LEDES Billing Kickbacks? How AI Can Help You Fix Them Before You Upload

Hi there. Welcome back. Have you ever had this happen to you? Maybe you're new to doing leads billing, or maybe you are a seasoned bookkeeper who now the firm that you're working with is telling you that you have to now send this particular client a LEDES bill. And LEDES is spelled L-E-D-E-S. It is an acronym, of course, right? Law firm acronym. But it's a specialized program for text to come out so that a LEDES billing system can read the bill and quickly process it. So imagine that you just hit upload on a LEDES file. You put it up in a legal tracker, you walk and go get a coffee while it's processing. And the worst scenario happens. You come back and you have a kickback email that just, it's going to take up your whole afternoon trying to figure out what happened and what's wrong with the bill. This is a job that a lot of your billing staff and bookkeepers who support law firms have to deal with on a monthly basis or whenever they do the billing. Legal tracker and other portals, and not to pick on legal tracker, but other portals like this have messages that come out that explain why the invoice bounced. But a lot of times it's very confusing and hard to understand. So there's a long narrative, a long explanation of why the bill didn't process. And we as bookkeepers have to go through it and figure out exactly why it didn't go through and how we can make the corrections on it to help it go through. So it could be a myriad of reasons why it didn't go through. It could be maybe the way they wrote the words in the paragraphs and the descriptive lines were wrong. Maybe there's a semicolon in there. Yeah, that actually happened to our firm where the attorneys were typing in semicolons into the body of the text of the bill. And what happened was this particular tracking system didn't like that. It didn't like that you had these in the system. It could be the wrong UTBMS code, right? Maybe block billing caused it. A rate that doesn't match the card can cause it. Your attorney already signed off. The client's expecting the bill. There's pressure there. And really, I'm going to let you know that you can let AI help you clean up that file before you do a secondary upload of it. And it's still a job that a bookkeeper is going to have to do. You're going to need those human eyes on it to make this process happen, at least in today's world. But AI can help you clean it up quickly. So what does a kickback feel like? A kickback is not a gentle nudge. You could have uploaded maybe 20 bills in the system, and then it's rejecting invoices in the queue. Sometimes the whole bill can fail, and sometimes just one line can stall everything. You reopen the leads export, you dig through the lines in Clio and LeanLaw have created in the leads bill. You hunt for the guideline memo. That maybe somebody saved maybe years ago that told you how to do this leads billing. And then you compare your rates, you compare them in QBO, LeanLaw, Clio, you compare the rates. Maybe there's a rate problem. Maybe the rates were supposed to go up or maybe they went up and they shouldn't have gone up. Anything like that can really make it a lot easier to find. You're trying to find the issue, find the problem. And sometimes the words you get back on the kickback email doesn't make a lot of sense. Then you end up on the phone with support trying to figure it out. So the detective work is really on your shoulders and you're working under a deadline. And AI is not going to take that ownership away, but it can take a lot of that rewriting out of the way and maybe even help you take that work off of your plate to pre-check the wording if you're not using a system that converts to leads. So what are the common rejection causes? There's a program out there that you can actually purchase that will analyze a lead's billing and help you do it before you upload it. I think it's Elite Billing. They have an eHub. You can use that. But they actually wrote an article that showed that invoice rejection rates are rising. So they were only 11% in 2024, and then they went to 18% in 2025. And while the days to payment From 62 to 50. So the client's getting paid faster, 'cause that's another part of it. You could have budgets that you have to deal with, and is it in the budget? There's a lot of pieces that go into this type of billing. And sometimes it's just maybe the legal tracker system or the billing portal that you're uploading to, they make you put in a special code for the law firm that you don't have in the system. It can be something like that. So a lot of times you get back the message that says something's wrong. It's an attention to the matter. You need to figure out what happened. You can write it back saying that. An AI can rewrite what the timekeeper actually wrote in a way that will be accepted. Or maybe the note's empty. You can flag the timekeeper. Please let me know what work was done here. There was no work in the body of the context of the bill. UTBMS mismatches can really wreak havoc, right? So the narrative will require you to do some research. Your code may say court appearance, but maybe you were doing research and development. Who knows? It could be anything. It could be copying and you picked court appearance. So you can ask AI to flag any disagreements and then you can pick the right code. You can go through the list, usually inside of Clio and LeanLaw, there's a list that'll give you what they are. And you might have to just correct it. Maybe somebody mistyped it. Then there's block billing. So one entry that lumps all the calls, drafting, and research all in one lump. Some of the programs don't like that. They hate it. So AI can suggest some splits when the text already distinctly speaks to a specific task. It can't invent separate timing. So you're going to have to work with the timekeeper on that, but you can't just write that you did all these things in one big lump. And expect the system to pick that up and pay for that. It will kick that back. So rate cards are another place that if the rates are wrong or mismatched, it will kick back. If the timekeeper that did the work is not the timekeeper that was approved, can kick that back. Keep a current rate table so you know what you're supposed to have. Match the rules and let AI take that kickback. message, put it in there and ask it to please explain this to me like I don't know anything about legal billing. I don't understand what this means. Can you explain this to me in clear English of what is wrong on the leads file? You can look at it. It could be format issues. Like I said, the semicolon, right? Any missing fields, any bad dates, any expense lines without detail. Those are all going to kick back. As you can see, this is a really intricate part of the job that we do. And it's important. And you want to use real leads, a real leads checker for structure if you're having to get that. You can use AI for the language or the code narratives, but you probably want to pick another product to really produce that leads billing like Clio has it and LeanLaw has it. Flow before you upload, it is important to double-check the numbers, double-check the people who are doing the work, double-check that there's no blank spaces. You want to make sure that the dates are correct, the hours are correct, the rates are correct. You may have a budget where you're only allowed so many hours. You want to run a narrative pass through a fixed prompt. So no inventing facts. You want to make sure that everything in there is exactly what's allowed in the system. And then flag any UTBMS risks. So high-confidence mismatches go to billing staff first. So you really want them to look at it to make sure that it's correct. Check the rates, check timekeepers against your tables, which ones were supposed to be working on this project? What was the billing rate for this project? What was the approved billing rate? Sometimes that has happened already in the system when your client starts working with their client. And then have a human review. It's sensitive matters, privileged wording, write-downs, block billing splits. This is all going to need you to resend this back to the attorney to make sure that it's correct, especially with the block billing, because you can't have that big block billing where you did the meeting, you did the remediation, you can't have all that in one big So if they booked in 4 hours for that, you're going to have to have it go back to the attorney and have them split that out. Really important. And then upload it and then log whatever kicks back. Track that, keep track of that, and then feed those reasons into your next month's prompt library so that you're starting to build context. Maybe you're using an AI bot for this and it starts to get smarter and better as we go through this. Super important. And I highly recommend getting an AI bot to help you with this billing because this billing is so intricate. So what the humans are still going to own in this whole process, AI doesn't know your client's politics. It doesn't know that partner Smith has pet peeves and types a certain way and you have to correct this. It doesn't have a hold on any of the ethical duty on fees. You own the accuracy. You have the guidelines. You're in charge of the compliance. And in any trust retainer, Edges that sit next to the invoice. You want to make sure that if you start to use a bot or you start to use AI as an assistant to help you with this, that they're your junior assistant that's going to help you draft really quickly, but you're still the reviewer. Remember that you're still going to be the ultimate reviewer in this world. This is a great conversation that we should have in the Law Lab. If anybody does do leads billing, we'd love to have you come become a member and maybe help continue this conversation. If you're watching this video, please. Send us a message if you're having a problem with kickbacks. Let's see if we can help you with that. And that semicolon one was something that kind of caught us for a little bit. And I'm so proud of our account manager that she figured it out, but it's important. So it's almost time. You have that deadline hanging over your head. And I think that's what makes this a lot harder. So if you haven't tried it, use AI to help you do the rewrite and maybe help you with understanding and decoding what the kickback means. So I hope this has got your mind spinning on the use cases for AI or AI bots. And if you want, please join the Accountants Law Lab. We'd love to have you. And if you can't join the Law Lab, and there's a free Law Lab that you can join in Facebook, and then we have the paid mastermind group. But if you can't join, please write a note here and just tell us what was your kickback reason. I'd love to hear some of the kickback reasons that have us pulling our hair out every month when we do our billing. And on that note, we hope you all have a wonderful day. Bye now. Bye.