Up To 77,000+ Claims Auto-Aggregated: How Mattel Switched To An Automated Mode Using AI-Enabled Deductions

Highradius

Speakers

Up To 77,000+ Claims Auto-Aggregated: How Mattel Switched To An Automated Mode Using AI-Enabled Deductions_highradius_image

Tracy Falkowski

Financial Associate,
Mattel

Transcript

[0:00] Host:

Good morning, everybody. Did you all get a chance to attend the party last night? Yeah, there’s hydration IVs on the field for those of you that need it, and thank you for showing up today and joining us for the session- Deductions Automation: The Journey of Mattel. Our speaker today is Tracy Falkowski. Tracy has been with Mattel for 10 years, working in the collection, deduction management, and credit areas. She was part of Mattel’s original transition to HighRadius deductions module and is involved in its implementations as well as future adoption of collections and cash app. She will discuss Mattel’s success with HighRadius automation. Tracy, the stage is all yours.

[0:45] Tracy Falkowski:

Thank you. Welcome. Hi, I am Tracy Falkowski and I’m here to share Mattel’s journey with the HighRadius deductions module starting from our former, very manual process to our current state of automation and then a future state of further automation and using HighRadius’s artificial intelligence to help decide our deductions. I’ll talk a little bit about our company, a brief history of our manual process before HighRadius, our success with HighRadius’s automation, our future goals with a project called Deductions 2.0 and the artificial intelligence potential for deciding our deductions. I do hope you’ve heard about our company and that your children play with a lot of our toys, and if they don’t, go buy some now. We were founded in 1945 and are headquartered in El Segundo, California. We have many brands including Barbie, Hot Wheels, American Girl, Power Wheels and Fisher Price and I actually work at the Fisher Price campus in Eastern New York which is just south of Buffalo. We do sell globally and have lots of employees worldwide, and also regional shared service centers, processing deductions in Europe, Latin America, Asia Pacific, and North America at Fisher Price. All of our regions currently use the HighRadius deductions module, and will all soon be implementing the cash app and collection modules as well. And we’re really looking forward to having the whole suite together.

[2:31] Tracy Falkowski:

And the results achieved are benefits. When HighRadius went live, it immediately began to pull back up through CPA agents for our largest customers, and also proof of deliveries. For certain carriers, we put into HighRadius as well, and that saved a ton of time of manual work. We were able to come up with a standardized workflow in HighRadius. It did take some time for full adoption, everyone’s still kind of, wanting to do their own thing, but now everyone is on the same page and that saves a lot of time also, it’s just been streamlined. Also, HighRadius allows many of our deductions to be programmed directly to the final processor because the backup is already pulled and attached. So those used to come to us first, we would gather everything and send it along. And now they just go straight to the source and we don’t even have to touch them. Having one place for all our backup makes the research and decision making so much faster and easier. Now our invoice copy, bills of lading, shipping documents, it’s all right in HighRadius for us and as long as the deduction references one of our invoice numbers, that’s all pulled in automatically. So no more going to three different systems to pull that in, it’s very nice. Clearly defined SLS shortly before Mattel used HighRadius, they also decided to outsource some of our work to a third party. With our old system, there was no owner processor like there is in HighRadius. There was just an owner and it was very difficult to know who had responsibility for what deduction between our onshore staff and our offshore staff, and so a lot of things were getting lost in the shuffle. It was just very difficult, we had Excel sheets since trying to assign everything randomly was very difficult. And then reporting is also much easier.

Any time in HighRadius, we can pull a filter or query of whatever information we need, whereas in our old system, if you, if the report didn’t already exist, you had to kind of piece information together to get what you wanted. With HighRadius, it’s much easier to find exports and put into the report that you need. Okay, and then as you can see in the numbers, the benefits of CPA agents for claims in POD saved a tremendous amount of manual time and labor and we’ve been able to reduce staff, that’s mostly through attrition not needing replacement. Some other automation successes with HighRadius- At first, we were still composing our chargeback denial letters in our old format, which was a Word template, and then, we would pull out the documents from HighRadius, put it all together, send it in an email to the customer through individual outlook accounts. Again, very siloed still, it was difficult to find the information. We began utilizing HighRadius’s correspondence templates to create those letters for us and pulling all the backup at the same time. And then a second step to that, a few months later, we made a hard push to get all of our customer contact information loaded into HighRadius, because again, it was in, you know, this person’s email, this Excel sheet for this person. We gathered it all together. Got it all loaded for almost every customer into HighRadius and then we were able to start emailing our decline letters directly from HighRadius and it keeps a copy in there for us and there’s not that extra step of creating it outside the system and then putting it back in. It’s just all done in the system. And as we see changes and contact information, we update them in HighRadius and that’s our one source of good information. This and doing that, doing z the context now is also going to help us when the collection module goes live and we’ll be able to auto-dunning from the system we already have most of our customer base loaded already. We just have to make some tweaks if one contacts us disputes and one contacts us past dues. Through HighRadius in the correspondence you can program responses to go to a certain mailbox., so we created a shared mailbox so that any responses to chargeback letters would go to a place where everyone could see them. And again, they weren’t in an individual’s email that you may not have access to.

[10:08] Tracy Falkowski:

And, we did the same for our requests of backup. We have our third party requesting debit memos from the customers who don’t provide them or don’t have portals, and we use the HighRadius correspondence to do that as well and have that program to go to a specific shared mailbox where they can go and check and see if the customer has responded. And that way, anyone can do that. It’s not collector specific. Anyone can pick up that email, attach it in HighRadius and then move it along to the next step in the process. And that is helpful. Our third party, our staff changes a lot and or they’re out sick a lot, so we want to have full visibility of what they’re working on.

[11:42] Tracy Falkowski:

Okay, to follow up on our successful automation so far, we’re working with HighRadius on a project called deductions 2.0 and it is a global project. Two main goals and an important goal for all of Mattel’s regions is to automate our credit memos through HighRadius. Currently, HighRadius does not communicate with our ordering system, only with our ERP. So HighRadius and Mattel’s IT departments are trying to get that working together. And this will save a lot of manual work for us. Because currently, once we deem deduction valid and we want credit for it, we assign it to our third party, they create a manual Excel template of what they want the credit to be, that they then make a PDF, put all the backup with it, put it back into HighRadius, it goes through the HighRadius approval workflow, which is very helpful, you know, that stays in the system, then it has to come back out of the system for someone to manually key it into our ordering system. And then they go back into HighRadius to put the credit memo number and so it’s a lot of manual back and forth. So if we can automate going right from HighRadius to that ordering system, that’s going to save the template creation and the manual team. So that’s a big thing for us that we’re all looking forward to.

[13:08] Tracy Falkowski:

We can move down.

[13:10] Tracy Falkowski:

And then, well, the second thing is adding CPA agents, the more the better. You know, we have around 18 now, or maybe less and we’re adding 18 more. So anything that saves manual work is a win because that saves us time for other things. And then beyond deductions 2.0, HighRadius approached Mattel US to do a trial run of artificial intelligence decisioning for deductions. AI decisioning will help prioritize and valid deductions, so we don’t have to worry as much about what is most likely valid and low impact. All the deductions would still be reviewed but AI would allow us to focus on the most important with the most potential for repayment. Also, in our current model, an onshore collector has to review every deduction to decide it before assigning it to our third party to either do the credit or the chargeback. And we’re hoping that with AI, we can just pass those low value, you know, valid claims right on to them, you know, hoping that we’re assuming that the AI prediction will be correct, and they can just do the credit memo and clear it for us.

[14:37] Tracy Falkowski:

HighRadius is using our historical decision data to create predictions for both trade and non-trade deductions. During the trial, we’ll see just an additional field for each deduction with their prediction. And then we’ll compare that with how we decide to decide the deduction that’ll be during the trial period, you know, see how much we match with our decisions, and, you know, we’re looking forward to some good predictions.

[15:08] Tracy Falkowski:

I don’t know. How long is that typical track period?

[15:11] Audience:

So it will be around 3-4 months

[15:35] Audience:

Oh, yes, because, you know, before we start predicting what is that if you were to have a certain set of data, which is already classified, so you can hear a set of deductions, which are valid and invalid. And that way that gives you the input to the model till the model starts understanding, you know, in terms of how we can predict, so…

[15:56] Audience:

How much history?

[15:57] Audience:

We generally start with two years, it all depends on the volume of deduction, rather than the time of the deduction because, the more the volume, the better the deducing. But I think, you know, if you’re concerned about all of the applications, right, you can do the history of the data independent of the application.

[16:23] Tracy Falkowski:

And these numbers from HighRadius were based on Mattel’s US historical data. They compare our decisions with valid-invalid, with their AI decisions, and had a 94% accuracy by volume predicted is invalid and 82.35% accuracy invalid by the amount and we hope to see similar figures during our testing. And once in practice, we’re looking forward to AI assisting us in clearing the low-value valid deduction so we can concentrate on high invalid deductions, hopefully, resolve them faster to bring in more money faster. Whenever the current issue is the collection of our chargebacks, it seems that so much time is needed to work the deductions on the front end. You know, once that initial decline is sent out to the customer, kind of gets lost in the shuffle a little bit to go back and you know, keep at it. So we’re hoping that with AI and the predictions and moving the work faster and more efficiently, we’ll have more time to follow up on those chargebacks and collect more money.

[18:02] Tracy Falkowski:

Do you mean just for AI automation? Or do you mean HighRadius in general?

[18:24] Tracy Falkowski:

Well, I will say that I was not part of that decision-making process. I do know that whenever SVPs was a big fan of HighRadius, and it was always his goal to get us on it, and I think he was the driving force behind it and all of his reasons, I’m not sure. But I do know that he was a big cheerleader of HighRadius. Well, I think further, you know, headcount reduction, we hope to reduce our offshore headcount reduction. As more automated, you know, they are our automation in some respects now, you know, they do up those credit memos, they do oppose decline letters, the more that we can get automated and have that done for us, then you know, we can reduce them even more. And then we’re like I said, we’re also getting the collection module and the cash app module, which will all work together to reduce even more time. It was just using the deduction module, huge savings, huge, you know, made everything easier. But we are really looking forward to using all three of those together because that’s really going to streamline everything. But finally, you know, these are some of the figures that HighRadius provided that we hope were consistent during our trial. We’re just basically waiting for HighRadius to flip the switch to show us another column on our screen that will say what they think the decision should be. And then we’re going to see what we think our decision should be and then we’re going to see how well we match up to that. And I hope that soon, right? Yeah. And questions? Any other questions? Yes.

[20:20] Tracy Falkowski:

Yes, not automatically. We do still have to, like, hit a few buttons to create it but then it goes right out through the system and retains a copy in the system.

[20:42] Tracy Falkowski:

Not for us, but that is possible. I think it’s called Web Push.

[21:00] Tracy Falkowski:

We want it, but we don’t have it yet.

[21:12] Tracy Falkowski:

To come back into HighRadius? Well, as I understand it’s not available in just the deductions module but when we get the collections module, it is. So that’s why we’re really looking forward to more things that we can do when we get the other piece of the puzzle.

[21:29] Audience:

We just lost a million bucks.This is one place where all teammates would be searching. There’s a credit view to look at all the correspondences, so that way, the energy to send out emails from the solution, but then we do the attention to go back to your inbox, you know, to get some escape and now half of a million bucks part of the solution to the update on this one. So it’s also attached to the case.

[22:14] Audience:

So I mean that’s one of the biggest advantages to come.

[22:29] Tracy Falkowski:

Turn it up.

[22:39] Audience :

The customer provides the backup before they make the payment in a short payment and identify the backup and attach it as encoding.

[22:50] Tracy Falkowski:

We do have backup regulations coming into a shared outlook mailbox. We do not currently have any email parsing set up for the customers that’s like on our list of to do’s, but it can be done.

[23:14] Audience:

So the research, it wouldn’t be automatic…

[23:21] Audience:

What is the difference between, we have a claims mailbox that’s shared with an organization and then HighRadius takes the backup and puts it on the deduction? Is that a difference?

[23:36] Audience:

So there are two ways to look at it, right? One is, the customers would send out the claims or documents. That is where it needs to pass all the documents and then attach it to a specific deduction. You’re communicating as a company, you want to have some idea of communicating or collaborating with someone outside. It’s a request of course if I’m talking to someone outside. So that is a really locks that we’re talking about to some of these points back on that, you know, then it gets back to the same introduction it takes time to actually-

[24:13] Audience:

Shortcuts to one agent automation accessory. How long does it take to get to the action section?

[24:28] Tracy Falkowski:

Well, he could give official timelines from just personal experience.

[24:34] Tracy Falkowski:

We’ve been pretty lucky with our portals every so often you know, passwords expire and that might delay you a day until you change the password and you know, go into the system into HighRadius and put the new password in. So you might have like a day delay while that’s happening. Occasionally portals change how they have to be maneuvered and HighRadius usually knows about those big ones before we do and they’re working on Target or Amazon. Occasionally, we see a change and we notify them and then they change the script, so you might have a day or two lag there as well. I do have some staff who always think the CPA agents are broken, because things may not match up. But it’s because they don’t follow the rules exactly. Amazon’s one, I don’t know if people will get Amazon’s zero-dollar checks, where they take their benefit deduction, and they break it up into pieces and take it on like three different checks so they have to pay a zero. Well, the CPA agent pulls in the claim for $500,000, and then when the deduction comes in, the first one is $25,000 and the next one is $250,000. Here, they’re all different amounts. So the system can’t automatically merge them because the reference number and the dollar amount didn’t match exactly. So they always think the backup didn’t pull. It’s not there. It’s not there. It is there. It’s just in the down, the pre deduction, you can find it with a reference number, it just didn’t automatically merge on to the actual deduction. So it does take some education of staff to understand where to go to find things if it didn’t match perfectly, but for the most part, we’ve had good luck with the CPA agents.

[27:01] Audience:

But you have to be close enough in the partnership and manage it closely in order to handle the changes. So it doesn’t change the photo, you have to be there because you are the owner. You have to work with them to give them what they need in order to really take care of that change. If I may add one, just one question from our side. You know, we go to these extreme lengths to make sure we are monitoring the agents all the time.

[27:46] Audience:

You know, the configuration on their side, or if the password gets expired in stock, but it’s not a problem with an agent per se, but it’s just that you know, a human trying to access that, we have a problem with that. So from our side, there is a lot of proactive monitoring to make sure that things are running a business.

[28:14] Tracy Falkowski:

Yes. Yeah, we get notifications when it expires.

[28:24] Tracy Falkowski:

Yeah, we do have that. I can picture it on the screen.

[28:33] Tracy Falkowski:

Did you have a question over there?

[28:37] Audience:

Just commenting on the topic. The topic I was just discussing, how fast agents will get modified when something changes, and in your experience, one to two days is typically what it takes HighRadius to.

[28:52] Tracy Falkowski:

Sometimes last, but I’d say that’s probably the most times.

[29:09] Tracy Falkowski:

Carrier portals, sometimes those PODs take longer because the, you know, it goes in searches and it’s not available and then it searches again until it hits it. So sometimes there are delays in that. The actual agent, okay.

[29:25] Audience:

The agents need to be reworked. Depends on how drastic the changes are, what is their volume in terms of how many programs they have to change, right? So, if the change is massive, for let’s say, a major client that has a lot of activity with directors with multiple clients, they’re going to have to make that change to all sorts of programs they have. But there is also synergy because now they know what needs to be applied in order to make that change across the board. So what I think is, will emerge in the industry, it’s gonna be kind of a connector between HighRadius and the big place where they don’t allow HighRadius to come in and pull information. You can post it because it’s a business reality, right? They’re doing it for so many families. So I think it’s going to emerge into that level of relationship management between the major players.

[30:47] Host:

And I’m just going to make an announcement that the session is officially over. You can stay in, have like a sidebar, for sure. I just want to know, like if y’all have a different session to go through after, now would probably be the time to leave.

[31:01] Tracy Falkowski:

Well, thank you, everyone.

[0:00] Host: Good morning, everybody. Did you all get a chance to attend the party last night? Yeah, there’s hydration IVs on the field for those of you that need it, and thank you for showing up today and joining us for the session- Deductions Automation: The Journey of Mattel. Our speaker today is Tracy Falkowski. Tracy has been with Mattel for 10 years, working in the collection, deduction management, and credit areas. She was part of Mattel’s original transition to HighRadius deductions module and is involved in its implementations as well as future adoption of collections and cash app. She will discuss Mattel’s success with HighRadius automation. Tracy, the stage is all yours. [0:45] Tracy Falkowski: Thank you. Welcome. Hi, I am Tracy Falkowski and I’m here to share Mattel’s journey with the HighRadius deductions module starting from our former, very manual process to our current state of automation and then a future state of further automation and using HighRadius’s artificial intelligence to help decide our deductions. I’ll talk a little bit about our company, a brief history of our manual process before HighRadius, our success with HighRadius’s automation, our future goals with a project called Deductions 2.0 and…

What you'll learn

  • Understanding the real-life use cases of AI in Deductions for Mattel
  • How AI can help in accurate deduction research recommendations

There’s no time like the present

Get a Demo of Deductions Cloud for Your Business

Request a Demo

Request Demo Character Man

HighRadius Deductions Software acts as a powerhouse for proactive deduction management to prevent bottom-line erosion. It provides automation, process standardization, and a platform for cross-departmental and customer collaboration. It supports deduction management by providing some key features like back-up document capture which captures deduction data from customers and supplies the information required for resolution; auto-capture proofs of delivery (PODs), bills of lading (BOLs) from carrier portals & emails; structured deduction resolution, collaboration & approval workflows to streamline the communication and approval process; along with automatic deduction correspondence, and automatic data push to customer portals. The result is a proactive deduction management operation that recovers revenue normally lost to invalid deductions.