See how AI transforms your SSP analysis workflow

So good morning, everyone. My name is Lila Kobrosli. I’m a senior product marketing manager, and joining me is Cathy Pearson, a senior principal revenue advisor. So for today’s webinar, once again, thank you for joining. We’re gonna be talking about Zuora Revenue: Build the right SSP strategy with AI. So just a quick, um, forward-looking statement, and this is just a precautionary measure. Our legal team asks that we share this on all of our presentations. I won’t read it word for word, but please take a moment to read our forward-looking statement. Great. Thank you. And next, we’re gonna go through just a couple of housekeeping items. One, please introduce yourself in the chat. Say hello, let us know where you’re joining from.

Two, we have links and resources, so click More at the bottom of the Zuora browser labeled Resources. And ask questions. We do have a Q&A section devoted at the end, but as questions arise, p-please feel free to drop your chat. We do have somebody monitoring the chat, so we’ll take all of your questions. Uh, four, we are recording this session, so we’ll share a follow-up email with the recording. And five, we have a couple of poll questions, so please participate in our poll questions. We’d love to hear from you. All right, so before we get started on our topic of, um, the AI, AI SSP strategy, I wanted to introduce the speakers once again. So I’m Lila Kobrosli, senior product marketing manager, and with me is Cathy Pearson. Cathy is going to go over the topic of the AI

SSP Analyzer as well as our demo for today. So great. Now, we are going– So today’s topic is AI, and the way that Zuora is focusing on AI is that we are– we f- we believe that AI is in three pillars. We believe every company that wins in the AI era will be able to do these things, and we believe that we are the ones who can help them do it. First, we’re helping monetize AI end to end, from pricing to revenue. We’re helping companies actually make money on AI monetization models like usage, prepaid, hybrid, and to run the many intricacies that finance teams incur. Second, we’re supercharging quote-to-cash teams. With the AI Zuora’s products,

we’ve essentially taken every step of and the workflow within QTC and made it ten times faster while retaining all the controls and finance, uh, needs to feel confident. And third, we know that the days of long implementation are gone. We’re helping companies get live fast and keep adapting fast so their systems aren’t a bottleneck, and we are built on trust. We’re doing all of this with the controls and data finance can actually trust. For today’s webinar, we are actually focusing on the second pillar, supercharge quote-to-cash, and how our AI-powered SSP Analyzer helps create a faster workflow with the control finance teams needs. And on that note, I’m gonna pass it to Cathy, who’s gonna kick us off with a poll question. Great. Thank you, Lila. Yeah, so before we get started, just really curious

to see how people are managing SSP analysis today. So we’ve got some examples here of maybe what you’re doing. Just real curious where people are on this journey in terms of SSP and analysis. Yeah. And if… You know, you may not have a formal process, or it may not be something that you don’t have to do today, so that’s okay as well. And we’ll just give that a few more seconds here while everybody’s voting.

And just let us know when that poll closes, and we can keep going. Ah. W- Actually what I expected, it looks like sixty percent of us are doing it in spreadsheets. Great to see that we got g- about twenty percent that are actually already using Zuora’s SSP Analyzer, and a bunch that don’t really have a formal process, so this will be good. So we’re gonna start talking about a little bit about our SSP Analyzer and what we’ve had for years. And some of customers even

started with this way back at the beginning of ASC 606. What does the SSP Analyzer do? Basically, you create a template that allows you to say, “What are my attributes for stratification?” Think product, family, geography, uh, type of customer. Then what calculation method do I want to do? Am I using a median or what we call our optimizer approach to best fit approach? I then look at how compliant do I want things to be. Just give me some range of what things are happening. What am I looking at? Once I get that, I then create batches that take a dataset, run through that analysis based on those attributes up front, and come up with how compliant am I? What does it look like?

Users then can review, look at that, adjust the analysis. Then when they approve it, it can be used within Zuora Revenue to actually as SSP. What we found, though, is really it’s figuring out if you’re doing it for the first time, what are the right attributes? Is it product family? Is it customer? Is it currency? Is it region? Should I be using two of those? Should I be using four of them? Which two? Where do I get to? It’s kind of like looking for a needle in a haystack that first time, and that’s really where the SSP Analyzer is gonna come in. Because think of it as an AI coworker. It’s gonna use a genetic algorithm to automatically generate, test, and refine

through all those different permutations of stratifications that you can have in a dataset. At the end, it’s going to define what will be the top-ranked stratification for you so that you can look at it, and it’s got some logic as to what it’s going to choose as the top. So that’s really what we’re trying to do, is make that process easier, so faster to get there, and then picking a stratification is going to be easier m- to maintain over time. So with that, let’s go ahead and jump into our demo of the SSP AI. So when I go into Zuora Revenue and I click on my SSP, I now have another way to look at SSP, what we call the SSP AI. And for those of you that have used Zuora

Revenue’s SSP analyzer, it’s gonna have a lot of familiarities to it. But def- but what I don’t have is I don’t have a template, ’cause I’m not defining things there. I’m defining things in a batch. So I’m gonna go ahead and determine my batch, and as you can see, this allows me to actually use external data to pull in to look at this analysis, and then ultimately my database. This is something that we’re working on to be able to use this with data we already have. I’m gonna choose my file, and I’ve got one here, so I’m gonna pull that up, and it’s gonna say, “Do I want to run on this?” So I’m gonna upload that file, and now we’re gonna see that it’s gonna do very similar things to what was happening in our analyzer. I’m just gonna say, “This is

my webinar test.” And I’m gonna choose those types of configurations that are gonna help the system understand what I want to do. Am I going to do an analysis that’s based on sell price? Yes. That’s usually what we see. Some customers say, “I want to do it on the cost,” if there’s a cost thing, so I want to do cost plus. The value type, am I doing percentage or amount based? So think percent, percent of list, discount off list, or just an amount. Then, how do I want to do the calculation? This is where I’m doing median. What’s the middle version versus an optimizer? Think of that as going up and down a scale and finding the best fit where I have the most transactions

in that range. The count type, this is really when you think about a transaction line that I have. Am I doing each line as equivalent to one, or am I doing quantity sum, where I’m saying if a line has 10, I’m gonna count it as 10 versus counting it as one? Then I’m setting up some tolerances. So I’m setting a range for what I want to look at and what my compliance is, is 80%. Now, this isn’t, you know, old VSOE days. I was plus or mi- 80%, plus or minus 15%, and I didn’t have VSOE. This is just how it allows you to look at how compliance are in the system to run through and say, “Hey, based on your compliance band, what’s going to give me the

highest SSP value, the highest concentration?” So then I’m going to create that batch. Now, this is where AI really kicks in, because it’s going to run in the background, and it’s gonna take it a few minutes, so we’ll refresh it. It’s gonna come through and get through different groupings. So what did it do? Very quickly it said, “You had all these attributes in your data.” So I had the standard ones of category, family, product line, business unit. I also had some other attributes that I defined. My attribute one was region. My attribute two was a customer size component, say small, medium, large. And attribute three was kind of a level of service. Think silver,

gold, premium. So in my dataset, I had all these candidates of what the system could look at for my stratification columns. It then went through and said, “Hey, here are the top three.” I get the top three, product line, business unit, and then that level of support. That makes sense. Gold’s gonna be more expensive than silver. It’s gonna have a different rating on it. But I’m sure all of you are looking at this idea that my number one, my gold star, is only at 66% compliant, but my bronze is at 73. Why is that happening? That happens because we give it a fitness score, but before we go there, let’s go look at what I can see. It’s gonna set up what those groups are,

and it’s gonna tell me that same information that I had about it. How many transactions are in it? How many are compliant? It’s gonna rank them all down through the list. I can hide those groups. I can do the same thing for each group. More groups here. That’s why… And I wanted to point this out because now I have 24 groups versus having 12 groups. That’s really what we have, is this idea of a fitness score. So what do I mean by that? If we go back to the slides and we look at that, this idea that it’s gonna rank a couple of factors when it’s giving you that ranking of, uh, gold, silver, bronze. The first, of course, is where do I get the highest compliance?

Where do I have the most transactions in those bands? But the second is really about simplicity. Think of this as, how do I do the simplest, the easiest to manage? So if I can get high compliance in only using two attributes, that’s probably more likely for me to maintain and keep those two attributes as being there than doing five or six. Manageability is really similar in that it’s really about how am I gonna manage this? Think about how many different groupings do I have? How many transactions do I have in each of those groupings? ‘Cause I’ve got to have that thing that makes that group be, um, eligible, ’cause we don’t want to have only five transactions in a grouping. So it’s really doing this idea of looking at these, um,

factors to come up with that fitness score. That’s why we saw that top one as gold versus being silver or bronze, even though it was at that 67%. And if I think back to what… If I had done this originally, go back here, this probably, I would’ve had to run these all differently. I would’ve had to create a template, pick those two, run that analysis, look at it. I would’ve had to take another template, run this way, and I would never have known side by side where they look. So this idea that it’s going to give me that analysis right up front, then from there it gives me that starting point. I can then go farther, dig through, make any adjustments I want,

but it’s a way to get to that starting point much, much quicker and much more manageable. So I know that was quick. It was a lot of information, but it’s really a topic that I like, and if you’re interested in learning more, please let us know. And I’ve… If it’s something that’s interested in right now, any questions, if someone is interested in actually seeing the analyzer itself, ’cause we have lots of time, even though it’s actually a pretty quick webinar, um, we can do that as well. So

once we get those poll numbers going, we can move ahead and talk about what’s coming, and then we can definitely get to questions. So in looking ahead at Zuora webinars coming up, uh, in July 30th, this is a topic that we hear a lot when I go out and talk to prospects and customers.

How are we using, and Lila touched on this, AI to rewrite how we do our billing and revenue implementations? How do we get you live much, much faster? That’s on July 30th. And then on August 20th, another topic that always comes up about AI monetization. This is focused on Zuora revenue and really looking at those topics of breakage and overage on August 20th. So please be on the lookout for those. One thing that I wanted to point out is that I do believe that our product team is actually gonna have a table talk coming up in the next few months going into more detail about our AI SSP analyzer, so that’s something to be on the lookout for. Um, so with that, do we have questions that we can take? Uh, that’s… Could someone help

monitor for us? It’s my fault, I stopped sharing. Hi, Kathy. Thanks. That was great. We have a few questions- Sure… in the chat. Um, I can kind of read them- Sure… for you. We’ll go through it. Um, an anonymous attendee submitted what was the template that was selected in the analyzer tool? So there is no template selected. That’s… And the A- and the AI would… It doesn’t select a template. So that batch, let me go in and show what I mean by that. So that batch, let me go back in, just you, within the batch, you set up similar information to what would be in a template normally.

So what I mean by that is it’s in this batch. I’ll just choose that same file so we can get there again. I’m choosing what would normally be in a template, because the thing in a template that I normally do for the analyzer is this information about analysis and type that’s in a template. Now it’s part of the batch because you’re not doing the stratifications, ’cause usually that’s the key to the template on the analyzer piece. So there is no template, it’s just the information here that I did on how I set up the batch. Okay. Got it. Thank you. Um, which approach is the tool using? Is it median or weighted average?

Um, it can use either median or it’s not really a weighted average, the optimizer. So if you think about the optimizer, I have my range of plus or minus, I’m just say, 20%. I would start at zero, this is optimizer or an amount. I would look at and I’d move up a scale by, very narrowly up a scale, and the optimizer’s gonna go where on a scale of, say, zero to 100% of list price, where do I find the highest concentration of transactions? That’s the optimizer, optimizer. The other is truly a median approach. A weighted average is not something of a calculation type that we have today, but if that you’re interested in, I can definitely feed that back to our product team. Okay, great. Thank you. Can you share

what the CSV looked like that you uploaded? Sure. Let me go and get to that. See if I can grab it from another- So this is a very simple file. One just has unit list, unit sell price. It has those attributes that I had. It’s got a quantity, and then a discount percent to come up with. So this is a standard template that we would use to do this. Okay, great. Yeah. Thank you for sharing. A couple people in the chat have been asking for what the actual visual of what the CSV looked like. Yeah. So this is really helpful. Um- Absolutely. Another question that came in is, can we use an upload

if certain data points are not currently being collected into Zuora to see if they would yield better results that those currently are in? Yes. Yes, because you’re using that CSV file, if we… You would have to download it and then upload with that CSV, and then we would be able to do it. Yes. Okay, cool. And then does the AI optimizer show what methodology or logic was used to get the highest compliance percentage? So it’s gonna use this logic. So what… It’s using the logics in the batch. I think the question is within that batch, when I go back to a batch, can I see which logic it used? So if I look at this, I think that’s the question is if I view this actions, can I see which logic it used? I don’t know that answer, so I will take that back, and that’s one that we can share out

with the team. Yep, of course. And any questions that we don’t get to live today, we’ll make sure that your Zuora account teams follow up with you guys after the fact. Um, but we’ll try to get through as many as we can with Cathy right now. She’s in the hot seat. Um- It’s all good. [laughs] Another question is, um, is this GA already or an add-on within the product already? Uh, I don’t think it is GA yet. We can get the date for you. I think it’s in the e- It’s not yet. Yeah. It’s coming. I believe per Catherine that the AI part is, um, but- Yes… the normal part, yes. Um, that we know of. The normal part is an add-on. It is an add-on. Yeah. I think the AI part comes with, is not with that, but the SSP analyzer is an add-on.

Yep. Okay, cool. Um, is it being used on Zuora for rev- revenue and billings or a prereci- prereq to use this SSP tool? Can you ask that question again? Is it do I… I think the question is, do I need to be a Zuora billing or revenue customer to use this tool? Yes. I think the answer is yes. Mm-hmm. And then is it, is the SSP ana- analyzer a standalone software? Uh, no, we sell it as part of the, the product. Okay, cool. Um, okay, another one. Once you run the batch and you get to your compliance ranges, does the AI tool upload the SSP template onto Zuora, or would that be a separate step?

Uh, that is where I would say that’s what the product team is working on in a couple of months as we make this so that it can be, that it becomes, um, finalized. Right now it would be a second step. Okay, great. Um, and then just for people who may have joined a little bit late, what is the SSP analysis method that is being used, just to make sure we’re all clear? So in this I used what we call our optimizer. That’s what we’ve used in our analyzer for years. It’s where it’s a best fit approach. So think about going up and down a scale, and it says, “Where do I get the highest compliance range, the most number of c- of transactions in a range?” So, like, I’m going from zero discount to 100% discount. Where, maybe it’s 50%, that I get the highest one. So that’s median. I mean, excuse me, that’s

optimizer. Median would just be where do I get… Where’s the middle of the transactions? So we use those approaches. This was optimizer. Okay, great. Thank you for clarifying. And then another question that we’ve been getting a lot, can you explain the differences between the gold, silver, and bronze stratifications in the example that we have up here? Right. So, so if we start with, if we look at the different a- um, stratification columns it used. So it looked through these one, two, three different columns and came up with and said, when I use product line and, uh, like, level of service, which is attribute three, that yields me 12 groups, and in those 12 groups I get to

overall compliance across them of 67%. So it’s looking, if I view those details, it’s gonna say, of them, on average, I’m at 67% compliant. So I can see I have a few that are lower, I have a few that kind of are in that range, and I have some that are great. So the first thing it’s doing is going through and saying, “What do, what groups do I get? What do they look like?” I’m doing that for all the different stratifications. In this group, I now have 24, so my compliance overall is higher. So I’ve got, I’ve still got the outliers, but I’ve got a bunch… But if you look, I’ve got very small groups because my data set’s small. So what the fitness score of the gold, silver, and bronze does

is say, “Let me look at, A, what is my compliance rate?” And I don’t know the exact weighting that it does. Um, that would be something that we can walk with our product team members in te- in terms of, is this weighted? Is the number of, you know, is a compliance percent weighted 60%? And then the number of columns weighted 30%. And then that aspect of it. But it’s trying to come to, where can I meet a, a high compliance rate overall across the groups with the least possible stratifications that I can come up with? Because when we think about it, you always want your SSP to be constant over time. You want those stratification

groupings to stay rather consistent over whatever analysis you’re doing. You don’t want to have to change those stratifications every time you do your SSP to get to a higher compliance rate to meet it, because then is it really the right attributes to be using? So the tool is trying to take into account that aspect of it. Okay, great. Thank you. Um, and another question, if our data lives within Snowflake or another data store, does the analyzer have the option to choose which fields I can select from those data stores and perform this analysis? Uh, that’s gonna be a question for our product team. I don’t know the answer on that. Okay, great. Well, we will follow up with you, Sandeep, on that one. Thank you. Um, and then final question, we do have an option to set the low band

and high band as a relative or percent based on the profile. I don’t see that option in the AI analyzer. Do we know what default option does it consider? It’s doing, um, the percent, not the, uh, absolute. Okay, great. That is it question-wise. Okay. If anyone has any other questions in the chat, feel free to share those now, and we can follow up accordingly. I know there are some points that our product and account teams will be reaching out to with you after the fact. Um, but that is it for Kathy question-wise. Um, Kathy, is there anything else that you wanted to share? We have a few more minutes, but we can always give some time back.

No, I would just say I, I think SSP is an area that I think we all do, and it’s definitely something that we’re trying to see ways to make it better and easier for people to maintain. I know for me it was something to… that took almost a week or two before I started using, um, Zuora Revenue, and I was in spreadsheets, so I know that pain that you’ve been in. And hopefully between this and what’s in our SSP analyzer, would, it helps our customers do what they need to do a lot faster. So I appreciate your time. Uh, lots to come, you know, on the webinars that we’ve got coming up. Please join them. You know, I learn something every time I join them. Yep. Okay. Kathy, thank you again for showing us this. Great. Really excited to get this rolling. Um, we will share the recording in a follow-up

email in a couple days to everyone who registered for this event. Um, and please check out our other events that are coming up. We’re hosting another Zuora Revenue demo on August 20th. I threw the link in the chat, so if you guys are free, we would love to have you back. Um, and if not, we’ll see you next time, hopefully soon. But I hope you guys have a lovely Thursday. Thank you. Thank you all.