DonorsChoose.org is the leader in classroom giving in the U.S., they make it an easy, fun, and rewarding experience to find a classroom to support. We talk with Oliver Hurst-Hiller, the CTO about how they have made breakthroughs in the way they use transactional emails to engage and reengage donors. Transactional emails – emails that are triggered automatically through the site (think ‘you have a new comment’ or ‘welcome to the site’ emails)- account for 10x the amount of donations that are driven from scheduled eblasts from the marketing team.
Listen to how DonorsChoose.org thinks about creating a great experience for donors and approaches their technology with this mindset.
Here are some of the email tools and services mentioned in this podcast.
- sendwithus – simple transactional email setup
- Mandrill – mailchimp’s setup for transactional emails
Speaker 1: This is Using the Whole Whale a podcast that brings you stories of data and technology in a nonprofit world. This George Weiner, your host and the chief whaler of wholewhale.com. Thank you for joining us. I don’t think I’d be alone in saying that it’s often the small interactions of conversations that can really define an experience or even a product for you. I was recently flying from Reno to, to Scotland actually and had, a flight had been cancelled unfortunately and I needed to get on a United Airlines flight. It was like ten minutes before it boarded and I was like I’ve got a snowball’s chance in hell basically to get on this and wouldn’t you know but the attendant there was incredibly nice and actually helped me get on to that flight so I got to where I was going on time. She definitely didn’t need to go to the extent that she did to get me on the last seat of the plane but I’m definitely going to remember that the next time I’m think about United even though they tend to cancel flights on me, they’re really great with these small interactions. So why am I, why am I going on about this? This is episode 16 where we’re going to be talking with Donors Choose. The largest online platform for connecting teachers and donors, providing the resources to these teachers online. And one of the things I think they get really, really well is what are called transactional emails. Those transactional emails are not what like the marketing team puts out there as the official, if you think of like United saying “Hey we have a discount on Blah Blah Blah today.” They are the small interactions, right, with the attendants or the stewardesses that really define your experience with a product and that’s what Donors Choose gets. They’ve really worked on those small interactions that happen when somebody gives to a classroom. That teacher receives it, and then sends back a message to the person who donated. We’re going to be talking with the CTO of Donors Choose, Oliver, and figuring out the secret sauce that makes it happen on the technology side.
Speaker 1: Alright today we have got Oliver from Donors Choose. Oliver is the CTO of of course, Donors Choose an amazing organization. Can you tell us a little about yourself and of course Donors Choose?
Speaker 2: Well first of all, thanks for having me on. I’m psyched to be talking with you and your listeners about all this stuff, it’s, it’s really fun. So DonorsChoose.org it’s a website were public school teachers post their best ideas for classroom projects that their schools can’t otherwise afford and then folks on the internet help bring them to life with their credit card, and then we, sort of our operation fulfills the project beyond just forwarding a check. We actually, you know, buy all the stuff for the classroom and deliver it to the, directly teacher. And we facilitate kind of the feedback loop from the teacher in the classroom communicating back with their supporters and saying, “Thank you” and telling them how the whole project went. Kind of like Kickstarter for classrooms and we’ve been around now almost 14 years it was originally founded in 2000 by a public school teacher in the Bronx and he still runs the operation. His name is Charles Best. And yeah, it’s been, it’s been an exciting ride. I’ve been with the org. almost 8 years, and it’s, it’s, it’s, it’s, there’s been a lot of growth, we’re really excited about that. It’s taken a lot of hard work but you know when I joined the org. I want to say in 2006, that year we did 5 million dollars. That school year we did 5 million dollars in donations to classrooms through the website and this year, the school year that just ended, we did 63 million bucks in donations. Really exciting. Hundred thousand projects funded, more than three hundred thousand donors mobilized just this year, just this school year, so, so yeah we’re feeling really good about the impact, almost at 250 million dollars donated to classrooms through the website since inception. More than 50 percent of schools, of public schools in the United States have a teacher who’s used the website to post a project and there’s something like 95 thousand public schools in the country, so, we’re really excited about that reach and one of our big goals is to get all high-need public schools in the country having at least one teacher at that school using our website. So, we’re not all the way there yet but, but we’re working on it.
Speaker 1: So you got some room to go there. Those are some incredible numbers, some high level thoughts there. It’s as simple idea right? Like, teachers need equipment, teachers need support in the classroom. You brokered this way of people simply being able to identify that need and directly donate but of course the, the Devil’s in the details here and I’m curious. The next layer down, like the day to day, what kind of metrics are you looking at let’s say as the, you know, the technology side of things? What kind of metrics are you looking at that are really driving things forward?
SPEAKER 2: So, you know on the tech side, we probably look at a lot of things that maybe are typical to any web business, underlying things in terms of, you know, uptime and performance. And then I think across different channels for driving donations and for teacher success you know we, we look at different metrics. So you know, big things we look at, we actually divide the business into revenue from what we call marketplace giving, which are donors, citizen donors on the website. And then we also look at donations from partners, mostly corporate partners, and that’s about a fifty fifty split in terms of donation revenue. So we look at those separately, we have different features and products for both sides of those businesses and we watch them both pretty closely. We, we pay attention to dollars being applied to projects that, that tells us sort of the velocity of money through the system. We pay close attention to the number of cash giving citizen donors on the site. Because people can come to site and they can redeem a gift card, they can give in ways where they’re not opening their own wallet and, and that’s welcome and that’s awesome and, and many of those people were able to convert in to cash giving donors but for us a donor a who’s opened their own wallet to help support a classroom is, is sort of one of our fundamental customer metrics because that’s someone who, you know, there’s a lot of things you can do on the internet that require different amounts of commitment and hitting the “Like” button is one thing, and opening an email is another thing, and actually opening your wallet to do something is a pretty high bar, relatively speaking, so we pay attention closely to that. We, we look at year over year cash giving citizen donors as a way to measure our people coming back. It’s very hard to get people to come back to be honest. And I, And I think that’s true of e-commerce in general, most you know, for everybody who’s not Amazon, but so we’re, we’re really watching, you know, for the people that, that have discovered the site and made a donation. Are we getting them to come back? And we also, of course, are monitoring the teacher participation. Are teachers getting their projects funded? You know one thing we know is that if a teacher gets one project completed on the site, then they believe the website works and they become a net promoter, will then tell other teachers that this website works and, and may post more projects in the future. And if a teacher has no projects funded then they believe the website does not work and they’re not coming back and they’re not telling other people good things. So another thing we closely watch, in addition to sort of overall project success and dollars in the door, is whether new teachers using the system are succeeding.
Speaker 1: This is all about actual relationships and getting back to how the, the system delivers but I think one of the things that you guys are an absolute model for is, is something called transactional emails. Right? You shifted the tide, and made it easier for these relationships to blossom, so , if one of your team metrics is really making sure teachers comes back, making sure donors come back, can you tell us a little bit about how you approach this from a technology standpoint?
SPEAKER 2: You know I think that Charles would. Charles was a public school teacher and when he founded the org. His “Ah Ha!” moment wasn’t that teachers in high need communities, you know, needed help and you know could do more for their students if they, if they only had certain critical resources. It, his “Ah Ha!” moment, because that’s sort of not, everybody kind of knows that. The “Ah Ha!” moment was that if he could deliver an experience for citizen donors, people giving five bucks of ten bucks, or forty bucks, if he could deliver an experience for them that was akin to the philanthropic experience that big money donors have when they give a million dollars. In terms of choice and directness and transparency and communication and feedback. If he could deliver that experience for regular people then he thought it, it could sort of galvanize support around these classrooms and he was sort of inventing, or at least, he was kind of inventing crowd funding or at least he was the first one to reduce it to, to practice in a successful way. And so, at its cornerstone for us it, it is as you mention, it’s about the email communication because that’s the way to do this at scale. And, for us it’s about touch points throughout the project on a life cycle so I’ll give you a bunch examples. But really what it’s about, it’s, it’s about the teacher connecting with the donor and vice versa. You know, in a sense we want to get out of the way and just let that donor form a relationship with the classroom and it’s actually ok with us if the first two times someone gives they kind of don’t even realize that donorschoose.org exists and that we’re an org. and we have people and operational needs and we run this thing. It, you know, if they just feel like I’m connecting directly with a teacher and donating directly to them that’s fine with us because it’s that directness which we thing is really powerful. So I’ll give you some examples of, of, of critical emails along the life cycle. So one is as soon as you donate we tell the teacher and the teacher usually writes a thank you back to the donor within about 24 hours, that donation, that communication we hear from donors is sort of their first kind of mind blowing moment where they say “Wow”, you know, I guess I knew there was a teacher on the other end of this thing then they think “My gosh!” There’s that, it’s sort of like the relationship is then formed. When a project is fully funded, a teacher writes a note to all their donors, thanking them again. That, that’s a very powerful email. Once the teacher has the materials in their classroom, and they start doing the project they take photos of the project happening, they write an impact report, and we follow up and we bring that back to donors via email as well, notifications and there’s one analog part to the process which is handwritten thank you notes for donors who give fifty bucks or more and those, those are mailed through our office and we, we, we haven’t had, those are so sort of creative and amazing we haven’t had the heart to digitize them yet. We’re afraid that, that they might not be as powerful digitally. But we all, we do want to do some pilots because in their analog form their really compelling but we can’t deliver them to all donors so we, you know, just for, for costs and scale reasons so we have to have this fifty dollar cut off. And we are doing some experiments to figure out if we could digitize that, and then we could deliver it to 100% of donors and, and that might be, that, and we know that would be a win but the essence of all of it is that, you know, is creating the relationship, it’s the donor feeling direct connection with the classroom which is hearing the voice of the teacher through email through these pictures etcetera and we all do that with email notification, so if the teacher finds out about the donor giving and actually every donor when they give is encouraged, almost required to leave a message explaining why they give. So, if I gave because, you know, my aunt was a math teacher or they say, “I gave because I grew up in the same town and I went to that school and I want to give back,” and whatever the reason they explain, you know, I gave because I used play high school volleyball or whatever and that connection, and at often times the teacher will respond specifically to what that donor said and now there’s a rapport, now there’s a connection. And, you know, perhaps most significantly in terms of the business, you build the long term relationship and also a lot of the notifications bring you back to the website to say Hey, there’s photos for you to look at that the teacher has posted and of course you, you know, you have to view the photos on our website and now the donors back on our website and not to be crass but you know perhaps, while they’re there we’ll be able to entice them to give to the same classroom or to another classroom
Speaker 1: Yeah, I mean you guys are definitely what I would call, like, best in green for these transactional emails and building relationships and I think it, it started because you started with that seed of, I don’t want to make it easier for people to just give money to classrooms but you say I want you to experience, to feel like, like a big donor, like I’m the big deal, I spent money, I spent time doing this and so, while most of the time technology is like oh how easy can I make this donate button. You’re like, “that’s fine but what happens afterwards?” If we could maybe dig in a little bit more into the technology of running… running a place to allow all these transactional emails. I would love that.
Speaker 2: Yeah, sure! So we use 2 things, Send Grid and Exact Target. And Exact Target we use, that’s been a great tool and we use Exact Target to do outbound email marketing and we, to our donors and teachers. And we never send to people, you know, we don’t buy a list and send to the list so these emails are only going to our own donor base and our own teacher base and there are things that can be done sort of inmass so it’s announcing to all donors in California about a special match offer that’s available only to them or it’s announcing to all the teachers in New York that there’s a special campaign coming up and they can participate in a certain way or it’s reaching out to people saying, “Hey for mother’s day would you consider giving the gift of helping the classroom to your favorite mom?” ‘Favorite mom’ meaning your mom or maybe in my case my wife is a mom and all the moms; we want you to give to as many as you can.
Speaker 1: All the moms!
Speaker 2: And so that’s “ so the, Exact Target is sort of the outbound marketing engine there and it’s and it’s and the beauty of it’s, you know, it’s operated by a non-technical team member, it’s owned by the marketing team and, and technically we, my team does that. There’s relatively little to support it. The main thing that we have to do is to maintain a nightly feed of basically donor and teacher data into the Exact Target system and that’s containing, all of, all the attributes that our colleague on the marketing team needs to send; be able to do a blast email. So it’s everything from, you know, the donor name to maybe the, the data, the last time they gave to where they live etcetera etcetera. And then corresponding attribute for teachers and once that’s all in the system, then our marketing colleagues can kind of go nuts and they’re great, they do lots AV testing of the emails we send they’re all up in the metrics, you know, the opens and the clicks and of course the revenue as well because Exact Target pulls that all the way through and so they really have sort of their own kind of command center, and management panel and, and the AV testing we’ve learned a lot and in fact in corporate it’s some of the learnings from the AV testing back into our, what I call system emails or, or what other people call transactional emails. So that side of the house is sort of is different. The transactional emails we’re now doing through Send Grid. Send Grid is really sort of email as a service so it’s our application, our software our web system that generates all these emails we’ve built from scratch; a kind of templating engine that takes inputs and then can cranks out whatever is needed, a donation receipt.
Speaker 1: So this is the point where Oliver is explaining literally how they’re creating these templates so that he can create customized, a bit more, you know, better looking emails that will be triggered by interactions on the site, when somebody makes a donation, when the teacher comes back, and submits a thank you. This is the secret sauce, they care. Right? They started off, right, with Charles Best’s vision of let’s really actually focus on the user experience end to end, what is it like? What if this person instead of giving ten dollars, gave a million dollars, how would we want to treat them in even the smallest communication we send. And I think the results really speak for themselves as we hear Oliver describe some of the successes they had but you know it takes a little bit of time to, to program that in but you can see at least for Donors Choose. This was time well spent.
SPEAKER 2: Alright so, so that’s all coming through Send Grid so it, and it’s our application that’s doing it all so the system will send out a notification to a teacher that a donor has given, we’ll send notification to a donor that a teacher has responded and so forth. And that’s all done internally through our own templating engine, we have a big java app that runs our whole web system and it’s doing that. So, Send Grid’s been great in terms of getting the email service and delivery part out of our backyards then we no longer have to worry about we can pay them to, to be experts in it. They do stuff like signing of the emails which helps delivery they can provide opening and click stats. But it’s still a little frustrating for our team that we can’t AV test what comes out of the system emails and we also can’t have non-technical team members make changes. I want to say that our system emails drove on the order of something like 12 million dollars in donations this year.
Speaker 1: Oh God!
SPEAKER 2: And we, we know from the Exact Target front where I think we’ve driving on the order of one and half million dollars of donations we know that when can AV test and we can make the content management, not gated on, tech engineering bandwidth we know that we can make improvements that, that will increase giving and help teachers so it’s been frustrating to us that we don’t that level, that we don’t have that for system emails. So it’s something that we’ve been looking closely at and trying to figure out how to solve. And like I said that Send Grid was one piece of the puzzle and another thing we’re now looking at closely is AV testing. We’re looking at how; a shop called Send With Us that is integrated with Send Grid and our hope is that they may be able to help us on the templating front and on the AV testing front.
SPEAKER 1: Yeah! Definitely the AV testing of transaction is difficult. Can you give, or share any findings that you know have maybe come from the marketing side or have been changes in small ways that has surprised you by a huge up side?
Speaker 1: You know… I don’t. I don’t have the numbers right at the tip of my tongue but what I can tell you is that just religiously with the stuff that we send through Exact Target, the outbound marketing emails. For those we’re very diligent about sending to a small sample group, testing a bunch of different subject lines and almost always the subject line you think is going to perform best, you’re wrong about. And the test helps you figure out what people respond best to and there’s just no question I mean, we, we moved the needle on open rates and on click throughs and on revenue all the time with, with the AV testing we do through the Exact Target tool which makes it easy, and this is how, this is how all real businesses on the web are running, you know, they AV test what’s on their front end and they’re AV testing what’s in their email and not to get, not to get all, you know all talking about business opportunities but it’s really an untapped business opportunity that the big boys have all built their own frameworks and on the front end of the website organizations like Optimizely, which by the way we discovered thanks to, to you when you introduced us to Optimizely because you had so much success with it at dosomething.org and since then so essentially AV testing as a service as a framework is now available to folks to the front end of the website is almost really a turnkey Whizzy Wig.
SPEAKER 1: Yeah!
SPEAKER 2: Proposition and but it hasn’t yet reached system email and, and it just we know it can unlock so many donations so there’s a couple players out there and we’re about to pilot Send With Us and it’s not a trivial change because sending system emails is sort of built into the very fabric of our web application but for us we think it’s worth the experiment. Because if we could let our marketing team and content experts loose on this, what is today you know a twelve million dollar source of donations we really think, we could help a lot more classrooms.
SPEAKER 1: It’s great because I think a lot organizations aren’t even at the level of What is this transactional email? You know isn’t the automatic thing that the tech guy just, you know, typed in the first day he launched the site which kind of kills me but getting into the easy testing is where we’ll hopefully be as a factor shortly. As you move, like, kind of wrap up here I’m curious, what is, what is on your plate for like the next six months? Tech Speaking.
Speaker 2: Well one of the things, that, that has been successful for us recently and that we’re going to invest more on over the next 6 months is, in the area of system email our, the recommendations that we send out to our donor base, so we have an email that goes out every month and it suggests projects to donors, and, and this is part of the whole thing of getting people to come back which like I said is hard to do, I mean we survey our donors and they say amazing things about us and they’re net promoters they love what we do they had an amazing experience then we, between, you know we’ll look at the data and this donor hasn’t come back in a while and I think that’s, it’s not that they’ve done anything wrong it’s just that people are busy, life is busy, it’s hard to get people, you know, we’re never going to be one of those websites where, like the New York times or your bank or your email where you check that every day religiously so, so how do you bring back and how do you send them, you know, without being spammy; nutritious reasons to come back. So we’ve been working on email notifications, sorry, recommendations via email for some time actually we had the feature for many years. It was a feature that was doing about seventeen grand a month until October of 2013 and we came out with yet another attempt to make this email better targeted and smarter and, and now it seems to be really working the emails are doing like 150-200 thousand dollars a month.
Speaker 1: Holy Cow!
Speaker 2: and it, and you know it really was just about experimentation it was probably our 4th or fifth attempt to do it right and, and at its heart what it does is, basically look at what were the searches that a donor did on our website before they funded their last project. So basically, I want to fund projects in New York, focused on music at, high need schools we’ll then do that same search for them automatically pull a project that meets that criteria and then email it to them once a month. And if, if they’ve given multiple donations with different search criteria we’ll include sort of one project recommendation for each and also if we know there their zip code, which we do for most people, we’ll also include a project that’s very close to them that, that is a high need project and it’s, we’ve been sort of playing with it and this is working well and so you know big area for investment for us is, is like “My Gosh! We finally have something that’s working in terms of recommending projects and getting donors to come back” so I think a big area of investment for just going to be doubling down and, on that product and part of that will be sort of very micro tactical fine tuning of language and you know making sure we’re presenting their recommendations as best we can and then, and then maybe some more macro things like are there more people we can target are we targeting on all the things we can and so forth. But, you know sometimes the least sexy thing, the sexy thing to do is to sort of invent a new product but actually the disciplined thing to do is to take something that’s working and see if you can just double down on it and make it work better so I, so I think that we’ll be doing that with these email recommendations.
SPEAKER 1: So the, the change in October, the big change was increasing the accuracy based on prior search results and focusing the email or was there like more nuance to it than that?
Speaker 2: The essence of it was recommendations based on specifically the projects you donated to in the past but based on the search that led to them. Because, not to get too far into- but one of the issues that we always had is, if someone gives to one project the project may have thirty attributes and it was hard for us to infer what was the attribute that made you give, and therefore what are we targeted on. Was it because it was in New York? or was it because it was music? or was it because you went to that high school? or was it because a friend told you about it? or was it because the age group of the kids…and it could be any of these things and with only one data point we’re not amazon with like a hundred data points.
Speaker 1: Yeah!
SPEAKER 2: So, with only one project, how do we know which dimension it was that motivated you? So, the “Ah Ha!” for us was that looking at the search you did before you donated, because it’s through search that you express the criteria that was meaningful to you. And then it’s that criteria plus a zip code targeted project recommendation.
Speaker 1: Oliver, I love all of this. I love the heck out of it, and you know I think you’re definitely headed, your train is on the right track with doubling down on what’s working and doing more of it would work certainly. How do people find you on the interwebs?
SPEAKER 2: DonorsChoose.org is the place to go and if anybody ever wants to get in touch with me, and you know, follow up on anything of these things or hear more about how we done them, I’m [email protected] and yeah I think that’s it, I’m also on LinkedIn, I have a Twitter account but I’m not very active because I’m too busy trying to get these system emails figured out.
Speaker 1: Oh boy! Too busy filling people’s inboxes, well keep on, keep on doing amazing work over there. Thank you so much for your time today Oliver.
Speaker 2: You bet! This was fun, thanks George!
Speaker 1: Man I love Donors Choose, I’ve known Oliver for a while and I’m a huge a fan of their work. I encourage you to go there, donate to a teacher, classroom and just go through the experience of it. It’s brilliantly done. Learn from it, take notes and then apply it to what you could potentially be doing in your own organization. Two big things I want to call out here. First, that moment where he said it’s about having the discipline to double down on something that’s working but could work a bit better. It takes a bit of courage to do that. Those transactional emails are performing at 10 X, as he said, 10 X the rate that the marketing emails are at bringing people back to the site to donate again. This second big point here is customization. If you haven’t you realized by now, the web, what, 2.0, web 3.0, will all be about better and better customization you’re going to be able to tell my heart rate when I’m about to buy something in 10 years, right? So how they’re starting to do that with you know somewhat limited resources was they found the right attributes, what was the last thing I searched for? Searches are often times a fantastic indicator of intention. And then of course location,right? Something they can pull up and then that triggers an email that can be sent to somebody, saying Hey We found technology projects being put up by teachers in the Brooklyn area that’s something that will get me to donate. We covered a lot of resources today and they will all, as always, be available at wholewhale.com/podcasts. Thanks so much for joining us.
This has been using the whole whale, for more resources on today’s show please visit wholewhale.com/podcast and consider following us on twitter @wholewhale and thanks for joining us.
And of course special thanks to Scrambled Mache for our music today. Coming out of Michigan, the song that we used, the title was “Music Teacher.”