools so you can be a little more sophisticated about how you handle those but we’re just getting to the point where we’re even trying to figure out what our strategy might be around that.

HANNAH: Yeah, one of the ways that we’ve been doing this is by doing the prompts in the show, we’re also kind of gaming the programming you would have to do if you were trying to do more chatty stuff. So it’s like we give them what we text so that we can give them a reliable return basically.

SAM: Trying to give constraints.

HANNAH: So text home but variations how they might spell that. That the user would reliably be able to text and deliver a result them. I mean, that’s a dream, what we’re aiming for, but, I mean, getting started with with it, definitely out there. So you can get some pretty odd responses. We try to make them not sound institutional but yeah. Did that answer?

AUDIENCE: Yeah, absolutely, thanks.

HANNAH: And has anybody who’s done SMS done successful programming out that’s natural looking at the GroundSource folks? If you don’t know GroundSource, it’s a great way to do this kind of plug-and-play and these two raising their hands are kind of our plants in the audience. Any…?

ANDREW: How much time I have? It’s finding this be balance between the institutional voice and the conversational voice. And if there’s any institution involved it’s hard to get that conversational tone intact. Emoji, it’s like eh… do we send emojis?

HANNAH: Like, what’s our policy?

ANDREW: Honestly, it’s important.

AMANDA: I worked on a Facebook Messenger bot, but, similarly, one-on-one conversations that were kind of broadcast. And I was at BuzzFeed so the tone was well established and also was appropriate-to-medium already but we worked really closely with the social media team to vet that kind of work while we were assembling it because it needed — because it was anonymous, it needed to come from BuzzFeed News, it wasn’t coming from an individual reporter on the ground.

SAM: Yeah, and that’s something that we’ve experimented with over the course of the last year or so. Like, that very first one, we try to make it clearly come from the voice of Al, but some of these other ones where it’s maybe from a different reporter — there was one where we had Anna Sail that had a guest episode that went along with it and we tried to write it in her voice, and used her as the voice of exchange. And it’s interesting. Sometimes that works really well, and sometimes it’s less appropriate depending on the content that you’re trying to deliver.

HANNAH: Yeah, so in these examples, the things that have — the examples that we gave were SMS, and how we’ve used it, and been excited about it, and in building these relationships. So a relationship with Al, or a relationship with a reporter. To Sam’s point, letting those people have a bit of a voice and texting back, rather than being like, this is in review. It’s always manufacturer exciting to have it come from a person because that’s who you get texts from in your everyday life. And second, giving people information that’s really personal to them. SMS is really valuable real estate for people. You don’t really get spammed the way that you do on email on SMS. So making sure that we’re actually providing value when we’re sending things. And the other thing that has been really helpful, or exciting with SMS is it’s been a really effective way for us to open up a dialogue so that we can hear back from people and it’s actually, like, engaged with our listeners and give them a space to write to us, again, and more importantly ask us questions of things they don’t understand and things they wish they knew. So with that regard, it’s opened up new reporting products for us.

So in this third example, we did a show on immigration — immigration policy — and it was back in October of last year. And it was a really wonky show. Like, we’re the Center for Investigative Reporting, so it’s kind of a given. So we go really deep into these topics. But this one, for us, so policy heavy and confusing, I think, for a lot of our listeners because we got a lot of questions back, when we asked, “What do you wish you still knew about immigration.” We got feedback about lack of understanding about how things work, how things used to work, what steps people could take to make change and things like that. And so we reported out an entire show and dedicated a whole hour to those radio questions, which is really important for us, because that’s a big investment for a company to make. We’ll do blogs and stuff like that. But to take an entire team of reporters and our entire radio show for a week and just answer questions that people had was I think really a credit for us getting back. So for that show I think we got 450 questions for the Redlining show, and then for the top one, we got over 7,000. So does anyone have examples for using SMS for the folks who raised their hand. Does anyone want to share a time that it worked really, really well?

AUDIENCE: I work for the Post and we did a messenger bot called Gillsbot that shared emotions for the election. And it lasted 30 days. It feels so long ago. And I did the language for that. And the unprompted stuff was what we had to pivot for the project because we just didn’t plan for it. And the nice thing with messenger is that there’s a lot of weird quirks with messenger is we can just write back straight on Facebook and surprise the heck out of somebody if they were trying to test the bot or do whatever. And I don’t know… that was funny. But yeah, it did well. It was smallish. Like, it did scale incredibly well, and I think it was because it was over a 30-ish day period but it was also pretty resource heavy and yeah.

HANNAH: How many people worked on it if you don’t mind me asking?

AUDIENCE: It was me and one other person from the newsroom, one person from product, and then an engineer. So when I say resource heavy, I think it’s just ‘cause that’s all we were doing for the 30 days. Pretty much 80% of our day was there. And it was daily check-ins and, like, writing prompts for that. So yeah.

HANNAH: Cool.

SAM: Is there anyone else?

HANNAH: Anyone have one that, like, massively failed? Those are also really interesting. We had a bunch fail before we got to the interesting that were really good.

BEENA: I did one earlier this month for ProPublica for first responders and the PTSD communities. So we went on the ground and went to all these different Facebook groups and resourced every group that I could asking them to send us their questions and we would do a story afterward asking anything that they wanted to know. This was also pegged to the release of a story and so we were trying to do some more engagement around it. I don’t think I got, like, a single question texted to me using this. And so perhaps it was like my own language that was off in some cases I was, like, messaging it to sources, messaging the number to sources that I wasn’t friends with so there’s a chance that they might just have seen the message so yeah.

HANNAH: We haven’t actually seen good pickup where we do digital messaging. So with Redlining, we tried it, and radio, and podcast, we have really large audiences but then, like, we embedded the number on a story and put it on social.

SAM: And a reporter, it was literally, you mention the number and stuff like that. And still the audio ones would outperform the digital ones.

HANNAH: And yeah, maybe 10:1 or even better return. So I think there is something about audio that is helpful that even if you could figure out — like, on interviews or something or do a short video and read the number or something, but I do think that there’s something to hearing-it-calling-in behavior that’s worked well for us.

AUDIENCE: Yeah, it’s that intake. It’s like we’ve had customers who are, like, they think that they’re not getting enough responses from their call-outs on digital on their form, so they’ll put, like, an — they’ll text us instead because the perception is, oh, people want to text us. Because it’s the call-outs that matter because distributing social is the number one factor. It’s crazy, yeah.

AUDIENCE: You said used different keywords to track where they were coming from?

SAM: Yeah, exactly.

HANNAH: And we’ve now switched to be even more clever…

[ Laughter ]

AUDIENCE: Tell me more!

HANNAH: And I should just pat ourselves… we changed it — because we have reliable distribution channels where we know where we’re going to distribute this. We’ve set up a number specifically for our broadcast audience, specifically for our podcast audience, and then for all the other channels so that we can see where the channels people are coming through. And the analytics tough is set up so you can easily track return users and things like that is something you do want to be thinking through it. We weren’t thinking through it. We’re like, “We’re testing!” And four tests in, it was like, how many people are return versus a new visit.

SAM: And it’s like, we’re not measuring that!

HANNAH: So it’s doing that — and we can talk in exhaustive amounts about metrics so I’m happy to do that.

AUDIENCE: So that was actually my question: if you could talk a bit more about the analytics that you used, and you started talking about wanting to know more about your audience and what you wanted to know beyond the anecdote, what kind of data you can get.

SAM: It varies on the campaign that we’ve been running. So the tool that we get, we get their phone number so we know reasonably have a good idea of where they’re interacting with us from. We can set variables on each interaction. So now that we realize we want to be able to count how many times they come back, we’re tracking that. Tracking repeat visits. For the Redlining one, people were entering a physical address. So we could — we could add that in.

HANNAH: Which, historically, our lawyer would be very mad about that, so we don’t store those pieces of data together.

SAM: One of the conversion things we ask is asking people to sign up for our newsletter. And once we do that, we’ve actually created a segment in our newsletter so that we can track the behavior of these people. So, if A, they’re podcast subscribers and they’re using the texting tool, does that group behave differently from any other groups over time? So those are the things that we’re trying to look at over time.

HANNAH: And did you have a second part? You said something about learning about our audience.

AUDIENCE: Also being in radio, you also want to get a sense of the numbers, and you’re finding more people to get more data with these tools.

HANNAH: We haven’t used any of these — we call them conversion asks — it sounds clinical, but does that term mean anything? When you ask someone to do something basically for you. There’s room for one of those per time that you’re engaging with someone so you want to be clear for what that is. So if you ask someone to sign up for a newsletter, that’s the only click you get, it’s not like sign up for our newsletter, and Facebook, and donate. Everyone in your newsroom you to put six. But be protective of that, do one. We haven’t had the real estate to do a survey. But one of the things, internally, we have this — we call them our reveal insiders testing group. We source them through, like, all our channels and it’s like sign up to test products for reveal and we were thinking of doing like a texting club or something similar which is sign up to get text updates from reveal. Figure out what a reasonable — something that would be valuable to that group. We haven’t figured out what be that would be and so we haven’t launched anything yet. But I could see joining that group and surveying being the first — maybe not the first, you would probably want to wait, but you could do more regular surveys, I think, with a group like that. I don’t know if anyone runs a newsletter, but kind of like with a drip campaign once they’ve received four newsletters, you can ask people to fill out surveys, and you can get more audience data that way and it’s really nice. I’m going to jump us right along so we can get out of the presentation part of this.

SAM: So why use SMS? Pretty much everyone has a smartphone or even a dumb-phone they can use SMS and it’s really easy to use. People engage with them and one thing that’s been really valuable to us is it’s not mitigated by an algorithm when we try and get our stories out on Facebook and other social products. If the algorithm gets between you and your audience, that can be bothersome, and we do find that people engage these text messages. In some of these campaigns where we have multiple call-outs during the show. Some people drop off — some people drop off after the first one, but over half of the people are getting all the way through them so that’s been exciting for us to see. And I think one of the things that I really appreciate about this channel is the intimacy of it. You’re staring at your phone in your hand, and it’s conversational, and it really does open up a dialogue. You get a message and you want to be able to respond back and keep it going. So those are one of the reasons that we’ve found SMS particularly engaging for the work that we’re using it for.

HANNAH: And to add onto this, but there’s like 90% open rates on this. We use MailChimp. 20% is looking good. 50% — that’s a killer newsletter. And with SMS, with great power comes great responsibility. You don’t want to spam people. But people look at their texts. So if you’re looking to reach people in a reliable way, not only are you going to have to worry about an algorithm but you have to worry about, they’re probably going to open at least one. They’ll probably get annoyed if it feels spammy, something that they didn’t sign up for, but they’ll open it. But this is usually where you’re texting with family and friends. So just by, again, being in that same space instead of email wherever you’re on whatever lists you got put on, you don’t get that on SMS because there’s more legal regulations. You can text people. You can’t do the same things with email. They’re also expensive. And on this sustainability one… this isn’t something that we’ve fully experimented on but one that we’re really curious about is could this be a pathway where you start building this audience funnel so that people who are texting you are actually donating on that same channel and just keeping it all in SMS and that can help offset some of the expensive costs of the platform. So we say: why not SMS? It’s expensive.

AUDIENCE: I was wondering, how do you ask for donations in SMS, or do you think — I mean, are we not there yet, would it break trust or something?

HANNAH: We have, but it’s clunky the way that we have — just with the the ways in which we take donations at CIR and so we have to send them a link to a website, where they can then donate which is another, like, three-step process. So until we iron out — you want to donate to Reveal, just say “yes” or something that keeps them just in SMS. We’ve limited the amount of asks but, you know, if you’re a non-profit, always be asking for money and so we’re trying to figure out how to make that…

SAM: Streamlined.

HANNAH: Make that happen. And I know other organizations do it but it’s just that we’re non-profit. So dealing with the non-profit tech sometimes. So cost is just a reason it can be hard for organizations and in particular, organizations with a bigger audience like the Post the Times, I think, had to stop using SMS in the Olympics. They have to kind of round it if I’m remembering right. Because if you have a billion rig audience, that path scales really fast. And people are protective of their text inbox. So just be careful. So if you aren’t really set up to think through what you’re really doing in SMS it’s probably not a really good marketing channel, and I would caution away from that just because people feel — whenever we do user feedback or deep user research, one of the things that makes it easy is you can call people because they’ve texted you. So you can instantly say, hey, you know, and thing on the phone and ask them about the experience on the phone, which is awesome.

And yeah, but people get really overwhelmed really fast. So you text them, like, three times in a row or five times without them sending you anything back, they’re going to be like, “Why are you blowing up my phone?” And it’s like I sent you five texts and if you think about it, if it were your mom or a friend, they have that same reaction and people have that same reaction even more with news. So you have to just be aware of that. And character limits. You don’t get a huge amount of room with text. So just, you know, you have to be really pithy, which is a really interesting challenge for investigative long form. So we’re taking, like, 3,000 stories and thinking about text messages.

SAM: How many characters does this document show?

HANNAH: And you have to add in, like, fields. So we’re going to do this, and then we’re going to kick right into our activity.

SAM: So basically we kind of covered these. So basically put if you have a companion audio piece that you want to send additional assets to, that’s one way to send SMS. The listening-post model where you’re surveying communities and getting their input on a particular topic. Feedback from communities. So when we first launched these SMS campaigns, we were just asking people for their feedback. Did this work for you? Is it interesting, would you use it again? That kind of thing. And the thing that we’re really interested in that comes along with this intimacy idea is that you’re building a relationship with them.

People really identify with the hosts of these radio shows and podcasts and things like that. So if they can feel like they’re having a relationship with that person, that’s all the better. So that’s why we found people were sending us their own selfies, and texting us, oh, I’m stranded in the airport, I don’t know what I’m going to do for the next three hours. There’s a door to conversation there and we’re interested in exploring that, and how we can tap into in order to strengthen our relationship with our audience. And I go, of course, the format is really good for conveying a personality, right? So because you’re not locked into this institutional format, you can have a little more fun and be more expressive with the conversation you’re having. So that’s another exciting area to explore.

HANNAH: [ Off The Record ] And prototype just a test SMS campaign. So we’re going to ask you to get into groups of three and we’re going to do this in ten-minute increments.

SAM: So there should be a stack of worksheets on each of the tables, and one thing that we ask people to think about is a story that they’re working on or a story that they’ve recently worked on that they think could be a good fit an SMS campaign to go along with. So take a second, think of the story that you want to work on, and we’ll take three minutes and just tell that story to each other, and in the group, you’re going to choose one to prototype.

HANNAH: And it could be a story idea. You’re like, we have assets that we want to send with this story. It could be a dataset. Just content that you think would work well for SMS. It can be like an engagement campaign or something where you want to get information from an audience so that it can inform your reporting. So there’s a lot of ways that it could bow. But, yeah, if you guys want to chat amongst each other, and you pick one, and fill out the worksheet based on that. So three people in a group, one story. Go.

[ Group Work ]

SAM: Okay. So if you’ve chosen your story to work on, now let’s move on to actually filling out the script for it. So think about the call and the response and also think about if you have any supporting assets that you want to be able to send to your user. You want to make sure to download those to your phone so you can just send them right back to that person.

[ Group Work ]

HANNAH: Okay, guys, we’re going to get back together in, like, two minutes. So maybe wrap up where you are in your conversations, and then maybe we’ll do which have a little debrief.

[ Group Work ]

All right. Did everyone get through the activity what we thought would take ten minutes? We wanted to call that out, And we wanted to give you time because it seemed everyone was jamming on the scripting and getting their heads arounds what they wanted to try. We wanted to spend time to talk about that because it’s kind of exciting to talk to people who were interested in it. But we aren’t Task Masters, but we would have had you do user testing with especially… so getting it pretty much immediately onto a phone and into SMS form and, like, simulating that experience of what it’s going to be like in is really important ‘cause if you’re, like, in a Google Doc or these forums, you’re going to write in that natural medium. It’s, like, three sentences on a Google Doc, it looks small; on a phone, it looks really long. So it’s a good way not to waste too much time when you’re doing these scripting exercises with SMS. So when you’re back at your organizations or when you’re thinking it through, I would just say, try to get yourself to that point, maybe do an outline, or do scripting, but get pretty quickly to that point or what you’ve written out into a text form, where you can send it to the editors, send it to your reporters, and get a feel of what that’s going to be like. And the really important part: you guys all know this is then taking that feedback and applying it to your product. So things are going to change a lot. And, for us, we changed our text complaints probably up until almost the moment that they’re ready to send. So it’s pretty easy to do, but yeah, there’s a lot of pressure on them, and so know that they’ll change a lot, and you’ll think of something really brilliant and you would say, that would be the perfect image, or that would be the perfect thing to send with it. So it’s a fun thing to play around with, but take the time to do it.

So we want to do — this is not an advanced thing — help! We want to hear from you guys because you guys were kind of chatting out. So if you’re design thinking fans, or New School, or people from Stanford, you’re going to be familiar, what worked well, what didn’t work well, questions that you had and anything that you wanted to say. Did anyone have an idea that completely changed. Did anyone have an enlightened moment where you’re like, “This is what I actually want to do!”

[ Sneeze ]

AUDIENCE: Gesundheit. We thought about school newspapers and because we didn’t have a specific news story to think about, we thought about what would be especially useful to students in a university environment and pretty quickly it shifted to a class registration bot because class registration is pretty traditionally a horrible process for most students. So basically a kind of a bot that — and I think they’re actually building it.

AUDIENCE: Everyone can text SRCCON CLASSES. But you can experience this thing.

AUDIENCE: But it’s a bot that basically establishes what your major is, and then how available you are, and then you can search by day for your major and then you would link directly. Totally awesome. I wish I had that when I was in school.

SAM: Waiting in line at the registrar’s office. You get to the front of the line — oh, that’s closed.

AL: Precisely.

So we were focused on the Chicago Housing Authority help people live in housing complexes report out issues like the elevator is broken and this is the effect that it had on my life and also to increase the accountability. So the mayor, so the city council knows what’s going on in these housing complexes; there’s not a lot of transparency. So we started talking about just the context for why people living in these complexes would actually want to text out, and whether they would actually be able to see the connection between them sharing their voice, and actually something getting done. ‘Cause it — a bunch of journalists coming to ask you via text message doesn’t necessarily (Andrew) necessarily add up to immediate change, maybe what you want. So we talked about that a little bit and had some thoughts around the voice of it being very straightforward. What are the words we used?

AUDIENCE: Set up a system rather than promising that a reporter would get back to them right away. Like, we would introduce the bot as a person. Like, this is so-and-so from the newspaper, I set up this system, too, so we’re not overpromising what they’re going to get back right away. But we do mention the goal of the bot, which is to get this information into, in front of change makers, decision makers.

ANDREW: And we had four menu options which were… what were they?

AUDIENCE: Right, did we change any of these?

ANDREW: So learn more about my building. So we’d have one branch would actually have to find a database for all this. But all the key contact information for the landlord, the property manager, what their rights are, if something has gone unsolved for a period of time, what their rights are. Asking them questions to share a story. But that felt too much like, “My toilet is broken and I need it fixed tonight.” And we couldn’t promise that. I wanted them to get a picture of a cat or something, just to lighten things up a little bit, or something that’s not so much like, “What’s your problem? Tell us your!” Seriously, what’s your problem? Which it can be that way. And texting can keep it conversational. That felt, too, a little out of bounds but something along those lines, maybe. But yeah that’s as far as we got.

HANNAH: We don’t want a rain of emojis. We don’t want like a frowny face when it’s way more serious than a frowny face. So I want to hear — oh, yeah.

AUDIENCE: So we were tackling a podcast that I’m working on for my news component of my newspaper and it’s a murder — unsolved murders podcasts for local unsolved murders. And at first I wasn’t sure how to make the campaign because it’s about murders but we thought that it’s not only about the murders, but it’s about the location, and how the location has changed. In particular, the case that we’re reporting on right now, the older gentleman is a staunch conservationalist, and so he fought a lot of this area called Groton in Connecticut. And so we started going through this process of trying to get people to share their memories of the area, or pictures of the area or the murder occurred at a Christmas tree farm where this man ran. You know, where people — you got your Christmas tree from there. And Amanda, at some point thought, maybe we would ask them to share their memories with a voice message that we could then incorporate into a bonus episode or whatever.

HANNAH: And even if it’s not a voice mail, you can call them because it’s a number.

AMANDA: I do want to sort of throw out, that what I thought was classic before. As we were talking, I was like, actually we did a lot, a lot, a lot of experiments with this. So I want to throw out if you go to buzzfeed.com/openlab, you’ll find a little bit, but I think it might be a useful resource.

AUDIENCE: The one for the News Lab is useful. What was that URL again?

AMANDA: Buzzfeed.com/openlab.

HANNAH: And that’s in the stenographer’s notes, yeah? I know we want to spend more time on this but we owe you guys to wrap up. Is he selfishly did this session but we just want to know new stuff about this, and if you ran into anything that’s exciting to you, please email us, and chat with us after and all that stuff. We wanted to just get down to some of the brass tacks of what some of the ins and outs of if you have haven’t done SMS before. So Sam’s gonna run through these.

SAM: So just to be aware for budgeting, you have to pay for inbound and outbound if the user is sending, as well. And if you’re sending MMS messages, videos, images, things like that, those cost extra. You also have to pay for the phone number that you’re using, and if you’re using a shortcode, it gets super-duper expensive — like, $3,000 to $4,000 a quarter. So if you’re doing it at scale, those costs ramp up pretty quickly.

HANNAH: With SMS, long codes, though, people prefer to be texting those sometimes because they prefer texting a person and the shortcodes feel markety.

SAM: And we said earlier that you don’t always have an algorithm between you and your audience but in this case you actually have carriers between you and your audience and if you try — if you get into a situation where you’re sending a high volume of messages in a short period of time, they can actually block you and shut you down. So we basically found ourselves in that situation for the first time a week or so ago. It was only a temporary situation, but it just it was like what the hell is going on? Why aren’t our messages going through, and you have to get on support and everything, and figure out what’s going on. It’s not a fun time. But if you work with your provider and just have a clear plan for who you’re going to be communicating and the volume of messages that you’re going to be sending, you should be in a pretty good space. As far as the legal concerns, you have to — you have — people need to be able to opt out of your campaign. And you actually have to tell them they can opt out. So you have to have support for some reserved coordinates like STOP and HELP and things like that to be able to respond for people get the help they need. The length is 160 characters and sometimes you don’t even get 160 characters before your messages are broken apart which is extremely confusing for the user. So it’s a really good idea to test your campaign out before you go live. And when you’re doing that, sometimes it’s a good idea to put in a delay between messages because depending on the message size or the message length, the carriers can actually deliver them out of order. So not only is your message being broken up in a way that you don’t intend, it’s being sent out of order, in which case you’re being royally screwed and you have little control over it.

AUDIENCE: Is there a difference between different mobile operating systems and devices in these particular situations?

SAM: I don’t know if it’s due to the device that you’re using or more with how the carrier is sending the data, so AT&T, versus Verizon, versus whoever.

AUDIENCE: Or the Androids and iPhones, too. I just know that emojis take up a crazy amount of characters on Android.

AUDIENCE: It’s almost always the carrier.

AUDIENCE: It’s the carrier. Thanks. I didn’t know that.

SAM: Like Hannah mentioned before with your calls to action, we try to limit ourselves to one call to action is so that we’re not expanding to too many people. You know, like the title ever this talk, not to feel creepy or needed and to be as efficient as possible. In terms of measuring success, we would say, just be clear what you would want to accomplish with your campaign. For us, it’s the number of people that are interacting with it, how far through the experience are they getting? Are they coming back multiple times? And did they convert to whatever call-to-action we were giving them? And then we talked a little bit about voice and tone already. We tried to keep it playful, light, and succinct. But, again, that’s really going to matter — it’s going to vary based on what the content of your campaign is and the kind of information that you’re trying to deliver to people. The relationship that you have with them already. You know, our show, our host, Al, he’s a very strong character, very strong personality that comes across in the show. And so we try to emulate that in the SMS campaign. And so we felt that was a strength we had to play off of but that may not be the case. Like the Redlining case that we had, that one kind of veered away from some of the earlier tones and sounded more institutional because of the volume of information that we had to communicate to people and the subject matter, as well.

HANNAH: And then Bill asked this is something that just came up with that immigration slideshow I mentioned where I brought up the questions. But perpetually, you always have to be aware of who this is reaching and who it isn’t. And when you’re talking about the questions that you got and the responses you got, acknowledging that, and who it reached and who it didn’t because we had questions about immigration from our audience and those questions then, obviously, mapped to the demographics and behaviors and — yeah, personalities of the people that make up our audience. They aren’t representative of America’s population on immigration. And it was something that we had to talk about in the newsroom. So just being aware of how you’re — like, when you’re thinking about that information, it can be very tempting to be like, “These are everyone’s questions on X.” We certainly got caught up on it because it was like, there’s over 2,000 or whatever. But just a really nice flag all the time. I should just put it on a sticky on my desk because it’s something to constantly remind yourself of. We’re not reaching everybody. Whose voices are being reflected and being reached in the responses? And the legal stuff, the budgeting… that stuff is really detail oriented so if you want some guidance. We have lawyer on staff, and I know not everyone has that benefit. So we’re happy to share what she’s recommended. We’ve put in our text messages as best practices. We’re happy to share on budgeting, and GroundSource is an awesome resource. I swear this is not product placement but they’re great. But if you’re getting started, it could be really nice not to have to think through this stuff. And we’re happy to talk this in detail if you want. And we just wanted to say. But in terms of our key metrics, and I know this was a question earlier. We look at the rate of listeners that use Amplify which is our SMS product, so how many users listen versus text us, users and answering — it’s an interesting persona who would listen to a podcast and text, and sign up for another medium that’s three mediums that they’re crossing for us, and so we’re interested in seeing that behavior when it applies to membership and donating. So if you have any membership people in your organization, or if you’re interested in that we’re going to be tracking that in the — over the course of the probably next six months. The users answering engagement questions that we asked, and new versus returning users. And users donating directly —

SAM: Cha-ching.

HANNAH: So those are the things that we pay attention to. They’ll probably change as we get smarter about thinking about this stuff but if they’re helpful for you at all, we wanted to share those, too. I know we’re over time. If anyone wants to chat with us after, we can do that. If you want to email us, call us, like I said, we love talking about this stuff. So thank you for spending this time with us.

[ Applause ]