Dive Deeper with Niche: Using Multiple Attribution Models to Paint the Full Picture

Ready. Welcome folks.

People streaming on into Share.

I just want to say hi in chat, say who they are, where they're from. Feel free to do that.

There's a lot of you register.

It's always fun to see see everyone coming in from all over the US and it looks like all around the world even.

Oh yeah.

There's a bunch of folks sign up, so we'll probably wait just a little bit for folks to stream on in.

Hi, Diane. Thanks.

Oh well, you already have fans here. That's awesome.

Rebecca, I'm sorry to hear about the.

That's that's a dedication though, to get a little break and enjoying this, but hopefully they're they're doing well or taking a nap at least.

Alrighty, Well, what do you think? Should we?

Should we start, We'll obviously we'll start with some, some housekeeping, so nothing too.

Too exciting quite yet. So hello y'all, my name is Max. I'm a technical engineer here at Technolitions. I'm joined by Will Patch from Niche and he's going to have so much good stuff for you today. I'm just here at the beginning to do a couple quick housekeeping things. Looks like I have a bunch of you in chat here, so jump ahead to this slide. So this webinar is being recorded and will be available for viewing later. So if you want to review what you saw.

You know, send this along to colleagues. You can do that if you need closed captioning. There's a CC button in the top right of your window that you can click. That's also where you'll see the expand button if you wanna make this full screen.

If your audio or video gets out of sync, you can just click refresh on the share window. That should refresh your browser. Get that re synced up. And of course lots of you been posting in chat so feel free to post questions. There's anything I can answer as we go. I'll be happy to, but I'll be collecting those for the end of the Q&A with Will SO.

As we go along, just ask those questions and enjoy the presentation.

All right. Thank you, Max. So we're going to be talking about something that I hope that either you have some passing knowledge of or you've heard of. I really like attribution models.

They're they're great way to measure what's being done, but I think there's some other hidden benefits that they will get into here as well.

You already know how to do file exports, so now what do we do with them? That's our second step here for those of you who don't know me I haven't met before.

My name's Will Patch. I'm a Senior Enrollment Insights leader here at Niche. I've been here for five years now. Before that spent nine years at Manchester University. It's their giving day-to-day. So I'm excited about that, taking a little break here, and I'll be excited to see how things look when I come back. Yeah, it it's. And then before that spent two years teaching. So I've spent a little bit of time around education. I come from a family of teachers.

Having a good conversation with Max beforehand too about sort of the the path. You know, I took the the traditional path into higher Ed.

Studying chemistry and psychology and then going to consulting and teaching and ultimately here, you know, it's that normal path people take into the admissions world.

As everyone's kind of we're we're going here, I wanted to throw a little poll out here. Let me get this entered in.

I want to know, are you currently using attribution models?

And how?

This is it'll give me a good baseline to know where people are. Poll is coming.

Fill in suspense. There we go. You can go ahead and vote on that.

Attribution models. We'll get into what they are, how to use them. It looks like most people answer is no. You know, I absolutely understand what people say, that they wish they knew if they were using them. Sometimes it's held by other people in the office or in another office, like IR, institutional research, anything like that. Yep, looks like mostly.

Mostly not using them.

Alright. So I think the first question we're going to have then is what is an attribution model? What in the world is this thing? Well, an attribution model, if you're in the marketing side of things, you might be a little more familiar with it. That's a good question. Whether it's in the office or in slate, an attribution model exists to give credit where credit is due. You always want to give credit to where the students come from, who's working hard, things like that. So think of it as a way of giving that high 5.

Giving that kudos the pat on the back to the the sources of where your students are enrolling from.

So when we do this, when we think about it, an attribution model, ultimately it tells you where your students are coming from. And it's going to tell you how to maximize your budget, maximize your reach. Because this attribution model is assigning value to every source the students come from, whether it's a certain type of inquiry form or a campaign or a visit program, you're going to get more full picture of where your students come from.

O When we use an attribution model, I don't know if anyone here is on the marketing side of things. You can drop in the chat if you are, you'll see it much more used there. In terms of digital marketing, different campaigns, this is slightly different. It's a similar idea, but applied to student search.

But with anything, you're going to need good clean data, strong reporting to make use of these models.

O What does it do? Well, throughout the student journey, we have this awareness stage where students are coming to you either as a prospect that you're purchasing the name of.

And or they're just window shopping, they're in the awareness phase, you don't know about them yet. So they're talking to their friends, they're reading about you on review sites, they're looking up online, going to your website, They're in that education, educating themselves phase. Once we move into consideration where they're an inquiry, an applicant, they've been admitted.

Now you can really get more into this because you're sending the mail, you're sending them emails, you're helping them kind of weigh their their options there and then ultimately we want them to make that decision, hopefully you, but at least somewhere where they're depositing. Sometimes that is ultimately about the financial aid. Affinity is always going to play a role in there and sometimes it's just luck of the draw. You know, these are 1718 year olds. They're going to have some interesting reasons for making their choices, but we want to make sure that.

You are top of mind.

And that you can then see, OK, what did we do along the way and how did that help the student get from point A to point B and ultimately graduation walk across that stage?

So when we think about where our students are coming from, you have these direct tactics of influence. Think about your Comm flows. So your emails, your print, your your in person events that might tie into that.

QR codes can be easily measured. They're very direct. We can definitely say this student scanned that code and took this action. Your yield events. Easy to measure students who deposited while they were on campus with direct admissions. Very easy to see. This student received an offer, they received their scholarship.

They accepted the offer and then enrolled. Very easy to measure those other things. Much more indirect, but very much a part of the process. It's so hard to try and apply value there. Things like word of mouth, you know, they heard from their friend, they were talking to their parents, talking to their counselor. We have these passive campaigns. Think about your awareness campaigns, like billboards, just sending out Flyers, kind of reminding students you exist. Much harder to measure.

The the impact of those Same with social media.

Unless you have very targeted campaigns with a URL that you're tracking with UTM codes, it's starting to sound like acronym soup. I know there's there's much, much harder to measure the value there.

So you can still do some things, but just not going to get the best value I mentioned you have to have clean data, right? How do you keep your data clean? How do you make sure that even that sort of fuzzy stuff you can get more value out of?

Well, as we go through all this, you make sure you keep your data clean. There we go by making sure that you don't have delayed imports. This is something that if you want to measure time from when a student comes to you or raises their hand as an inquiry or applies until they take the next action.

You need to think about make sure that you get that student into your system as soon as possible. I always recommend as soon as you get that file, import it. If you can. Set up an SFTP, direct imports, things like that, do it. Don't do a weekly import, don't do things like that. If you have poor sourcing where let's say you take all of your students from niche and just bundle them together. Whether they're a direct admission student, whether they are inquiries, whether they are cross interest prospects, general prospects.

Or if you say, well all of these students came from College Board, well you might be an AP name, might be a PSAT name, might be a SAT name. Having that will will muddy the water. It's going to make things harder If you're not cleaning up your duplicates, that's really going to make things hard.

Data silos exist, hopefully much less so for you since you're in Slate, and hopefully you have a lot of people using it from across campus. But it still exists. But then there's just this reality of complexity.

You know, students throughout their search, there's a lot of interactions. There's there's not just this. A student receives an e-mail and they instantly say, yes, I'm going to apply in a role right away because you also have then the personal interactions. You're going to have the high school visit, you're going to have the visit to campus, you're going to have all the other things they're receiving and considering.

In a perfect world, everyone would just one action, one outcome. But we have to think about, OK, this isn't. This isn't a clear 1:00 to 1:00.

And and we need to be honest about that when we think about think about our attribution.

Getting a little lag here in the slides.

There we go. So why should we use attribution models? There's a lot of reasons. First and foremost, it's all about value. To determine the price of a partnership or a tactic, that's not going to tell you how valuable it is. What you need to know is this ROI analysis. You need to know what your return on investment is. So based on this campaign, this lead source, this inquiry source prospecting, what are we going to get out of it? How many students actually enroll?

Graduate and what's the dollar value assigned to that? I know sometimes people don't like to say the student has a dollar value, but it's kind of the reality of it too. We we always have heard that. No margin, no mission if you're not getting.

If you're not getting a value there from that source, if you're getting students who are costing you more than than what you're getting out of them, unfortunately that hurts the university. It's hard to keep educating the next generation if you can't keep your doors open.

So this is absolutely something that is essential. I I hope that I'm getting that point across. One of the nice things, once you start calculating the the value there, it gives you a tactic, it gives you something in your belt there to be able to pull out and say this strategy, this visit, this source yielded us such a positive ROI, we need to invest more in it. So when you're talking about your budgets, you're able to go to your boss, you're able to go to your president if you're already the VP.

And say we need more budget for these tactics because look at that ROI, look at the outcomes. If we could double the the number of students enrolling at this net tuition revenue, isn't that worth it to you? And I've had success in the past getting that additional budget. I know you can too, but you have to have the numbers, you have to have the outcomes to be able to do that.

There's also a little bit more there. These attribution models allow you to customize and optimize your comflows. So here's the second reason. How does it do that? I hope you're wondering. Well, when we have this number, we can speak to the origin story. So you can do a comparison within your own leads, within your own sourcing and being able to say, OK, these are behaving differently. Now what do we do about? So threw up an example here based on some anonymized data here.

We have two different, essentially inquiry forms on the top two lines.

Your general inquiry form, maybe that's site wide, maybe it's just on your admissions pages and then specialized athletics inquiry.

We see, unsurprisingly, higher volume with that general inquiry form, but then we look at that app rate 20% of inquiries applied but 42% of athletic inquiries applied.

OK. Well that, that alone, OK, we can we can think about what might that be. We look at the admits, we look at then enrolled 7% of students who came from that athletic inquiry enrolled versus 3% from the inquiry form. OK. So something is going on here. We see that one of these two is performing much better.

When we start looking at outside of the inquiry forms, you see your prospects very general there. The agency source names, similar app rates, same funnel.

O OK, those are behaving very similarly similarly.

What can we do? Well, we take a deeper dive. There's always that next step for context. Whenever you're doing this type of thing that you need to make sure you're looking for what the rest of the context is. It's more work, but you're going to get a lot more out of it. You have to think about the qualitative along with the quantitative. So we're looking here at the quantitative. We're saying, OK, why are we getting more? If you only looked at this, you might say, well, we just need more athletic inquiries, end of story. OK, let's take the qualitative look. So let's look at their Comm flow.

If someone just inquires through the normal inquiry form.

The first e-mail they got just told them to apply today, so the student raised their hand. They said I have questions and the response was apply.

And then they were told to visit. And then they were told who their counselor was, and then they were told that it was an affordable college. And then they were told they could visit. Again, this student asked a question, and that question was never responded to.

Now let's look at the athletics response. The student inquires about sport. They give the initial recruiting questionnaire so the coach can learn more about them. They get a welcome from the coaches, the information about balancing student athlete life and then into the affordability and then stories about student athletes, this student said. I'm interested in athletics.

This comflow spoke to that. It told them what life was like as a student athlete. They get to know the coach. They get a supplemental.

Supplemental form that helps the coaches then speak more directly to what the student needs and helps them figure out how that student would would fit on their roster, fit in their philosophy.

Both of these, you can see both of these are students who raised their hand. They said I'm interested and we have very different responses. This is why you have to have that qualitative piece to it.

O When you're doing the analysis, which we'll get to right here, when we look at what the attribution models are, there's always that second step of these two are behaving differently. Let's see if we can figure out why.

My my recommendation, if we see this, is then you take that next step. OK, let's experiment. Let's try and optimize our inquiries. We see a lot of them. So what do we do? Well, let's make the inquiry com flow more relevant to the student. We make it speak to what they care about. We make it speak to their question rather than just bombarding them with what they what we want them to do.

So that that's, I know it's a little bit of a rabbit hole before we jump right in, but I don't want to lose that important message that.

Yes, an attribution model is all about the quantitative analysis, but you need to have the qualitative aspect of understanding the why and the So what.

So you need to be able to say, not just.

This source performed better than this other source you can look at. Why might that be if they're all receiving the same Comm flow? OK, now you have the opportunity to try and tweak one to see can we get better results or is it just the source itself?

But until you have the data, you're not going to know the answer there.

So let's talk about the attribution models. There's a lot of different ones that we're talking about marketing, we're talking about student search. There's five basic attribution models.

I'm sure you've heard of at least one of these. There's first source Attribution, first Source inquiry, Last source, full source, and unique Source.

Five basic ones. You can get all sorts of crazy things if you want to relate it to the marketing side with time decay models, all that. If you have a ******** data team that wants to dive into that, do it if you are a smaller team, if you're a smaller shop VS1, know what the differences are. I would go with one or more of these.

O We talk about first source. That's attributing all value to how the student first came into your CRM. How did they first get to you?

Well, the pro here is that it's very easy to do. It's easy to say this is where the student first came from.

The cons, though? It's ignoring when students actually start engaging. If you import a name of a prospect as a freshman and they don't engage until their senior year because you have the wrong contact information, you had the wrong interest, you know.

Then we're still assigning under a first source attribution model. We're assigning all that value to that first source of prospecting, regardless of whether it gave us any usable information.

It's not a evaluate I can talk. It doesn't assign any value to the recency of when they're engaging.

And it's really overvaluing that early prospecting. It's assuming that, well, if we just buy every name of every freshman in the state, that's all we have to do. Again, accuracy. Are they actively searching? Are they ready to receive the information? Very, very different.

First Source Inquiry. This is very similar, but it's attributing value to the first inquiry source from a student, even if they were previously a prospect.

Slightly different.

So the the positive here of being a first source inquiry, sorry, we're getting a lag again, there we go. The pro here, it's emphasizing when the student takes takes action. Where are they actively searching? Where are they actively engaging? Big positive there. The downside? It completely ignores prospecting. It says the prospects do not matter. All that matters is the action and if you're measuring a conversion action from prospect to inquiry.

You get around that. I don't know that everyone does though. It's also not assigned value to recency again.

So if a student inquires as a junior, doesn't really engage until they come on campus for a visit as a senior, and then they really start engaging, we're saying that all that matters was that first inquiry form. Partially true, but it's not telling the full picture yet.

Third one here, last source. This is value. We're assigning the value here of that last source before conversion. Whether that's are we saying the conversion action is visit, application, enrollment. Whatever it is we're saying was that last source that they came from, whether that is an inquiry partner, whether that has an inquiry form, whether it's a visit, high school visit, whatever it is. That last source is all that matters.

Ros There, This may do the best job of identifying sources and campaigns that directly affect actions, but also discounts A nurturing. So it says that all that matters is that last piece. But if they've been slowly engaging, building interest over the past year, OK, we're saying none of that matters. Just what's that last thing? Excuse me?

That's why I always keep tea handy.

There's some positives there, OK.

Fourth source here. Full source. We're looking at the full picture here. It's assigned value to every source based on the behavior of students from that source, and not the timing at all. O you're going to see every student that came from a source, regardless of whether it's the first, middle, last, where it falls.

The positive here this gets around all of your timing of imports. The only it's also the only complete report of funnel behavior by a source and shows the total impact of that source. So this is a big positive there. Downsides. You know it's not going to be one student to a source. The students will appear in multiple sources, which some people really don't like. But it's also a reality that students don't just come to you from one place. They might apply and then later fill out an inquiry form. I've never understood that, but it happens.

You know, there's lots of things that can happen there. It's also discounting the impact of nurturing on conversion.

Because you're not really being able to say where where through the source the student is engaging.

Last one here. This one is much harder and gives you a much smaller picture, but it adds a piece that we're missing from all the rest of these. That's the unique source.

It only reports on students who had a single source.

So you might say, well, why would we need that? It's going to tell you how many students you would miss out on if you didn't have that source. So if you eliminate that inquiry source, inquiry form on your academic major pages, you know, what impact would that have? How many students would we lose who didn't come to us from anywhere else?

The pros here is showing the the absolute value of that source in terms of students who are missed elsewhere, because if you cut that source out, whether it's prospecting, whether it's inquiry source.

That's how many students you wouldn't have enrolled because they didn't come from anywhere else. It discounts the multi touch reality. Again, it's not showing the full value, only what would be lost without it. Also downside, it's going to be very time consuming to do especially if you have a large number of students in your system. It's going to take a while because you're going through and you're looking for basically just where are we seeing duplicates and removing those duplicates it's you know it's it's going to be one of those that.

It's worthwhile to do maybe once a year, once every other year, but set aside the time for it too.

How do you know which one's right? Or is there a right? I I would argue I don't think there necessarily is a right, So let's use the chat again. I want to know what which one of these feels best to you, What jumps out? Or if you are using one, what are you using right now?

Jump in the chat here.

Using first source I see that a lot.

So, most problematic, but it's certainly the easiest.

Help.

Yeah, I I like unique source, but that was one I I only would ever do once a year. It was it was not something you want to keep trying to to update.

Leslie, love it. You're you're you're going right into what I'm going to say here.

Awesome. Yeah, it's it's going to vary based on your timing, how many people.

Can actually go in.

Yes, Aaron. No, it's the sign back here.

The Maybe I should just start calling it first contact and first say a first source. There's a lot of ways we can kind of mix it, but.

I'm not going to be someone who comes on here and I I hope no one else goes in and says, oh, you know, you have to be doing these things. You have to do it this way because every situation is going to be slightly different. The way if you are only going to use the data in a certain way, it doesn't make sense to do much of work that gives you information that you're not actually going to use. Or if yes, plane don't have the time to do a bunch of these analysis after you export the data, OK.

It's it's one of those that.

I know a lot of people like to come in and say, hey, do it this way. This is going to solve all your problems. The reality is that's not that's not the case. You have to figure out what works best for you. I have some recommendations of course, but these are these are what's worked for me. This is my opinion. I don't want to come in and say if you're not doing it this way, you're wrong. So let's talk about what this looks like.

My two cents. If you can only use one model, I would really recommend full source. That's going to give you the full picture. You're going to be able to see all the students and all the behaviors from every single source you have.

Excuse me if you can layer two. I would do a full source and then add a unique source on and sort of your end of year analysis. Once all the work's done of the class coming in, everything's locked in. Then take a look at how many students if we cut.

These line items, basically when you're talking about budgets, what would we miss out on?

Right. That you may find and I was surprised with some of this when I did back in the day, there were a couple sources that yes, we had great engagement and great conversion from them. But overall we didn't get any students who we wouldn't have gotten elsewhere. And so we had to start thinking about what was the full value there in terms of engagement. And in several cases it made sense to keep them because we got additional information that we got data points we wouldn't have gotten elsewhere or we we got just better engagement there.

And it helped us along the way or you may see, OK, this one is costing us quite a bit more than we're getting out of it. It gives you some leverage to do some negotiation or to cut it entirely if you can manage three sources, look at full source, look at unique source, look at first source inquiry. So that way you see when they're first engaging, you're picking up on that full source, you're seeing how your prospects are engaging unique source, you're seeing how the prospects are engaging. So let's see then where are we first getting that engagement.

Alright, so example here. I'm going to go back to what we were looking at before.

If we have just this single model applied, we have an inquiry form, we have a vendor and we have an agency, and this is all just anonymized data. Looking at this is how how I would recommend setting up a view, setting up all that data that you're going to analyze, look at the total volume, look at the applications from that source.

Look at your admits enrolled students and then you can see what's that full funnel outcome. So inquiry form 3% of the total volume enrolled .1% from the vendor .1% from the agency you look at was the cost of this source inquiry form, you know that's one that you know.

You're going to know.

Some people like to include it as a part of, in your case, your slate cost. I would say you're going to have acrm no matter what. You're going to have an inquiry form. No matter what inquiry form costs you nothing.

Recruitment costs, OK, that's going to be your mailing, Your other costs for that. If you're mailing to all these students, there's a cost to that. You have your cost to enroll, then how much did it cost you each to get each student to enroll?

And then you have your ultimate four year value minus the cost.

Whenever you look at ROI, I always recommend looking at that four year value. So how much? How much tuition or total revenue do you get per year multiplied by four years multiplied by your attention rate?

Because you're not. As much we would like to say all students graduate, reality is reality.

So that's that's just a single source here. If we're just looking at full funnel, here's what this setup might look like.

You're going from total volume. You're getting your enrolled students.

And then the ROI.

So what does it look like when we start layering multiple models?

So I've I've shortened this because I didn't want to do all that. So you have the full number of enrolled students from every source. Well, what if we looked at the first source inquiry, What would that look like? Let's look at then just the first source number of students enrolled and then we have the unique enrollments. How many students would we have not have gotten from anywhere else?

And then you're saying U your ROI for each one and here's why it gets really interesting. We're looking at four different models here, but you're telling four different stories as well. So if we're looking at full source right here off the bat while this vendor had had a great ROI, so at our inquiry form.

If we look at first source inquiry though.

Totally different story because you're seeing here, well the inquiry form still had a great ROI but this vendor, they had no ROI as a negative. Well the reasoning there is these were all prospects.

So in reality, when we think about that way, none of them were a first source for inquiry.

Well, if you're only using a single model, you're getting different stories each place. Let's look at the first source. If you only export first source.

Boy, look at that. All of a sudden you gain all these great prospects. Well, when you're buying a lot of freshmen sophomores, of course, that's going to be the first place they come from. So you have a great ROI there and now all of a sudden the agency has a negative ROI.

And the inquiry form has a lower ROI. You gain a totally different story there too. And if we look at unique, so how much would you lose basically? Well, if you didn't have an inquiry form, I don't know why you wouldn't, but if you didn't have an inquiry form, you have a loss of revenue there. And then we have some negative ROI from the agency and from the vendor there as well.

This is why I think it's so helpful to look at this and say here's why we want to have.

We want to have multiple models to look at because each one gives you a slightly different story. Each one's telling you something else.

There we go. So ultimately you need to find out what works best for you, whether that's the first source, first source inquiry, last source, full source, or in the unique source.

And let's chat, I know we had some questions come in there.

It's a it's a lot it's a little like drinking from a fire hose. But try this out. Do some exports based on your source and start working with some of these and figuring out. I I always recommend, you know, especially later once we get into the summer and you start getting numbers of deposits, expected enrollments. I know a lot of budgets roll over on on May 1, but on May 1, on June 1, July 1 depending on your budget cycle.

But this is like, these are good pieces of information to have when you have those budget conversations and when you're planning for next year.

So Max, I think you were tracking some questions here.

And I've been keeping track of Elsas questions. I'm going to jump around a little bit.

Yeah.

Yeah.

So one question just from a slide you recently had up with the table with the different sources, it looked like Svetta had asked what you meant by agency and maybe you have like examples of what you're thinking and these comparisons.

This is just anonymized data, but this was an agency who managed student search and managed marketing for them.

Yep, Yep.

Got it. Cool. That's kind of what I was assuming you meant. She was asking about like a testing agency, obviously, that could be.

Now that that'd be the vendor actually was a testing agency, but.

See. Yeah, that totally makes sense.

That's why they were all prospects there.

No.

Yes.

Awesome. So just wanted to help help get that one clarified. And then there was a few questions and I think it'd be interesting. I'm curious to hear about this as well. So from Daniel and Chelsea and then Jason had a similar question. So Daniel had asked, you know, how do you access this data in a system like Slate, you know, where do you find full source And Jason is kind of following to that, right. So like, is this happening outside sleep in niche? Are these calculations on a spreadsheet? Yeah, if you could speak to that, that'd be awesome.

Yeah, Yep. So I I because it is so complex, it's not something you're just going to click a few buttons and get it to do. It's going to require multiple exports by source.

And then having a spreadsheet. The great thing though if you do it in like a if you're.

I'll use an example here. If you are a Google school and you have a file on your Google Drive, just a spreadsheet where you're keeping all this. Then you can connect it to a looker studio and people can filter it and in ways they want to, and you can share that out internally. If you're a Microsoft school, you can do it in Excel and then import it into Power BI. Either way, the nice piece there is that you make the data more accessible if you have these handy dandy charts where people can sort by year, sort by type of student, sort by state, however you want to filter it down.

That's a way that you can get the data in the hands of people. If you are reporting out to cabinet whomever, it can be very helpful to give them the data in advance.

That's a nice way of sort of reducing some of the more self-serve you make it. I'm a big advocate there, especially if you're on a small campus. If you can make it self-serve where people can look it up and they don't have to come for you for every request, maybe you have a massive office and tons of time on your hands. But if you're like I was, the more self-serve you can make it, the better.

Give people the ability to look at the data themselves and ask questions. Then instead of saying, hey, can you pull this or what's going on with that source?

It's just going to be a easier experience and they're going to feel like they have more ownership of it if they can play around with it as well.

I definitely can relate to to that as well.

Yeah, you probably get a few requests like that.

Awesome. So you have great questions. Kathy had a question earlier that I'm curious about, Chris, if you've you know, experienced this as well, well which is we have lots of stealth applicants. So what would be the best way to try to identify where they're coming from? I know you focused on you know, inquiries and prospects but.

Well, your stealth apps are essentially going to be a source. That is, a source that you would analyze is your stealth apps, because that application would be their source in that case.

It's going to look a little different because if you're looking at.

All all these examples are what I I think ultimately is the highest value which is enrollment. And by measuring that as a source against others, you're going to be able to see how they progress from application to enrollment compared to other sources.

If you were instead if someone wanted to know.

A attribution model for applications. Well boy your your stealth apps are going to look awesome because it's going to be 100%.

A lot of students are window shopping. They're going on your site, they're doing all sorts of things. If you if you have some ping data there, awesome.

But in reality, treat a stealth app like any other source.

And then as I mentioned, if I can go back to that slide even.

Yeah. Where we're looking at this origin story, so we can look at some of the comflow behaviors. You might see that, well, yes, stealth apps come in, here's their acceptance rate. Here's how they move through to enrollment. Well, OK.

Can we improve that? Can we speak more to them? Because these are students who will have some passing awareness of you. Whether it is family connections, whether it's they just know from word of mouth from your website. They've been reading about you for years. They've seen you March Madness is over. That you know Iowa I'm sure has gained a lot of great applications right now and and as someone who lives in Indiana I'm excited to see what happens to the fever this year. Now. But it's.

You're going to be able to say, OK, we had this, this glut of applications come in stealth applicants, they know us, but they missed all that nurturing behaviour. So what happens if we start adding some of that, some of that nurture, some of the here's more about us type content into that applicant comflow. Can we improve this and being able to see it that way And you can compare that then then to other applicant sources like when you have a common app come in as a stealth.

Where you have a niche direct admissions.

And it's a stealth These are students who weren't in your system before now. How are they progressing through?

Now that we have them in our system, it's that side by side comparison.

That was a very long answer to the question, I know.

It was very interesting for me to hear for sure.

I think that was it in terms of our questions in chat that we got so far that we do have time for a month.

Yeah, Leslie had some good advice there, it looks like.

Interactions by code with comments.

Yeah.

All right.

Looks like we don't have any other questions. Oh yeah, it is. This is one of those things that is it's very important. It's something that I I hope you learn more about the different models and you can. This is something that I, I my goal for all this is that kicks off conversations on campus. I'm not going to come in and say do these things check these boxes, but you can go and go back talk to everyone else on campus and say, you know we're doing first source right now. There's some drawbacks.

What would it look like if we added this? It's going to add some extra work, but we're going to also get a lot more information from it. Can we try this? Can we evaluate our sources differently?

And have those conversations figure out what works for you. But it's going to give you a much better picture if you're one doing this and two layering on different models, so.

Yeah, yeah.

Yeah, Elizabeth's question. Can we create a custom type of origin source in Slate? I'm thinking about first source inquiry.

So.

With it in Slate, what you'd want to look at is all of your inquiry sources. Look for what was the first source there.

If and Max, this may be a question for you or I may have to do some more research on it, but.

As someone converts from a prospect to an inquiry, can we code that in a certain way that we'd be able to query on that?

That would be really that would be the most helpful, being able to see at what point do they convert from prospect to inquiry? Because it may be before another inquiry source. It may be because of another inquiry source.

Yeah, the the way with origin sources and origin origin groups would be to make an inquiry origin group. The only difficult thing with that is, right, if you're familiar with origin sources, right, you just put certain sources into a group and so you'd have to define certain sources as ones that are inquiry sources, right. So your ROI form anything that defines something as being an inquiry. The only thing that's difficult is if you have any sources that could be.

More like prospects or inquiries, right? They're both coming from the same place in Slate.

That's where it's difficult, but if you can categorize them, the origin group would be the way to go there.

Yeah, and if you have prospects and inquiries, so Niche, for example, does both prospects and inquiries.

I'm.

Always recommend two different imports so you have much cleaner data because if you have there, there's just cold prospects. They'll behave one way across interest. Prospect is a student who's looking to appear and says they want to hear from similar schools. They're going to behave differently than just a cold prospect. And then you have the inquiries who have raised their hand and said I want to hear from your university, They're also going to behave differently. So in that case I'd have three different and then if you have direct admissions, there's a fourth all coming from the same.

Location essentially, but it's four different types of students. You would want four different imports for those so that you can really see the differences and you can communicate with them differently and you can query on them differently because they they are legitimately different students. A student who receives a direct admissions offer and accepts it is not the same as a student who is prospected because of their major interest in location.

I'm good. I'm good.

Makes sense. We've got a couple more questions and then, well, I don't know how much, I know we're getting towards time in a couple of minutes, but Diane had asked.

I've got a little while before I pick up my daughter from school, so I'm good.

Perfect. All right. Well, yeah. So Dane was asking in chat, how do you reconcile multiple lead sources? Basically, you know, who gets the credit, the first source inquiry, how do you handle that?

Yeah, so that's that's what this is ultimately about. If you have a student who appears in multiple sources, it's going to depend on which model you're using. If you're using first source, then just that first place they get it gets the credit. If you're using full source, every source of the student came from gets that credit, and that's why it's painting that complete picture. You're going to have students appear multiple places, but it's going to tell you the value of that source as well.

Using unique source, there's students who won't be credited anywhere, and then there'll be students who are just credited to one place, wherever that unique source was.

This, that's ultimately what this is is about, is who gets the credit. It's a little bit like trying to think of a good analogy here.

If I had both of my kids.

Work together to clean up a room. Who do I give the high five to? Well, if I say whoever finished first, then one of them gets a high five. If it's whoever started first, which would definitely be my daughter, she gets the high 5. But ultimately that full source would be, well, you both worked on together, you both did the work, you both get the high five. And if we did as a unique source, well, neither one of them only did alone, so nobody gets the credit.

The the credit will be different depending on which model you use.

Which is why it's it's helpful to look at and layer it. Yeah, I'm glad. I'm. I'm glad that helps.

If you layer it, the idea of layering isn't to say, isn't too complicated. It's because when we look at this, am I around the right side? When we look at this, it's telling different stories.

So this is where that qualitative piece comes in. You have to be able to think about, OK, which of these makes the most sense for that source, Which of these makes most sense for what we want to use the information for? Because if we're only using first source here, well, this agency did nothing for us. Well, they may also be doing other things. Or we see here wow, that all these prospects were great because we bought all these sophomores And then you have to get into the qualitative of OK, but then they didn't actually engage for several years because.

We were sending them information about business when they're interested in art.

Or we've been calling them on a disconnect number. Or we've been emailing them at their parents e-mail address.

You have to always look for the extra context. It's not a simple thing. It's not a checklist thing. It's about giving you all this data that you can then make better decisions. It's not about telling you what decision to make, it's informing decisions. So I know, I know that that's maybe not as easy as everyone wants it to be, but sorry.

Yeah. No, I mean, I feel like it's nothing is ever simple and straightforward.

Yeah, Yep.

Yeah.

Umm, Will, if you want to stay on this slide, actually, Andy had an interesting question. You kind of touched on it a little bit as you're answering the previous one and Andrew's asking the best practice when it comes to cost. So the cost of a vendor often includes inquiries from multiple cards. So you mentioned sophomores, but obviously kind of it could be a range. So how do you calculate that cost when you're measuring one at a time? So just for fall, 2024 was Amy's example.

Give you an easy answer and a complicated answer, easy answer. Every year you're going to get multiple cohorts from them. So you can just use that full cost and and say you know next year we'll also be getting under class. So let's just look at full picture right now.

The complicated way would be my wife's, the accountant and I'm blanking on the right term here, but.

We're basically, we're going to say, OK, if our cost was $20,000.

And half of them were seniors. If that's a class we're analyzing here, half of them were seniors, then we'll assign a $10,000 value there.

You can do that after the fact and just look at what percentage of that cost was for students in this class. You can get even more complicated if it's a case of, well, half this contract was for search and then there's another part for marketing. Yes, that marketing also inform the search, but we'll just do that portion and then we can also then divide that. It just depends on how complex you want to make it. And that's where part of that you can just clean up in the aftermath and say, Yep, so here was the ROI, let's say it was.

A full source ROI of 1.77 million. Now of that though some of that cost also went to these other things.

OK, you just clean that up in the narrative.

But you can also then I wish I could remember the right term there.

It'll come to me after the webinar right now.

I'm sorry I can't help you there.

No.

Well, I think that was it in terms of questions in chat and I know we're a little beyond our 45 here. So well, thank you so much. That was super interesting, super helpful. And thanks folks who ask questions and chat.

Yes, thank you all.

Right. Well, we'll go ahead and close out the share please.

You know, check out the recording.

Yep.

Contact Will. He has his info up here and sign up for other Slate present sessions.

Yeah, thank you and hope everyone has a great, great rest of the day here. Thank you, Max.

Thanks well.