As anyone who grew up with The Sound of Music learned, the beginning is often a very good place to start. Unfortunately, this isn't true when it comes to legal process improvement.
I commonly hear something like this when working with attorneys:
"I don't have time to do a bunch of process improvement work, but every so often something about my workflow drives me crazy and I resolve to fix the whole darn thing. So I sketch out the different parts of my process and get to work making them better. Starting at the beginning, I take a hard look at my client intake system and make a few changes to improve it. But it is always harder than I think it's going to be, and by then the client work is usually piling up. So I go back to being a lawyer and never really get around to improving the other stages. Until a sometime down the road when it starts to drive me crazy again..."
There are several problems with this approach.
One is that it is near-insanity to try to improve an entire process all at once. Even the most experienced process improvement consultants will falter if they try to eat the whole elephant in one sitting.
Another is that if you start at the beginning every time, you'll never progress to the part of your workflow that is actually causing you trouble.
Besides, even if you wind up with a whiz-bang client intake system, you'll probably find that your new matters keep getting bogged down in the other parts of your workflow.
In fact, I'd bet this is exactly what is happening with your practice.
I'll let you in on a little secret I've learned as I work with my consulting clients: Most lawyers shouldn't try to improve their client intake process at all.
Why not? How can this be true when there are so many software tools and blog posts and marketing gurus trying to sell me things to fix my intake system?
Because client intake is almost never the stage of a law practice workflow that is holding back the rest of the delivery system. (And, by the way, unless your practice is brand spanking new, your marketing probably isn't the problem either.)
I'm not saying that your intake system can't be improved; it most certainly can. But even if your intake system is downright terrible, I'll bet you a nickle that fixing it won't do a thing to improve your overall practice efficiency. Because I guarantee there is some other part of your workflow that is worse.
Some time back I wrote about the Theory of Constraints. You should read that article for a more complete look, but the basics go like this:
- In any workflow or process, there is typically only one bottleneck that is constraining the flow of the entire system. There may be more than one slow point, but there is only one "weakest link."
- If you can improve the flow of work at your bottleneck — if you can shore up that weak link — then you can improve the flow of the entire system.
- Any effort you make to improve the flow of work at a part of your workflow that is not the bottleneck cannot improve your overall system. Again, it is like strengthening a link in the chain that wasn't going to break anyway.
Think about it. If you fix something downstream of the bottleneck, your newly fixed stage will be starved for resources due to the bottleneck. It would be like expanding the bottom of an hourglass thinking that the sand will somehow get there faster.
Expanding the top of the hourglass is no better, and it could make things worse. If you fix something upstream of the bottleneck, not only will you have wasted your time and resources on an ineffective fix, if that increased upstream flow puts more pressure on the bottleneck then you can create turbulence that slows things down even further.
Coming up tomorrow: The four metrics you need to be measuring in order to improve your operational efficiency. After that, how to find that single bottleneck in your practice and use metrics to improve it.
The key to any process improvement effort is to find your bottleneck and fix that first. Finding your bottleneck is easier said than done, but here are a few tips:
(1) It really helps to measure things. You don't need data about all of the things (at least at first), but you need to calculate these four:
Average Lead Time. How long does it typically take you to handle a matter from start to finish; i.e. from the time you send the engagement letter to the time you close the file?
Of course some matters will take longer than others, but understanding your average lead time across all of your matters — or at least all matters of a similar type — is essential to understanding whether any of your process improvement efforts are working.Average Revenue per Matter. Again, I want averages here, although if you typically handle lots of different matter types then you might want to group them to get averages for each.
Average Monthly Matters In Progress. This is a simple count each month of all of the matters in your practice that you've committed to (typically via an engagement letter) but haven't yet closed completely out of your system.
Average Monthly Close Rate. How many matters do you typically close each month? For this metric I'm talking close as in "finish all work on," not closing new sales.
This combination of metrics is necessary to know whether any process improvement experiment you run on a part of your system actually has a net positive impact on your entire workflow.
(2) If you are using a kanban board to track the stages of your workflow, then finding your constraint is usually easy. Your bottleneck is where you see work stacking up in a single column of your board. (If you're not using a kanban board, we should talk).
(3) Even if you aren't using a kanban board, you probably still have a good intuitive sense of where your bottleneck may be. Just look for the place in your workflow where work consistently gets stuck (often where it consistently needs your personal attention).
- How long does it take you to handle a matter from start to finish? This is your average cycle time. Yes, some matters will take longer than others—such is the nature of averages. At some point I may care about your outliers and your deviations from the mean. But at first I only care about your average.
- How much money do your typically make from a matter? Again, I want averages here, although if you typically handle lots of different matter types then you'll want to group them to get averages for each.
- How many matters, on average, do you close each month?
- How much money, on average, do you make each month?
(2) If you are using a kanban board to track the stages of your workflow, then finding your constraint is usually easy. Your bottleneck is where you see work stacking up in a single column of your board. I haven't yet written about how to set up a kanban for a multi-stage workflow, but if you want to build one please don't hesitate to contact me (schedule a call or send an email).
(3) If you aren't using a kanban board, you probably still have a good intuitive sense of where your bottleneck may be. The trick is to validate your hunch. Here's how: Run an experiment where you make a small improvement to the part of your workflow that you think is the bottleneck. Then measure to see whether your improvement actually makes a difference. So, back to #1, that means you also need to measure performance of that particular workflow stage. But there's a catch: you can't use improved performance at the bottleneck stage to validate whether your improvement worked. Why? Because improvements made to parts of the system that aren't the bottleneck do nothing to improve the entire system. So you have to check those four measurements I outlined in #1 above: If one or more of them improves after you make the change, then you know you've found the bottleneck (obviously keep measuring month to month to make sure you're right). If they don't improve (but your local measurement says that stage got better) then you now know that you haven't found the bottleneck. So pick another stage and run another experiment. And so on, and so on.
It isn't always easy, but it isn't complicated either. The trick, as with so many things, is patience and perseverance. And having some accountability doesn't hurt either, so if you work with a team make sure everyone is aware of your efforts. And if you're a true solo, then I strongly suggest finding a coach (and yes, I do that, but this isn't meant to be a shameless plug) or some other accountability partner.
Questions? Comments? I welcome them all. Please don't hesitate to get in touch.
* A caveat to this blanket statement: If you can do things at the client intake stage that will grease the skids for some other part of your workflow that is a demonstrable bottleneck, then making improvements at intake can be worthwhile.