Benchmarking, Project Planning, and Performance Tracking for Centralized Operations

[0:00] Krutarth Kousambi:

Good morning everyone. We have quite an audience. It seems that everyone was drinking last night. Alright, so we’ll start with the presentation.

[0:14] Krutarth Kousambi:

So this is a story, I became a full-time employee of Prime Source before six years and at that point in time, we had a decentralized credit and collections and from there we went into the decentralized model. It’s the story that I’m going to share with you guys what kind of challenges we had and how we overcame it.

[0:42] Krutarth Kousambi:

So few things about Prime Source. We have 34 distribution centers. Prime Source is the largest building materials distributor in the United States. We have around 1200 employees. If you go to Home Depot and Lowes those are our biggest customers. You would see a lot of brands called GripRite, ProTwist, GR-X, those are all our products. So we have thousands of materials and you know, fasteners or insulation products, wiring products, fencing, all kind of building materials that you would see.

[1:25] Krutarth Kousambi:

So Prime Source’s plan for AR transformation, right? This started somewhere before four years where we were in a decentralized operation for collections, credit, and disputes. So there are three main stages that we went through during this AR transformation. The first was obviously benchmarking. Once the benchmarking was done, it was how we planned the project and you know, how we went through the implementation. We use SAP as our ERP system. And at last, it’s evaluation and what kind of results we got. And we got some pretty good results, we had reduced DSO, there’s faster dispute resolution, more managerial visibility on what’s going on, what’s not, obviously, relocation of resources.

[2:17] Krutarth Kousambi:

And in general, the reporting, and, you know, I mean, where we were before. And at the end, when we were centralized where we are right now. I mean, it’s a huge difference. We’ll start with the benefits of a streamlined transformation. I always liked a phased implementation approach. It will give you milestones, it will give you dates that you kind of need to, you know, make sure what your scope is and you stay within that scope. We started with benchmarking probably four years ago, where we identified that hey, we wanted to make this move. What is it going to look like? What are the troubles that we will go through? There were some really bad metrics backed at that point. Everyone was very confused about what was going on. There were no reports that would show who worked on what. It was all manual, it was all through emails. So it becomes really hard. When a credit specialist or collections specialist says that, hey, you know, I mean, I worked on 300 customers today, right? But if it’s all email, I mean, how do you track it back? If they say that, hey, you know, I mean, we worked on 100 disputes. And we emailed it to either the DC managers or the sales guys for what’s happening, what’s not and no one replied back. So there was no ownership back then.

[3:51] Krutarth Kousambi:

We defined what our goals would be, what are the shortcomings and you know, what do we want to see in the future. I’ll come to that later on. Once we had defined what the objectives were, we kind of sat down and defined the scope of the project that these are the six or seven main things which we wanted to make sure that they are the ones that we target and how we are going to evaluate them later on. Within the project planning phase, we used HighRadius. Me and my business analysts back then, kind of for two years or more years followed HighRadius seminars and webinars just to see what products are out there and does it make sense to make that jump. There were some other companies as well we were looking at just kind of started following what’s out there. What can help in our situation?

[4:50] Krutarth Kousambi:

Later on, it’s an evaluation where obviously it was hard to measure just after we went live but today, it has been 3 years that we have been live with the modules that we went forward to, from the standard SAP collections and disputes we had back then. And when we see it, and later on, I’ll show some metrics on it. The last thing is obviously, you know, identifying the key areas that need more focus now. We have some future projects coming on as well. And we kind of want to see how we can do something better in those areas as well. So AR landscape, I’ll start with what we had before when we were decentralized, right? So we have decentralized as I said, we have around 34 locations back then it was 36 distribution centers. At each distribution center, we used to have one credit manager or credit analyst, right. They were responsible to see what happened at the soul 2 levels. What went wrong Was it the pricing? Was that something damaged? Something was not delivered? So 36 people right there working, they’ll have worklist collections where it is and we used to use standard collections and disputes in SAP, we use to use that model. They review the credit blocks, every now and then, you know, the customer will keep going into credit blocks because of the rules that we had in place and they were not very good rules. All the disputes within our distribution center were being worked by that particular credit manager or the analyst. And on an average, you know, the dispute cases used to be like, hundred and possibly the worklist for the collections call used to be around hundred to 200 as well.

[6:52] Krutarth Kousambi:

On the corporate side, here in Irving, We used to have six corporate collections specialists. They kind of overlooked everything as well as whatever was, was corporate-related stuff. And if there were any disputes or collections related to calling the corporate office. And both the teams together kind of looked at customer maintenance on what needs to happen if it sold to level it would be the distribution center, credit manager analyst. If it was something related to the payers, that would be the corporate credit. So this was the team structure before when we were decentralized.

[7:34] Krutarth Kousambi:

Some challenges that we had during that time was there was no standard process, right. So we have 36 individuals in the distribution centers, we have six here, so almost 42 people, everyone had their own style of working. Everyone has their own style of writing an email. Some are maybe more formal, some are more strict. Some are not formal at all. So there was no standardization on how we are going to approach the customers. No clear visibility for the DC teams, because here at the corporate office, the manager or the director wouldn’t be able to know how much is being worked on, right? Everyone would say that we have emailed them because an email was the main source, everything happened through email, be it contact to the customer or be it contact internally to the sales team. Or to find out where the proof of delivery is any kind of documentation backup, it was all emails and back and forth. Sometimes we wouldn’t get the response that we wanted from the sales team as well. They will be busy doing their thing. So how do we say that hey, you know, I mean, this is the workload that you guys have, we’re waiting on you guys for some kind of information and give that information back to us. So definitely, that was lack of manager visibility, no reports. Nothing to run where we can say that has someone worked? Has someone not worked? Are the internal teams who are supposed to get back to us with the kind of documentation or with the kind of resolution, they might not be coming back to us, right? So everything became very time-consuming. It was not coordinated at all. And there was a time when we decided that hey, you know, I mean instead of being a decentralized credit and collections, we are going to centralize it.

[9:32] Krutarth Kousambi:

Going to go through how we redesigned the AR process. The first thing I said right, I mean, there are three major parts to it, benchmarking, project planning and also implementation and then, later on, the evaluation. So we made the switch from decentralized to centralized, right? The centralized team looked like this. So from 42 people, we now had only 11 credit and collection specialists, one manager, one supervisor or the lead. Each specialist kind of got three to five DCs now. So they’re obviously their workload increased a lot from what it used to be before. And there were two people who were just looking at the customer maintenance, the customer master. So when we made the switch, we started to get even more challenges, right? And the challenges were there. Obviously from 42 people when you go to 11 people and each person is now looking at three or five DCs, their worklist increased a lot. We were at the soul 2 levels so what used to happen is that you have this one payer let’s say Walmart and you have all the stores that are sold to thousands of stores before since it was decentralized, whatever store that particular distribution center catered to was the one they used to call, Right? So when that happens if you would be calling multiple times, to Walmart, and different people like all 35,36 credit people would be calling them to ask for their set of invoices, their set of disputes their sort of collection invoices.

[11:19] Krutarth Kousambi:

The deductions and by deductions, I also mean that it generally results in a dispute. That dispute increased as well. So it increased four to five times for each specialist. So, the increase in workloads, the volume increase, every person was kind of frustrated that we have to look at so much more than we were then we used to. We had as I said, read collect collections, SAP collections, and SAP disputes. But the way it was set up was very simplistic, you know, I mean, it didn’t have any complex rules or anything that would help the analyst out. Dispute management was handled outside SAP and as I said, right, I mean, there were just emails back and forth. Looking for someone to answer back and say that Yeah, you know, I mean, the pricing was incorrect. The product was damaged. We do have a proof of delivery, here it is. A lot of the times we wouldn’t get the answers back in time.

[12:25] Krutarth Kousambi:

No manager visibility. And obviously, there was dipping productivity since everyone thought that, you know, from before we are doing so much work. So we started to define the objectives of what are the things that we want, right? I mean, where are the areas where we are really failing? What is it that we want to see? Some KPIs that we thought were important were obviously faster dispute resolution, we would have disputes sitting for three years, four years and there won’t be any resolution. There would be employee turnover. People will change. And it would become even more difficult to get the answers.

[13:06] Krutarth Kousambi:

And improve analysts’ productivity, as I said, I mean, there was no way to reach that big of a worklist if you get a thousand customers, how are you going to make it right? Even if you get a week to work on it, you’re still going to feel that you’re not making an impact on the organization.

[13:24] Krutarth Kousambi:

End to end visibility was one of the main challenges for the manager and the director, right? Because there is nothing there is no one report or no one system where they can just go and say that hey, you know, I mean, this is the average. This is where the dispute case stands as of now and who needs to act on it now. So streamline centralized operations. We wanted to standardize the process. Everyone used to do things on their own. So that was something which was definitely needed, someone would just write a customer a very informal letter, someone would write a formal letter, the tone might not be right. So we wanted to fix that as well and improve the customer experience and relationship because when you have all these people getting to the same customer from different directions, it creates a problem because you’re 35,36 people calling the same customer but only for one set of invoices, one set of deductions, that kind of is rated to their distribution center and not for the whole company.

[14:38] Krutarth Kousambi:

So our action plan was to definitely go back and check how the AR was set up and how we can change it. So we completely redesigned the AR process. We wanted to see where the bottlenecks were, what needed to be done to be where we want it to be. The Productivity as I said the morale was going down, everyone was working more than they should be. The collections worklist increased. The dispute cases increased as well. Faster dispute resolution, as I said, some of the dispute cases would take a year, two years, three years, we will have four years and still, the disputes are, are just sitting there and have not been working. And the last one obviously, was that hey, if everything gets together if we are able to do all those three things, how are we going to go back and check and evaluate that what we have done is as has been correct or not and if we are getting the information, which we need.

[15:43] Krutarth Kousambi:

So we go to the project planning phase, right. And as I said, we went back and we kind of started to redesign the whole process. We split it into two parts, collections, what are the things in collections that we want? What are the things in deductions that we want to have right? We wanted to first redefine the customer segmentation. We had different risk classes, different groups. But we didn’t have a solid strategy for collections. So all the groups, all the risk classes across the board, there was only one rule. And that rule was just through the aging buckets. That becomes really hard because now you might have some good customers, you might have some bad customers. But when SAP is giving you a collections work list, it doesn’t know which one to approach first and which one not to approach first.

[16:36] Krutarth Kousambi:

So we improved our collections through the as I said, rate, I mean, it was very basic, it was aging-related and I’ll tell you some of the rules that we got and how it helped us, in some cases to be proactive as well because you probably want to be proactive, not to customers who are always paying but for the customers who you think are trouble customers, right? That’s where you want to be proactive. How do we prioritize the worklist? Because across your customer segmentation, you just have that one rule and that one rule fits all. Just give you customers, which you probably don’t want to call because they would pay anyways right? Or you don’t need that much of an effort, but the collection workers would still bring it up. Because all you’re caring about is the aging buckets. We made a change to how our collections worklist was being generated. And as I said, right, we have 10 to 12,000 payers, right? So 10 to 12,000 head offices. We have 60,000 customers at the soul 2 levels, right 60,000 customers, instead of being at a soul level, which we were before we made it a Peer level, so that we only have to worry about the 10,000 or the 12,000 peers that we have.

[18:08] Krutarth Kousambi:

So straight out of the box, as SAP doesn’t have a report or you know, I mean strong reports on collections, on disputes. So we customized some of the reports on how we can see who worked on what, how many customers were reached, what was the result? And how much dollars that they’d come back against whatever was open at that time. So there was something we had to customize, even with the add ons that we had from HighRadius, but with that in place, now we are able to tell what’s going on and what’s not on the deduction side.

[18:50] Krutarth Kousambi:

As I said, we wanted to make sure that the inter interdepartmental communication is good, because the emails would be saying they would sit there someone would be busy, somebody would not reply back, even after your email three or four times, and it would become a pain. How do we make sure that once an item is disputed, right at that point in time, we are on it? We explained to the customer that hey, you know, I mean, whatever you were asking for, is probably not true or it’s true, we might have proof of deliveries, we might have invoice copies, which we can send directly through the system and not go back to the billing document, take a copy then, you know, save it as a PDF, you attach it, you send it to the customer. It was just so many steps at that point, that it became really hard.

[19:44] Krutarth Kousambi:

So with that being done, obviously we want to reduce the manual work and focus more on research and resolution. One of the things which we didn’t deduct was also the fact that all these customers right I mean, 12,000 customers, some of bigger customers have web portals. Or they would be sending us stuff through EDI, EDI 812, which is the debits and credit adjustments. So they are letting you know that hey, on a $100 invoice, I’m only going to pay you 90 and I’m not going to pay you $10 because of this reason. If you have 2030 portals, everyone would have to go login into their ID search for the invoice that same question. Download or take a snapshot of that particular reason why they’re not paying those $10 attachments and then maybe send it either for right or for some kind of an activity which would become your next step. Right.

[20:43] Krutarth Kousambi:

At the same time audit compliance because if you have workflows if you have a way to show that, hey, this case went from here to here, this is how it got approved. This is the note on it. And that’s why it’s been written off to this GL audit becomes a little easy then what it is, I mean, we, right now have an audit going on and like every day we have requests coming in what happened, what happened and we just have to, you know, dig down and find answers to them. So this definitely helped.

[21:17] Krutarth Kousambi:

I kinda want to go into how we made those changes at each section, right? So all I was talking about right now is that, hey, we wanted to do this, we wanted to do that this is how it should look. What did we do, which made that change to work? Improved collections rule, we only had aging buckets. So something if it’s between 30 to 60 days overdue. SAP would assign at a 25% rate, right? If there are some other invoices that are overdue, from let’s say 60 to 128 would assign another 25 points. It would become 50 points, a hundred points. But then the problem with that is that you might have smaller dispute cases, you might have dispute cases, just 2 dispute cases, let’s say which are about 120 days, but only having $30. Right? So you have two invoices with $30. But your rule is still going to pick it up because you have only set it up according to the aging buckets. It’s going to pick it up, it’s going to show it up to you and the worklist. Now $30 might not even be a problem to you and you probably don’t want your analyst to look at that line first. You probably want to bring it to lower down when you’re done with the bigger amounts first.

[22:44] Krutarth Kousambi:

So we had created different rules. Sometimes it was that hey, only 25,000, only 75,000 of the total invoices that are in your current AR become overdue only then assign some points to the strategy, right? And when you assign the points to strategy, you kind of have the max of hundred, right? You have a max of hundred, let’s say you have seven or eight different rules, each rule would be given some kind of points. So if I say that in one of my rules is that if the overdue items are more than $70,000, then break that rule so that that rule would be broken, and whatever points have been assigned to that particular rule 25 points, it will start getting 25 points, right? A second rule might be that if someone has promised you to pay so you have a promise to pay in your SAP system, someone broke it so someone did not pay for those invoices even after the promise to pay was broken. You probably want to assign it another 15 points, right that becomes 25 to 15.

[23:51] Krutarth Kousambi:

We also created some proactive rules for as I said for late customers, who were in that bad risk class. The reason being that you kind of want to ask them beforehand, you probably want to reach them a day or two before you think that the invoices are going to be overdue because that’s where you really want you really want your analyst to go ahead and give them a call or send them an email a day or two before in that particular rule reset that, hey, you know, I mean, if this particular amount is being overdue, within the next two or three days, right, then that rules should be broken. So when that rule breaks, and it comes up in your workplace because otherwise the rules are not broken. SAP collections worthless didn’t don’t show it to you unless any one of the rules is broken, right? So that’s how we started to become proactive. So we used all these different collection rules and assigned them points. So that when different rules are broken, and if all the rules are broken, it would assign more points and be higher up in the collections worklist, so that whoever is working, that particular worklist would know that hey, you know, I’m and this is my top priority. And then, later on, he’ll come down to a place where it’s only, let’s say, 10 points, right? If it’s 10 points, it’s probably a rule which got broken, but at the same time, the other rules were not broken. So you will probably find it.

[25:30] Krutarth Kousambi:

The second thing we did was automated dunning. So as I said before, everyone has their own way of writing an email. And when you reach out to your customers or even reach out to your internal department, it sometimes doesn’t look professional sometimes. It depends on how the other person takes the tone of your email, and it creates a problem. So with automated Dunning, what we did was in collections, we created seven or eight standardized letters, which were approved by our HR or your legal team if they’re going out. And at the same time, the ad hocs helped because we had those letters standardized, but at the same time you have the ability to go ahead and write a statement or two at the start if you wanted to personalize it as well, but the main content kind of remained the same across the board right. And we started to automate it when we started to automate it obviously we didn’t automate it for each and every customer because it would be very challenging situation where you have not been sending this Dunning letters and Dunning letters depending on the level of Dunning the first time when you say to someone that hey, you know, I mean, you need to pay me you’re probably very polite at saying that and would say that, hey, you know, this is the amount that you need need. You guys need to pay me the second time. You’re You know, your letter kind of becomes more professional and more strict In saying that, hey, you know, I mean, you already have this much past you please pay, right? The third time it becomes that hey, you know, I mean now if you don’t pay we might have to talk with our legal team. So we have these dunning levels. And for each dunning level, we created some kind of rules on why the wordings would be what it would be. And we automated them kind of slowly. So we just had the first letter sent out first for at least six or seven months. We didn’t even turn on the second, the second notice or third notice or even the legal letter. There. The reason as I said right, your customer is not used to looking at it. And if you just turn it on and start sending first reminders, second reminders, third reminders, Legal Notices, they possibly are going to freak out. So I mean, we did it very slow for the first six months. We just had the first reminder being sent, we turned on the second rule where we would send the second reminder. The third reminder, the legal letter is obviously something which you don’t want to automate, that is something which you just want to do yourself. But this helped a lot because now you don’t need a credit analyst to send emails, the job would run the job automatically from your Outlook, go and send all those letters. It will have a table where it would show you that these are the invoices that are overdue and needs to be paid.

[28:36] Krutarth Kousambi:

The third point is insightful reports on the production side of dispute management, we didn’t have strong reason codes. We have some we had some very, very common reason codes and root cause codes. We do want to make sure that obviously, you don’t have like 50,000 votes right, but having a healthy ratio like 50 into 20 good reason codes on why your dispute case was created and what was the root cause for it really helps right. So, it might be that your root cause is billing or it can be that it had something to do with pricing issues it had to do with damaged products it had to do with something not delivered within the sales order. So maybe you just felt 95% of the sales order but forgot 5% of items.

[29:33] Krutarth Kousambi:

Backup document capture as I mentioned before, right, we had like 20 different portals, 20 different customers, all big customers, every time the analyst has to go and type down and then look at the invoice which is in question. Go back and take that document or take the screenshot to save it to attach it to the dispute case, so that it can be processed further. What this does, and I’m going to give the name of the add ons that we used right after this line. And what it does is it goes back and checks for the invoice number of checks for the reference automatically logs on to those customer portals. And it downloads whatever document is there and at the same time, it attaches itself to the SMP dispute on the left-hand side. So on the left-hand side now you can see the documents that are already attached the credit analyst in didn’t have to go to that website and find what was going on with the invoice and create a screenshot so it’s already there, it becomes easy because now the credit analyst just needs to double click on the document and see what’s going on and start to see what’s going on further.

[30:56] Krutarth Kousambi:

So and the three modules that we use I’m just going to tell you what the three add on modules from HighRadius is. We already had SAP standard collections and disputes, right? We changed the design rate, we changed the reason codes, we changed the root cause codes for the dispute side on how we wanted it. On the collection side, on the standard side, we went ahead and changed all the collection rules. Now, there is an accelerator called disputes and collections accelerator, which is an add on onto SAP. What that did was that on your dispute case screen, on the left-hand side, it would give you the full document flow. So payment has happened, you probably have a residual item remaining. You can now track back what the payment document was, you can track back what the billing document was, what the delivery was and what the sales order was. Instead of going back and going to the original document and trying to find what the sales order was. It’s all the links that are already on the left-hand side so you can just click that link and be there. This saved a lot of time. Disputes and collection collections accelerator. Yes. The other, so as I said, right, I mean, the check images are on the left-hand side, the backups are on the left-hand side, you have the document flow within SAP which is on the left-hand side. Everything is right there now and now it becomes so easy for the analysts to work on it.

[32:29] Krutarth Kousambi:

The second module is ACA which is called the advanced correspondence automation. What that does is within your collections worklist UDM_specialist and supervisor and UDM_dispute. It gives you a standardized letter. So I said right, we created seven standardized letters for collections, seven standard regular letters for disputes. You can pick which one you want to send right there. And you can click Send and it will collect your outlook. And the Go out, no need to type in and write stuff and create those tabular formats or anything like that.

[33:08] Krutarth Kousambi:

And the third one is the CPA, which is the claims and POD automation. Right? So the claims and POD automation is the one where I’m saying that, hey, you’re on the left-hand side, the engine automatically goes to all those web portals or to the EDI files, it fetches the information. And it either converts it into PDF or whatever the document type is when they download it, and it gives you a link right there so that you can click and you would know what happened and why was there a debit and credit adjustment?

[33:41] Krutarth Kousambi:

Some of the results were achieved so we went live before three years. We have a reduced DSO on an average of 10 days. That’s a lot. For some of our customer segmentation, in fact, the DSO is, you know, we were able to gain 18 days and a few of our customer segments. Faster dispute resolution, we have 40 to 50% of the time that we have saved from what it was before. We have a very streamlined collection process. Everyone knows what the worklist is, what kind of rules are broken, why the customer is on the top and not at the bottom? Right. We have the relocation of resources as I said from 36 people in DC and 11 people working. Sorry, six people working when we were decentralized. We came to 11 and last month our credit team kind of had a realignment again and right now we’re just six people right. So I mean, just that is to six analysts, one supervisor, and one manager. The visibility as I said right now you have all these workflows, you have logs that come in, it’s very easy to see what happened to the dispute case before the dispute case was only for the people who were in disputes. And it wouldn’t go out to let’s say a sales manager. What we did with the DCA module is that we incorporated them as well. So if we have a national team, we have a major team, we have a sales team, we have some delivery people who might be on the operation side, everyone is now on that particular model. And what the dispute analysts can do is directly send it to that particular DC manager, for example, the distribution center manager right. What this does is now you can run a report and see that somebody sees they are working faster, somebody sees managers, their admins are working fasters, the others are not replying back. And you can tell them that hey, go ahead to this particular t code and start to work on those cases. Because there are 300 things which are depending on your side, whereas other DCs are working it on time, and probably they are keeping up with the stuff, so you can pinpoint who the owner is and who needs to work on it. Right?

[36:12] Krutarth Kousambi:

In the future scope, I’m going to complete it. How much time is remaining within? Okay. All right. So future scope, we’re going to do the same thing with credit processes. Right now we have problems with a lot of customers going to credit blocks. Because the rules aren’t sophisticated enough. So this is something which PrimeSource now wants to focus on. And we are working with HighRadius to see some demos as well on what the cloud offering is, what are some other kinds of offerings which might help us automate the credit limits by looking at the past behavior, by looking at the volume of the sales or how big a customer is? Credit agency data aggregation we do it right now in Excel it’s not integrated but at that point, if D&B or some kind of rating agency would be integrated directly in SAP the person doesn’t have to go somewhere else. And definitely credit management workflow right and right now it’s manual. We want to make sure that it’s an SAP and and and it’s as good as how it is in collections and disputes.

[37:28] Krutarth Kousambi:

Some lessons learned obviously vendor selection. I was very happy with HighRadius, not that I’m getting paid. But but but they were the first ones who kind of made sure that you know, whatever the timelines were the timelines, the scope remain the kind of the same there was not you know, a lot of back and forth, made sure that you know, each and every time we are going in the right directions are short, some industry best practice practices, which helped us a lot. So definitely vendor selection. The last thing was the day to give documentation. I have worked with multiple vendors. No one gives back documentation saying that, hey, these are all the changes we did and decide what the technical changes we did never received it. So I was very happy the first time I received it, and I was like, wow, this is something that is definitely going to help. Lasting phased implementation. We need to have phased implementation. If you don’t keep changing, there are so many iterations you just are not able to define the scope. And that really becomes a problem. So with that, anyone has any questions and anything that I can deep dive into?

[38:46] Krutarth Kousambi:

Alright, thanks a lot, guys. Have a great day. Thank you.