I was fortunate enough to have experiences in a number of industries before this. I was a pension actuary for a little while, that didn’t last too long. I was in finance for quite a while. I actually did some secondary education. I was in direct response television prior to joining Education Dynamics.
So, we’ll start with the first question: What is data analytics? Anyone want to take a stab? It’s okay.
Audience member: You analyze data.
Sandesh: You analyze data.
So I happened to be at a conference earlier this year. The Chief Analytic Officer’s Forum. I felt so important. [Laughter] And one of the things was, “Let’s talk about what is data analytics.” We had a very lively breakout session, and at the end of it, they came up with this definition:
“Data Analytics is the practice of utilizing statistical learning methodologies with complex data systems to enhance decision making.”
We were all so proud. I felt really good. Then we asked ourselves if that’s what we actually do. Very few of us actually raised our hands. That’s what we want to do. It’s not what we actually do.
So, I’m not going to give you a definition, but I’ll give you some views on data analytics, or business intelligence, data science. You guys have heard these terms quite a bit.
The first thing I want to mention is it’s not new. None of this is new. We’ve been using it in manufacturing, medicine. I think the place where most people see it is actually in sports. Right? You see a batting average, you see all sorts of stuff, right?
Second, the novelty of it comes from the fact that we now have almost ubiquitous access to some sort of information. Previously, you might have had paper applications, they got lost. You know, a storm came, and your warehouse was gone. You had calls that your people were making, and they’re not recorded. So they were all lost to history.
We now know because we live in a world of, kind of like Big Brother, that this information is available somewhere electronically. And that is why data analytics has started coming up. Because we can start looking at it for the first time.
So, what I’d like to say is instead of defining it, I’ll tell you what we try to do. So, later on at another breakout session at that conference I was at, we did talk about this. At its best what data analytics is doing is it’s connecting the people that do things, the processes you have to do things, and importantly, the outcomes that happen.
You guys all okay with that?
I was okay with it. It took me a long time to come up with it.
But that helps us understand the interrelationships between things. And helps us understand what’s actually happening with your teams, with your organizations.
But there is one very big important thing at the very end. You know, it’s all nice to know things, but it can help us iteratively test new marketing and enrollment tactics and strategies. And this is where the real value to an organization comes in. You know, it’s nice to know what is happening, but also because you’re tracking all this, you can try to do tests afterwards and understand if you’ve improved it or not. Sometimes you learn more from the not, but I’ll get to that later.
So, what’s the biggest challenge? If someone was going to tell you to do data analytics, what’s the hardest thing to do right now? Audience member: Keeping unbiased data. Sandesh: That is a big issue. Audience member: You don’t have a series of different pots of data that I’m not sure if they’re corrupted, I’m not sure if they’re timely or accurate. They’re all based on a traditional model, so sometimes it seems like what you’re tracking in a silly excel spreadsheet is not what you’re showing them as a system of record. Sandesh: Okay. Anyone else? Audience member: Maybe the data analyst or analytic people are communicating the results to people that maybe they shouldn’t. Sandesh: That is very important, and I’m going to get into an entire session on that. Audience member: Gathering and interpretation. Sandesh: Gathering and interpretation. Audience member: One of the biggest challenges I always have is redefining data elements. Sandesh: Redefining data. So, these aren’t all of them, and I’m not listing all of them. I just picked one, which–I tried to encompass it, which is the skills, knowledge, and ability to access, connect and look at the data in the first place. That’s why it’s such a hot market. And that is a major issue, and I’m going to completely gloss over this in the rest of my time. I promise you that. But that is the biggest challenge in using data analytics within an organization. There are ways to get to it, and there are lots of people that talk about that. But that is the biggest challenge. I’m aware of it, and I’m not going to talk about it again. I do believe an inquisitive person with a growth mindset can make a huge, huge impact in using data analytics. I hope that kind of person describes you. The first step is to try and get the data. I think many of you that have done this have understood that you get hooked at trying to get the data. Right? So let’s do some examples, and Dr. Richardson’s over here, so I’m going to use her first. It was really nice to see some real world examples that I could use instead of just making up my own. If you guys missed it, she had a wonderful presentation on the first day of measuring retention and persistence of adult students at Queen’s University of Charlotte. And what I liked about it, and I know this is an eye chart, but there was roughly, would you say, like five to seven pieces of data that you were collecting, manually, over a long period of time, and you were able to find some quite insightful learnings for your organization? At least six. You made some tactical changes. What I liked about the end was– there was an appetite for further advancement. What are the next steps? What are we going to continue to look at? So, I’m not going to go into your presentation. It was wonderful, and I encourage everyone to talk to her about what was done. But I like that fact that this was like, there was no system here, right? Your system was Excel. You had your assistants that tracked it over time. But that was enough to make a difference. Ms. Black. “Slouching (then nudging) towards funneldom.” I think it was four years you said, right? It was about four years that you guys worked on taking a process on your funnel, converting it to an actual funnel that people could, like, look at and understand how various metrics and various conversion rates actually help you. And that’s a reality. The data analytics and solutions aren’t fast in higher education. I get into the lifecycle of the sales process when that happens. And I’ll take a pause here for any questions so far. I am about to do the live demo. Cool, we’re all good. Alright, so I’m going to use the number one data analytics platform in the world right now. What is it? Audience member: Excel. Right, Excel. So, I was told specifically I’ve got to make this a little more accessible, so I thought, let me show you how to create an enrollment curve. Here is a link to the excel file.
I have a piece of information over here, thousands of records. There’s a random prospect ID. I added some random program level, you know, bachelor’s, associate, graduate. Is the planning date in which a prospect was delivered? You know how you receive the contact information, let’s assume, and here’s a time and date in which they enrolled. I assume most of you can try to get this data. So we’re going to make what’s called an enrollment curve, and it’s going to be really pretty. And then, later on, I’m going to show you how it can be really useful. So, an enrollment curve just looks at how long it takes for things to happen. So I’m just going to add something, a lapsed time field, at the very end here. I may go through this a little quickly. You can always reach out to me, and I’ll tell you how to do some of this stuff. In Excel, you can actually just subtract date and times, and it’ll give you the number of days in between two milestones. I don’t like the decimals, so over here I’m just going to round up for our purposes right now to make whole numbers. Okay. So this one took seven days to enroll, three days, six days, etc. Let’s just highlight all of this. How many of you guys have used pivot tables in Excel? Awesome! I usually have to tell people what that means. Let’s make a pivot table. In this pivot table, I’m going to make this a new field, lapsed time, and bring it down. I’m going to count the number of records. Again, I probably am going a little fast, so I apologize. So, this is data analytics right here. You’re looking at your data, and you’re like, why are there 55 people who enrolled before I got them? Right? But that’s what this is saying. So, let’s go back here. Because this is the elapsed time: the number of days between which you got them and when they enrolled. And I actually put these records in here on purpose. Because one of the things you do in data science is you look at these numbers and go “what is it in these numbers that caused this?” Is this a real issue in how we’re getting people? Maybe these are reactivated students from a year ago. Right? And you have to do something in the system to say, hey, they’re activated again.
The system is now–a day later it says, “Oh yeah, you exist.” Maybe. That might be an option. There might be something more hidden. But you start a conversation. I haven’t done anything, and that’s already a conversation. But let me ignore it for now because, well, we data analysts like to ignore things we don’t like. So, you have this thing where you see how many people are coming in every day. Let’s do something fancy here, and I’m just going to do percent growing total in, fancy stuff that tells you how to do a cumulative percentage to a hundred. So, in 337 days I’ve had all the enrollments in this data set, this is very ugly. Let’s do a chart. Let’s do a combo chart. Obviously, I’ve done it a few times so I can do this very quickly. This is actually one of the first things I did the first 30 days I started at Education Dynamics, which probably saved my job I assume. So now, here we have an enrollment curve. You have the number of days between a prospect being delivered and when they actually enroll. These blue lines are telling you your volume on each of those days. And this line is telling you what percentage. So, the percentages down here, 0 to 100, Excel likes to go a little above, but 100 percent. This is the number of the days. So, one of the things you would look at in a curve like this is how long it takes to get to about 80 percent of your conversions. So, that’s about 80, and it looks like it’s in the 45-day range.
I want to show you how we can use this in a case study that I was going to start talking about. Are we all cool so far? If you want more on the detailed steps on how to do that, just reach out to me. I can run through it very quickly with you one on one if you want. I mean, you saw it took me like, three minutes or something to do it. I don’t know if a lot of you guys know this. The core business at Education Dynamics– we own a number of websites. Gradschools.com come is one of them. We have several websites for going back to school, and we’re challenged every day how to garner the best prospects for our schools. The best thing is people that actually want to apply, enroll or start at a member institution. We have to do that, right? We spend millions of dollars in marketing every or month. So, over the last 3 years, we partnered with our schools to collect information such as, what actually happened to our prospects? How many of them actually applied or enrolled or started at the school? We now collect information from over 150 schools a month. Or roughly about 150 schools a month regularly share their information with which prospects engaged, in some form. And the first thing this has helped us do is to iteratively test and tweak our marketing tactics. Who are the people we should be bidding on? Who are the people we should add maybe one more funnel step to see if they have more of an intent? So, it’s all more on the marketing side, but you know, it’s worked for a lot of our schools. And this is something interesting we found. On the bottom here is the number of schools considered or chosen by a prospect on our website. So, they could have chosen just one school, two schools, three, four or five. This has indexed their performance, the likelihood of them enrolling or starting at one of these institutions. It turns out that if they choose 3, sometimes four but not often, schools they are 2.3 times more likely to enroll or start.
That’s amazing, right? People want choice, I thought I’d found something new, and then I Googled it. Have you guys ever heard of Sheena Iyengar here? She had a wonderful TED Talk on choice. I really thought I’d found something new, and I Googled it, and she’s like, “You know, back when I was a grad student at Stanford, we went to a grocery store, and we found out that if you give everyone 20 bottles of dressing or something like that, no one bought anything. But if we gave them like three or four, it was like 20 percent higher sales.” I really thought I’d found something right there. And then I read her book, and it was amazing, and that’s her TED talk, and I highly recommend it if you guys are interested. It’s less than 10 minutes long. Audience member: What’s her name? Sandesh: Sheena Iyengar. Audience member: Thank you. So, this school actually garners a lot of graduate leads actually. They kind of have some niche graduate programs. So, off the grad schools’ primary website we sent them about just under 11,000 inquiries for the year 2015. That’s actually pretty small for most of our schools but for GradSchools that’s actually a decent amount. Roughly 28 percent of those prospects only chose that school to be contacted by. But 70 percent chose that school and at least one other, and generally it was about two others. Okay? Very few people actually in this world want more than three choices. I guess we’re all lazy, right?When I had quotes for getting some work done on my home, I was like, “I need to get five quotes, but I think two’s enough.”And I got the third one because the third guy just showed up like five minutes after the second guy did. What’s important here is they got 20 enrollments, and in this case, the definition for the school is they were in the first class for at least seven days at 120 days later or 90 days later, I honestly don’t remember. That’s a .65 enrollment rate. Not very good, but very good for them because these are graduate students that are hard to get. Twenty-nine of these 7700 enrolled. So, they had a lower enrollment rate then. What’s amazing is 132 of these people enrolled with one of their competitors. One of the other schools that they chose to be contacted by. So, the likelihood was way higher, like 3.-something times higher over here, right, that someone would actually enroll or start, not even apply, to the school when they had a choice. So, my background is in mathematics, so I had to figure out what the expected value was, and there’s a way to figure out the expected value. It was somewhere between 58 and 63, so they probably lost out on about 9 to 14 students. That’s like an entire class, right? In graduate level. We’re also able to look at their application curve against their competitors. So, this took me three minutes to do. It’d take you three minutes to do. This green line is the maturation curve, the number of days, and that’s 0 to 100 percent. The green line is the school we’re talking about. The red line is the school that’s kicking its butt. Blue line is a little bit better. These two schools still have the majority of these 132 enrollments, and they are getting applications much faster. Actually, the 80 percent range is about the same for both of them, but they’re way faster in the beginning. I don’t know if that’s a contact strategy or not.
Green, again, is the school we’re talking about. Red is the one that’s kicking its butt. Blue is the one that’s, you know, generally the same for the first month but seems to just keep at it for a while. As I understand, grad students need a little time. We also find that this school doesn’t have very many starts. And this school has about three times as many starts. So, I don’t know if this implies this, or this implies this. I have no idea. Pretty sure though that this school is calling a lot faster. I don’t think the speed’s everything, I think contact strategy is actually as important. You have to be fast enough. You also have to have the right thing to say. That’s my own take on it. It’s not my area of expertise. I’m going to stop there for a second. Any questions so far? Audience member: Can you say who the grad school is? I’ll tell you this one’s on the west coast, this one’s in the northeast. This one is also on the northeastern side of the United States.
Pointers and Pitfalls
I’ll start with some pitfalls in my career in trying to use data and analytics effectively. The number one thing our people forget is you are not the student. I’ve had a conversation, I had a terrible conversation actually, saying well, “All these students doing this and this and this” and I’m like, “I don’t know that.” I don’t know why that prospect is doing what they aren’t doing, I’m just telling you how they are behaving given the information I hav. But I’m trying not to ascribe my intent to what I see. That is really hard for us as humans not to do that. Right? I think that is one of the biggest pitfalls we all fall into. You are not the data you’re looking at. You’re not the student you’re looking at.
“Data analysis and the way we’re looking at it is by definition retrospective.”
That’s an important thing to remember because it reflects how you operated in the past. It’s important to know how you were operating because that is telling in the data as much as the data is telling about you. It’s a two-way street. Everyone says that correlation is not causation. No one knows what that actually means, but what that actually means is that if you suspect a link, you have to figure out ways of testing that link. For example–. I’m going to use an example from Dr. Richardson’s talk on transfer students that have 30 credits or less are less likely to persist throughout your course. And I have no idea about how the course is done, but what I would suggest is to figure out is if the assumption is that people who have fewer than 30 credits have less experience in being in college, and so it’s harder.
One of the tests you can do that I would suggest is, well, what if we do a survey class that builds some skills for them? Maybe a month-long class. And then you tested those 30-credit people persistent or not. I suggest this test because there’s not a lot you can do about the people themselves, but what you do know is they are human beings that have started a class with you that are not persisting. Your choice can be to say, “Well, they’re not good enough because they don’t have the experience”, or you can say, “I’m going to test it because they’re already in my door.” You know? I just have to figure out how expensive it is to test that. You can do the test and still fail, you know, it’s true, but you’ve done the test. And that’s why I say if you suspect a link, you have to figure out ways of testing it. And that is a central tenet in data science. I get this a lot actually, and maybe it’s’ because I work with a lot of humanities majors: It’s easier to knock something down with verbal arguments based on theoretical possibilities of what may or may not happen. Actually, this happens all the time. I see it on CNN and Fox News all day long. –than with facts and figures. And this is just a reality of working with people that, and working with human beings that have emotional ties to what they’re doing and feeling threatened. But the number-one thing I’m going to talk about is watching out for cognitive biases. In higher education specifically, the life cycle of a consumer makes it ripe for cognitive biases. And what do I mean by that? Let’s look at the funnel. Everyone knows the funnel. You have some traffic that comes in. The traffic becomes inquiries. Inquiries become contacts. Contacts become prospects. Prospects become applications, hopefully. Applications become enrollments. Enrollments become starts. We usually stop around here, that’s why there’s a blue box there. Those starts become graduates. Those graduates find jobs. They become alumni. That’s your basic lifecycle. And I’m not going to get into conversation rates here. I’m going to talk about the time and lifecycle in each of these things. That’s seconds. Right? Traffic becoming inquiries generally involve seconds getting onto your website. That’s about minutes depending on how slow they are typing. Then should be minutes contacting them, but it could be hours. Hopefully, it’s not days. But generally, this step over here is weeks. And this step over here is months. And hopefully, you’ve got them for years. And they stick around for decades in some capacity. And you’ve got them hooked for life. Right? Let’s go back here. This is really–. These two here are the first two real things that everyone’s looking for right? Enrollments and butts in seats that are at least in class or–. That is weeks or months. What happens is it takes three or four months to fully measure the effectiveness of your tactics. It is very easy to ascribe meaning to either examples or worse yet, on memory. I literally have people at schools that say “Well, this must have worked because, you know, I did that mailing, or I did that ad in the Washington Post. That’s why it worked. Cause you know, what happened two months ago? You know? Miss XYZ was on maternity, I had those two interns, the interns worked on all the mailings. Yeah, that must be why.” Right? We do this all the time. it’s easy to ascribe meaning when we don’t really know. Turned out for that school, they had changed one of the questions on one of their sites a few months ago. So their admissions people already knew an answer to something that made them a little more qualified, so they were able to be a little more efficient. That’s how it turned out for that school. But I had a conversation where they thought it was just a direct mailing. And it was a conversation like, “This person was on maternity, I had these two interns, they worked on this project, it must be that project. Right?” Because humans have been shown to make a decision first, then look for evidence. And this is a good book about that. How many of you guys read this book?
You haven’t? I highly recommend it. Wonderful book. Taught me a lot. Made me a better communicator with my girlfriend actually. [Laughter] Audience member: Wait, wait, what was his last name? Kahneman. Daniel Kahneman. He won the Nobel Prize. He’s an accidental economist. Behavioral economics, is the field he founded by mistake.
So, I will give you the biggest pitfall that I see people make all the time. I know you guys aren’t all marketing people. Doubling down, you have this urge to double down. So let’s take this–. Here’s a segmentation categorization. Let’s say you take all the types of people that come in, and I think these are nine different personas let’s say, whatever it is. A through I. And people that fall in the A category are twice as likely to, let’s say, enroll with your school. Right? Then E, which are exactly as likely as anyone else, and these I people are, you know, half as likely to enroll. Then you get someone in and they’re like, “I want you to double enrollment.” and you go, “Well I gotta find all the A, B, C’s.” Right? That’s not always the right answer. Because we’re working on one piece of data over here. You know what we don’t know about this? How much does it cost when we get another A? And how much does it cost me to get another D? And are there things that I can do about these D, E, F’s and G’s to make them more likely to start with my school? Generally speaking, the answer is I can probably spend $10 a head, $10 per unit on a D, and get maybe a 40 percent increase in volume a year. But maybe a 10 percent increase in spend on this might only get me five percent more because I’m already doing really well with them. So, when I say resist the urge to double down, I don’t mean you shouldn’t do it. But realize that you’re looking at one piece of information, and one of the most important pieces of information is inefficiency.
I looked at cost. That’s what’s going to help a marketing team a lot. Capacity might be an issue, right? You’re doing really well with the A’s, but you’ve got one class for it, and they don’t want to put another one together. I don’t know. I’m kind of making that up. So, resist that urge, Look at it holistically. So, let me give you some pointers. You guys want to hire a data analyst now? Or be data analysts? If I were to hire one, obviously they’d have to have those skills, that challenge that I talked about. Now I’m not going to talk about it the rest of the time. Of course, you’re going to measure their skills technically speaking to make sure they can do all this stuff. And then these are the softer things they look for.
1.They have to be intellectually curious. Demonstrate intellectual curiosity. They have to be numerically inclined, that is a given. I have a degree in math, and I will tell you, you don’t have to be a “Quant.” The best analyst I have on my team has a fine arts degree.
2. Growth mentality is key. You need to believe that you can stick to it and learn something.
3. Be willing to challenge their own assumptions at all times. This is really important. If they can’t challenge their own assumptions, they can’t get into a testing mindset, in my opinion. It doesn’t mean they have to be willing to change their mind easily, that’s a very different thing. They can be willing to challenge their assumption but allow a high bar to change their opinion. A bar in which you see the information, and it’s compelling enough to change their mind.
4. Seek information from subject-matter experts. In data science, it is about building coalitions, building partnerships. It goes back to your point of being able to communicate with people. You can’t’ be in a little room and do the math and make recommendations without talking to people that live the process every day. Now, these are not qualities you need for like, a big corporation. Amazon doesn’t necessarily need this. But in higher education and smaller companies where it’s disparate, you need these skills, I believe.
5. Tenacious. You need to be tenacious because you’re going to have a lot of failures. Have you read this book: Superforecasting: The Art and Science of Prediction? It’s a terrible name I’d say, but it’s an easier read, its’ a way easier read. And it is a fascinating take on these people, forecasters, who are really good at figuring out what might happen in the near future. And what you find out is they’re like normal people that just happen to have these characteristics that I’m talking about. They’re kind of like the people that call into like, Wait Wait, Don’t Tell Me.
So, the last part, I don’t know how much time I have, ooh, five minutes, I think I can do this in five minutes.
Encouraging a data-driven culture
So again, I told you, the devil is in the details. Education on data systems and processing, some amounts of information is still important nowadays. It is a big pain, but you can’t not look at that. So, you need that to create a culture of innovation. I’ve found that it is much easier when you find a willing, or sometimes, unwilling partner in an organization to start with. I happen to find, in my experience, that marketing departments are usually it. They’re willing to do anything if it means they get, like, you know, $2 off their cost-per-enrollment. I can’t stress enough that testing and iterating continuously is so important. What is true today may not be true tomorrow, but it may be again the week after. That happens all the time. All the time. And unless you’re continuously testing and iterating, you really don’t know when this is true or not. I believe in something that I call democratizing data. If this is possible. I know you have HIPPA, you have all sorts of regulations. If there are ways to share data, giving access to as much as information that can help them, these are the different groups you have to work with. And give them a voice and a feeling of ability to participate. This helps in building coalitions, partnerships, talking about why information exists that way, what are the pitfalls of trying a new tactic, etc. You might have different departments that have different goals, and they help actually, in my opinion. If you can do it right, it helps ground you in your–. Because we all have narrow views at what we’re looking at. We always realize the narrowness of our views. If you talk to your partners, and you give them access to data, and they can have intelligent conversations back with you, you can build beneficial conflicts for you instead of delaying each other working towards a solution.
We’ve got two more points. One, everyone knows. Don’t blame, instead, learn from failures. And I’m going to steal something that I found in a lot of presentations. The first one was in Miss Black’s presentation. Celebrate every win, no matter how small. I want to thank you both for letting me use pieces of your presentation. And you can email me anytime you want, actually. Do you have any questions? Let me know. Right now? I think we have a couple minutes. Audience member: I have one. And it’s more of a comment than anything else. One of the things that we’ve been frustrated with is we get data and then we have nothing to compare it against. Right? Especially in the adult and graduate world. I went crazy with this about a year ago. Somebody was asking me what’s a good retention rate at the graduate level. I don’t know! You start with what you have and then you test it over time. Sandesh: Right. I know what we’re doing. But it was sort of like, people were saying, “Well how does that compare?” and yet there’s no data, it’s frustrating. Because at the undergraduate level, there is. You can find that kind of data. You can find people that’ve written dissertations on that kind of thing. Unfortunately, at the graduate level and at the adult level it’s very hard to find anything that is comparable, and much less comparable, to the same situation that you’re in because we’re all so different. And so, I think one of the hardest things about data for us is the fact that you almost feel like you’re doing it in a vacuum. I mean, you have to use it, like what I said earlier about student service, you know, it’s my benchmark, I’m not going to compare it against anyone else because we’re unique in our situation and what we’re doing. So, to various others, yeah, it’s nice to know, but what I care about is am I doing better over time because that’s all about I can do. And that’s really the key. We’ve had similar issues. I mean, when I first started looking at it, I’m like, less than half a percent of these people are enrolling. That’s terrible! And they’re like, actually that’s better than we’ve ever done before. I’m like, really? I guess the answer is if there’s something you need to look at, if you can’t get your friends to tell you what they’re doing, you just measure against what you’ve had in the past. Or you talk to Carol. [Laughter] Any other questions? Audience member: Great presentation, obviously. I think it speaks for itself, the passion that you have for analytics is infectious. But um, we work together, Sandesh and I, and just to briefly touch on your point. We are probably partnering up on just aggregate data that Education Dynamics has as a whole that we can provide to the industry. So, definitely look out for that in future emails that you’ll receive from me if you haven’t received a whole bunch already. But to the extent that you can, because I know that you work on a different team, how do you think the development of GlassPanel, from an analytics perspective to help obviously track these students, these inquiries, and speak to S.I.S., and view the enrollments, will kind of change the game for higher education, right? Because that’s a big initiative of Education Analytics, and I think you have a decent amount. Sandesh: I helped develop some stuff, you know. The pain point is real, right? Having disparate systems, that’s a real pain point.The value you can get is if you can make it easier to get to that all together, you ‘ve just saved yourself, I don’t know, probably two hours of your analyst’s time. At a minimum. It sounds like nothing, but over weeks, months, that is a huge amount because you can spend more time on tactics and strategies and testing rather than actual data munching. And that’s a hard thing to do. Data munching is a hard thing to do. I know because I spent weeks and weeks talking to the developers of GlassPanel on how to do that. But having a system that does all that would help. That’s where I see the benefit. It’s just easier to get you in. We have to leave, but I’m available for additional questions. Thank you.