Wim:
00:00:00
So throughput I view as the number one metric that any MSP or
Wim:
00:00:04
team or company should be looking at.
Wim:
00:00:07
And this comes out of lean theory, comes out of manufacturing actually.
Wim:
00:00:11
We can find a lot of examples in the other worlds.
Wim:
00:00:14
You look at how many, let's say, passengers per hour on a bus, or a
Wim:
00:00:18
train system, or a ski lift, or vehicles produced per day by a car manufacturer.
Wim:
00:00:24
This is the short video demo that will now introduce the costume
Wim:
00:00:34
feature under a couple of slides.
Wim:
00:00:40
And that's what they're focused on improving.
Wim:
00:00:42
There's different ways to improve that with looking at how people
Wim:
00:00:45
are getting on the lift, on the lift, that can improve that flow.
Wim:
00:00:48
You can add more seats on the lift, you can make the seats
Wim:
00:00:52
wider, you can speed up the lift, um, different, or add more lifts.
Wim:
00:00:56
And, but ultimately, throughput, the number one metric.
Wim:
00:00:58
So if you look at an MSP, how many projects are we doing in a month, how
Wim:
00:01:02
many firewalls are we doing, deployments, how many migrations, how many tickets,
Wim:
00:01:08
um, throughput shows up in a lot of ways.
Wim:
00:01:10
And I think we can look at a lot of other metrics, like your kill
Wim:
00:01:14
rate on tickets, marketing, so many things, but ultimately it comes down
Wim:
00:01:18
to throughput, how many tasks, and again, that can be a wide variety of tasks, from
Wim:
00:01:24
small, from five minute things to five month efforts, whatever it might be, but
Wim:
00:01:28
how many can we get through the system?
Wim:
00:01:30
And I think it requires, absolutely requires the team working
Wim:
00:01:33
together, because you can look at individual person performance.
Wim:
00:01:37
And, be fooling ourselves, people are in quotes doing their job,
Wim:
00:01:41
but how is it impacting the throughput of the entire system?
Wim:
00:01:45
Most PSAs have the ability to add some sort of widget or calculation.
Wim:
00:01:49
Um, some are primitive, but you want to look at the total number of tasks
Wim:
00:01:52
completed in that specific time frame.
Wim:
00:01:54
So, it might be how many sales opportunities have gone through
Wim:
00:01:57
the pipeline in a month.
Wim:
00:01:59
or how many tickets per week or per day are actually completed.
Wim:
00:02:02
We're not looking at how many have been started or in progress, but how
Wim:
00:02:05
many have actually gone through the entire process and are done done.
Wim:
00:02:08
So it's a simple calculation.
Steve:
00:02:11
What is the purpose of monitoring throughput.
Wim:
00:02:14
Ultimately about performance, what's the result that we're producing?
Wim:
00:02:17
If you look at a factory, how efficient are they with that number
Wim:
00:02:20
of resources or square footage?
Wim:
00:02:22
What's the output at the end of the day, literally at the end of the day?
Wim:
00:02:26
What are they producing?
Wim:
00:02:27
And once you measure it, you can start improving it, which might be reducing
Wim:
00:02:31
the steps required, there's training, there's automation, maybe robots, looking
Wim:
00:02:36
at the bottlenecks and improving that throughput, adding equipment, investing
Wim:
00:02:41
into different things so that we're, we're generating more, and ultimately
Wim:
00:02:44
that's gonna result in more money.
Wim:
00:02:46
If you look at a system and a factory or an MSP telling people just to produce
Wim:
00:02:51
more profit, that's not helpful at all.
Wim:
00:02:53
Like, you've got to connect what they're doing right now at whatever,
Wim:
00:02:57
8am or whatever time it is.
Wim:
00:02:58
What are they doing right now, this morning?
Wim:
00:03:00
How can they impact that and that will eventually cascade into profit.
Wim:
00:03:04
and more value for the company.
Wim:
00:03:06
We've got to break it down to things they can get their hands on and drive
Wim:
00:03:10
and visualize on an hour by hour basis.
Wim:
00:03:14
So I think this is fundamental, big picture thinking is what's the
Wim:
00:03:17
throughput of our team, of our company.
Wim:
00:03:20
How many clients are we supporting?
Wim:
00:03:21
How many seats are we supporting?
Wim:
00:03:23
How many clients can we onboard every month?
Wim:
00:03:26
And how would we maybe go about improving that?
Wim:
00:03:29
If we want to, if we need to.
Steve:
00:03:31
How often do you think MSPs should be checking throughput?
Wim:
00:03:34
It depends on the particular, I guess, subsystem you're looking at.
Wim:
00:03:37
If you're looking at the helpdesk tickets...
Wim:
00:03:39
You might want to look at that daily, but if you're looking at onboarding
Wim:
00:03:42
and you're doing one a month, well, maybe you're just looking at
Wim:
00:03:45
it a monthly or quarterly basis.
Wim:
00:03:46
How are we doing on that throughput?
Wim:
00:03:49
And, I guess, where is it a bottleneck?
Wim:
00:03:52
Where do you want to improve it?
Wim:
00:03:53
Um, the more you want to improve it, the more you get, the more
Wim:
00:03:55
frequently you should look at it, visualize it, be talking about it.
Wim:
00:04:00
Um, so, yeah, huge impact on profitability.
Wim:
00:04:03
Obviously, the higher throughput, we can...
Wim:
00:04:05
We serve more clients, we drive revenue, we're more profitable,
Wim:
00:04:08
especially if we can do that with the people we have, that's, that's
Wim:
00:04:12
a win for the entire company.
Wim:
00:04:14
If you can do more tickets per day, more projects in a week, that's only good.
Wim:
00:04:18
And then there's a huge impact on mental health.
Wim:
00:04:21
If you can complete more tickets or projects in less time, that's,
Wim:
00:04:24
people feel like they're achieving something, they're part of a winning
Wim:
00:04:26
team, the more productive, it's a much more positive working environment.
Steve:
00:04:30
Can you give me some examples of who should be responsible for
Steve:
00:04:34
monitoring the throughput KPI?
Wim:
00:04:36
A CEO is ultimately responsible for performance of the company and
Wim:
00:04:39
overall throughput, but he's going to be delegating that to his managers.
Wim:
00:04:42
It's going to cascade throughout the organization.
Wim:
00:04:44
So project manager will be looking at how do we get more projects through, having
Wim:
00:04:48
discussions with their team about how do we remove bottlenecks, reduce defects so
Wim:
00:04:53
that we're not reworking stuff, improving that flow, improving the handoffs,
Wim:
00:04:57
making sure stuff is not waiting.
Wim:
00:04:58
It's a team effort, but yeah, the managers.
Wim:
00:05:01
would be the main people thinking about it and owning those metrics.
Wim:
00:05:05
I have an example.
Wim:
00:05:06
So in our Crafty Penguins MSP, it's all Linux, the Linux MSP,
Wim:
00:05:11
used to have only one person doing discovery projects for new clients.
Wim:
00:05:15
And that's a bottleneck.
Wim:
00:05:17
We could only do so many.
Wim:
00:05:20
And so now we've scaled it.
Wim:
00:05:21
Any engineer in the team can do a discovery project at the
Wim:
00:05:25
required quality, required output.
Wim:
00:05:27
And we get a great close rate on that, so it's removed the overwhelm, the
Wim:
00:05:32
stress, and it's, um, so it improves the throughput on that discovery
Wim:
00:05:36
process, but it's really helped grow the company as well, because we can
Wim:
00:05:39
onboard clients so much quicker that way.
Steve:
00:05:41
How do you know what actions to take based on throughput and
Steve:
00:05:44
what actions should there be?
Wim:
00:05:46
Should we start with understanding that system?
Wim:
00:05:48
Like, what are the steps in it?
Wim:
00:05:50
What are the stages?
Wim:
00:05:52
And then only then can we look at, okay, what, what are the bottlenecks?
Wim:
00:05:55
Take a data driven approach.
Wim:
00:05:56
If you take like a factory example, you're manufacturing cars.
Wim:
00:05:59
Uh, always a simple example that everyone can relate to is you got assembly, you
Wim:
00:06:03
got the various points of that, and you get to maybe painting and final QA check.
Wim:
00:06:08
Is we look at where is work waiting?
Wim:
00:06:11
If it's waiting to get into the paint booth, well, why is that?
Wim:
00:06:14
Is an equipment issue, training issue, not enough people in the, in the paint booth.
Wim:
00:06:17
So looking at breaking down the process or process map, visualizing that,
Wim:
00:06:22
finally finding those bottlenecks and then obliterating those through
Wim:
00:06:27
whatever it might be, training, better equipment, automation, SOPs,
Wim:
00:06:31
better meetings, whatever it might be.
Wim:
00:06:33
So it's a great thing to do in the.
Wim:
00:06:36
In our world, it's getting daily huddles where we're looking at what
Wim:
00:06:40
tickets are hitting a bottleneck, what's neglected, what's waiting on somebody
Wim:
00:06:44
who's not there, what's waiting on maybe a technician who's been there
Wim:
00:06:47
for 10 years and knows everything, and kind of everyone's going looking up
Wim:
00:06:51
to him, um, we call that a Brent that comes from the Phoenix project book,
Wim:
00:06:55
which I've, um, Oh, we're on audio, so I would show you the book cover.
Wim:
00:06:59
I've got it on my back wall here, actually.
Wim:
00:07:01
I'm just a big fan of the Phoenix Project book, which gets into throughput
Wim:
00:07:04
and how do you remove bottlenecks.
Wim:
00:07:06
But Brent was a big one, because everyone kept going to Brent in this
Wim:
00:07:09
book, because he knew everything.
Wim:
00:07:10
He knew about the sand and the firewalls and weird stuff that
Wim:
00:07:13
was done seven years ago that all the new people didn't know about.
Wim:
00:07:16
Um, so that's, yeah, break down the steps, find the bottlenecks, see what's waiting.
Wim:
00:07:21
Keep it moving.
Wim:
00:07:22
We want continual flow.
Wim:
00:07:23
So when that car is going through that manufacturer floor,
Wim:
00:07:26
it should never sit still.
Wim:
00:07:27
It's always being worked on.
Wim:
00:07:29
You don't want it sitting there for, for four hours waiting.
Wim:
00:07:32
It should just keep continuous flow.
Wim:
00:07:34
And so it should be on our tickets as well, as they keep going
Wim:
00:07:37
through our system being worked.
Wim:
00:07:38
They're not waiting for a day or a week or three weeks.
Wim:
00:07:41
So focusing on getting it through the system, out the other end.
Wim:
00:07:46
The other analogy I have is, um, Having a pipeline, oil pipeline, is you don't
Wim:
00:07:50
want that oil sitting still, you don't want gunk building up on the inside,
Wim:
00:07:54
you don't want blockages, and keep it going high velocity, high pressure.
Wim:
00:07:58
So yeah, these daily meetings can really help, like a 10 minute meeting, be looking
Wim:
00:08:01
at, looking at your tickets, what's been waiting, waiting on a client, what's been
Wim:
00:08:06
started and not finished, and we, people get enamored with starting new tickets.
Wim:
00:08:11
Procrastinate on finishing the stuff they're ready to start.
Wim:
00:08:13
Is it creating that culture of completing overstarting?
Wim:
00:08:16
There's a lot of ways to hit it, and it depends on exactly why that
Wim:
00:08:19
work is getting delayed or hung up.
Wim:
00:08:21
But it requires a team effort, like the manager alone can't get it unstuck.
Wim:
00:08:25
It's like working together with the people, bringing in other people,
Wim:
00:08:28
more eyes on it always helps.
Wim:
00:08:30
I think a lot of people are trying hard to improve their performance and throughput
Wim:
00:08:34
and such, but they're optimizing the wrong areas, because they don't understand the
Wim:
00:08:39
real bottleneck, which could be, could be technology, could be a process step,
Wim:
00:08:43
it could be people, could be culture.
Wim:
00:08:45
And as long as you're optimizing something that's not the bottleneck, you're
Wim:
00:08:48
actually not improving the throughput.
Wim:
00:08:50
That's, um, you've got to be scientific about it, is what
Wim:
00:08:54
actually is the issue doing the five whys of why, why did this happen?
Wim:
00:08:59
Okay, and why that and just digging down that way?
Wim:
00:09:02
Because if you're, as long as you're proving them something, not the
Wim:
00:09:05
bottleneck, you're wasting time.
Wim:
00:09:06
You can let the secondary things be inefficient.
Wim:
00:09:09
That's okay.
Wim:
00:09:09
You got to focus on your primary bottleneck.
Wim:
00:09:12
So it's, you can be, you can get carried away with automation,
Wim:
00:09:15
for example, in your RMM.
Wim:
00:09:17
But if it's not your bottleneck, then it's a waste, like maybe it should
Wim:
00:09:20
have been training your technicians, or SOPs, or just the way we onboard clients
Wim:
00:09:25
might be a bigger, a bigger bottleneck.
Wim:
00:09:28
So being a big picture aware of what's going on is key.
Steve:
00:09:31
let's keep talking about that bigger picture.
Steve:
00:09:34
when looking at throughput, there are some other metrics that we
Steve:
00:09:38
should be looking at alongside it.
Steve:
00:09:40
So, for tickets, if we're looking at throughput, we may also want to
Steve:
00:09:44
look at, , overall the number of tickets opened, the number of tickets
Steve:
00:09:48
touched, number of tickets closed.
Steve:
00:09:51
So that way we can get a bigger picture of our service desk.
Steve:
00:09:55
Are there any other metrics that you can think of that go
Steve:
00:09:59
really nicely with throughput?
Wim:
00:10:01
Absolutely.
Wim:
00:10:02
Um, that's the work in progress would be...
Wim:
00:10:06
A very close one to this.
Wim:
00:10:08
Work in progress and lead time and the flow efficiency.
Wim:
00:10:12
So we'll get into those, those next.
Wim:
00:10:14
But work in progress is the number of tasks currently
Wim:
00:10:16
being worked on in the system.
Wim:
00:10:18
So it could be a ticket that's started, and not completed.
Wim:
00:10:21
If you look at manufacturing a car that's been started, it's got one part.
Wim:
00:10:25
Somebody's picked one part, um, from inventory and is starting
Wim:
00:10:28
to do something, but not done.
Wim:
00:10:30
And we want to reduce that because the more work in progress you have,
Wim:
00:10:34
It's more to touch, more to visualize, more to report on, more to talk about.
Wim:
00:10:38
It really bogs down the performance.
Wim:
00:10:41
And I'm really trying to drill that home with MSPs, is reduce the work in progress,
Wim:
00:10:46
visualize it, talk about it, put it on the dashboards, always focus on getting,
Wim:
00:10:50
getting that down, first in, first out, focus on completing overstarting.
Wim:
00:10:55
Because the less you got going on, it's just, there's less distraction,
Wim:
00:10:58
there's less decision fatigue.
Wim:
00:11:00
There's less context switching, less multi tasking, so that's a key one
Wim:
00:11:04
to improve throughput is reduce the work in progress in that system.
Wim:
00:11:09
Whether it's a helpdesk, a project team, a sales team, is get it through, get it out.
Wim:
00:11:14
That's a key one, work in progress.
Wim:
00:11:15
The lead time refers to how long it takes from the time a request
Wim:
00:11:20
is made to the time it's complete.
Wim:
00:11:22
So if you look at a car, how often from the time, let's say if you're doing
Wim:
00:11:25
a custom order on a car, you've On a car, you've gone to the Tesla website,
Wim:
00:11:29
submitted your order, how long until you've got that car on your doorstep?
Wim:
00:11:33
So if you look at where the customer submitted, emailed, support at your MSP.
Wim:
00:11:37
com, how long until it's done done?
Wim:
00:11:39
Or from the time they signed a quote on a 365 migration,
Wim:
00:11:43
how long until it's done done?
Wim:
00:11:45
Like QA's done, no bugs, documented, They're happy.
Wim:
00:11:49
How long is that?
Wim:
00:11:50
So if you can tighten that up, obviously that's going to improve your throughput.
Wim:
00:11:53
And then the third metric supporting throughput is flow efficiency.
Wim:
00:11:58
That's the ratio of time a task is taking between or is actively being worked on to
Wim:
00:12:03
the total time it spends in the system.
Wim:
00:12:05
This is a fascinating one.
Wim:
00:12:06
When you visualize it the right way or see an example, it kind of blows your mind.
Wim:
00:12:09
It's like...
Wim:
00:12:10
Okay, so I got these tickets that are taking a week to get done, but
Wim:
00:12:13
actually the technicians are only putting a half an hour on it, so why
Wim:
00:12:16
is it waiting a week or two weeks?
Wim:
00:12:18
A lot of tickets spent forever just waiting.
Wim:
00:12:20
A lot of projects waiting for people to have time, or some approval from the
Wim:
00:12:24
client, or some little bit of information that most likely we already have.
Wim:
00:12:29
So I mentioned that with the car in the factory is the Teslas
Wim:
00:12:32
is pumping out these cars.
Wim:
00:12:34
They don't want that sitting there at all.
Wim:
00:12:36
It's moving, it's continually moving down the line , and
Wim:
00:12:38
people are moving around it.
Wim:
00:12:40
As these cars come by, the people don't move the work moves and on.
Wim:
00:12:45
Yeah.
Wim:
00:12:45
More, more and more are being done on a multiple people working
Wim:
00:12:48
on it at the same time, the car is never waiting for somebody.
Steve:
00:12:51
Awesome.
Steve:
00:12:52
I really like that.
Steve:
00:12:53
I can't wait to dig into these other metrics with you in future episodes.
Wim:
00:12:59
Likewise.
Steve:
00:13:00
Thank you, Wim.
Wim:
00:13:01
My pleasure.
00:00:00
So throughput I view as the number one metric that any MSP or
Wim:
00:00:04
team or company should be looking at.
Wim:
00:00:07
And this comes out of lean theory, comes out of manufacturing actually.
Wim:
00:00:11
We can find a lot of examples in the other worlds.
Wim:
00:00:14
You look at how many, let's say, passengers per hour on a bus, or a
Wim:
00:00:18
train system, or a ski lift, or vehicles produced per day by a car manufacturer.
Wim:
00:00:24
This is the short video demo that will now introduce the costume
Wim:
00:00:34
feature under a couple of slides.
Wim:
00:00:40
And that's what they're focused on improving.
Wim:
00:00:42
There's different ways to improve that with looking at how people
Wim:
00:00:45
are getting on the lift, on the lift, that can improve that flow.
Wim:
00:00:48
You can add more seats on the lift, you can make the seats
Wim:
00:00:52
wider, you can speed up the lift, um, different, or add more lifts.
Wim:
00:00:56
And, but ultimately, throughput, the number one metric.
Wim:
00:00:58
So if you look at an MSP, how many projects are we doing in a month, how
Wim:
00:01:02
many firewalls are we doing, deployments, how many migrations, how many tickets,
Wim:
00:01:08
um, throughput shows up in a lot of ways.
Wim:
00:01:10
And I think we can look at a lot of other metrics, like your kill
Wim:
00:01:14
rate on tickets, marketing, so many things, but ultimately it comes down
Wim:
00:01:18
to throughput, how many tasks, and again, that can be a wide variety of tasks, from
Wim:
00:01:24
small, from five minute things to five month efforts, whatever it might be, but
Wim:
00:01:28
how many can we get through the system?
Wim:
00:01:30
And I think it requires, absolutely requires the team working
Wim:
00:01:33
together, because you can look at individual person performance.
Wim:
00:01:37
And, be fooling ourselves, people are in quotes doing their job,
Wim:
00:01:41
but how is it impacting the throughput of the entire system?
Wim:
00:01:45
Most PSAs have the ability to add some sort of widget or calculation.
Wim:
00:01:49
Um, some are primitive, but you want to look at the total number of tasks
Wim:
00:01:52
completed in that specific time frame.
Wim:
00:01:54
So, it might be how many sales opportunities have gone through
Wim:
00:01:57
the pipeline in a month.
Wim:
00:01:59
or how many tickets per week or per day are actually completed.
Wim:
00:02:02
We're not looking at how many have been started or in progress, but how
Wim:
00:02:05
many have actually gone through the entire process and are done done.
Wim:
00:02:08
So it's a simple calculation.
Steve:
00:02:11
What is the purpose of monitoring throughput.
Wim:
00:02:14
Ultimately about performance, what's the result that we're producing?
Wim:
00:02:17
If you look at a factory, how efficient are they with that number
Wim:
00:02:20
of resources or square footage?
Wim:
00:02:22
What's the output at the end of the day, literally at the end of the day?
Wim:
00:02:26
What are they producing?
Wim:
00:02:27
And once you measure it, you can start improving it, which might be reducing
Wim:
00:02:31
the steps required, there's training, there's automation, maybe robots, looking
Wim:
00:02:36
at the bottlenecks and improving that throughput, adding equipment, investing
Wim:
00:02:41
into different things so that we're, we're generating more, and ultimately
Wim:
00:02:44
that's gonna result in more money.
Wim:
00:02:46
If you look at a system and a factory or an MSP telling people just to produce
Wim:
00:02:51
more profit, that's not helpful at all.
Wim:
00:02:53
Like, you've got to connect what they're doing right now at whatever,
Wim:
00:02:57
8am or whatever time it is.
Wim:
00:02:58
What are they doing right now, this morning?
Wim:
00:03:00
How can they impact that and that will eventually cascade into profit.
Wim:
00:03:04
and more value for the company.
Wim:
00:03:06
We've got to break it down to things they can get their hands on and drive
Wim:
00:03:10
and visualize on an hour by hour basis.
Wim:
00:03:14
So I think this is fundamental, big picture thinking is what's the
Wim:
00:03:17
throughput of our team, of our company.
Wim:
00:03:20
How many clients are we supporting?
Wim:
00:03:21
How many seats are we supporting?
Wim:
00:03:23
How many clients can we onboard every month?
Wim:
00:03:26
And how would we maybe go about improving that?
Wim:
00:03:29
If we want to, if we need to.
Steve:
00:03:31
How often do you think MSPs should be checking throughput?
Wim:
00:03:34
It depends on the particular, I guess, subsystem you're looking at.
Wim:
00:03:37
If you're looking at the helpdesk tickets...
Wim:
00:03:39
You might want to look at that daily, but if you're looking at onboarding
Wim:
00:03:42
and you're doing one a month, well, maybe you're just looking at
Wim:
00:03:45
it a monthly or quarterly basis.
Wim:
00:03:46
How are we doing on that throughput?
Wim:
00:03:49
And, I guess, where is it a bottleneck?
Wim:
00:03:52
Where do you want to improve it?
Wim:
00:03:53
Um, the more you want to improve it, the more you get, the more
Wim:
00:03:55
frequently you should look at it, visualize it, be talking about it.
Wim:
00:04:00
Um, so, yeah, huge impact on profitability.
Wim:
00:04:03
Obviously, the higher throughput, we can...
Wim:
00:04:05
We serve more clients, we drive revenue, we're more profitable,
Wim:
00:04:08
especially if we can do that with the people we have, that's, that's
Wim:
00:04:12
a win for the entire company.
Wim:
00:04:14
If you can do more tickets per day, more projects in a week, that's only good.
Wim:
00:04:18
And then there's a huge impact on mental health.
Wim:
00:04:21
If you can complete more tickets or projects in less time, that's,
Wim:
00:04:24
people feel like they're achieving something, they're part of a winning
Wim:
00:04:26
team, the more productive, it's a much more positive working environment.
Steve:
00:04:30
Can you give me some examples of who should be responsible for
Steve:
00:04:34
monitoring the throughput KPI?
Wim:
00:04:36
A CEO is ultimately responsible for performance of the company and
Wim:
00:04:39
overall throughput, but he's going to be delegating that to his managers.
Wim:
00:04:42
It's going to cascade throughout the organization.
Wim:
00:04:44
So project manager will be looking at how do we get more projects through, having
Wim:
00:04:48
discussions with their team about how do we remove bottlenecks, reduce defects so
Wim:
00:04:53
that we're not reworking stuff, improving that flow, improving the handoffs,
Wim:
00:04:57
making sure stuff is not waiting.
Wim:
00:04:58
It's a team effort, but yeah, the managers.
Wim:
00:05:01
would be the main people thinking about it and owning those metrics.
Wim:
00:05:05
I have an example.
Wim:
00:05:06
So in our Crafty Penguins MSP, it's all Linux, the Linux MSP,
Wim:
00:05:11
used to have only one person doing discovery projects for new clients.
Wim:
00:05:15
And that's a bottleneck.
Wim:
00:05:17
We could only do so many.
Wim:
00:05:20
And so now we've scaled it.
Wim:
00:05:21
Any engineer in the team can do a discovery project at the
Wim:
00:05:25
required quality, required output.
Wim:
00:05:27
And we get a great close rate on that, so it's removed the overwhelm, the
Wim:
00:05:32
stress, and it's, um, so it improves the throughput on that discovery
Wim:
00:05:36
process, but it's really helped grow the company as well, because we can
Wim:
00:05:39
onboard clients so much quicker that way.
Steve:
00:05:41
How do you know what actions to take based on throughput and
Steve:
00:05:44
what actions should there be?
Wim:
00:05:46
Should we start with understanding that system?
Wim:
00:05:48
Like, what are the steps in it?
Wim:
00:05:50
What are the stages?
Wim:
00:05:52
And then only then can we look at, okay, what, what are the bottlenecks?
Wim:
00:05:55
Take a data driven approach.
Wim:
00:05:56
If you take like a factory example, you're manufacturing cars.
Wim:
00:05:59
Uh, always a simple example that everyone can relate to is you got assembly, you
Wim:
00:06:03
got the various points of that, and you get to maybe painting and final QA check.
Wim:
00:06:08
Is we look at where is work waiting?
Wim:
00:06:11
If it's waiting to get into the paint booth, well, why is that?
Wim:
00:06:14
Is an equipment issue, training issue, not enough people in the, in the paint booth.
Wim:
00:06:17
So looking at breaking down the process or process map, visualizing that,
Wim:
00:06:22
finally finding those bottlenecks and then obliterating those through
Wim:
00:06:27
whatever it might be, training, better equipment, automation, SOPs,
Wim:
00:06:31
better meetings, whatever it might be.
Wim:
00:06:33
So it's a great thing to do in the.
Wim:
00:06:36
In our world, it's getting daily huddles where we're looking at what
Wim:
00:06:40
tickets are hitting a bottleneck, what's neglected, what's waiting on somebody
Wim:
00:06:44
who's not there, what's waiting on maybe a technician who's been there
Wim:
00:06:47
for 10 years and knows everything, and kind of everyone's going looking up
Wim:
00:06:51
to him, um, we call that a Brent that comes from the Phoenix project book,
Wim:
00:06:55
which I've, um, Oh, we're on audio, so I would show you the book cover.
Wim:
00:06:59
I've got it on my back wall here, actually.
Wim:
00:07:01
I'm just a big fan of the Phoenix Project book, which gets into throughput
Wim:
00:07:04
and how do you remove bottlenecks.
Wim:
00:07:06
But Brent was a big one, because everyone kept going to Brent in this
Wim:
00:07:09
book, because he knew everything.
Wim:
00:07:10
He knew about the sand and the firewalls and weird stuff that
Wim:
00:07:13
was done seven years ago that all the new people didn't know about.
Wim:
00:07:16
Um, so that's, yeah, break down the steps, find the bottlenecks, see what's waiting.
Wim:
00:07:21
Keep it moving.
Wim:
00:07:22
We want continual flow.
Wim:
00:07:23
So when that car is going through that manufacturer floor,
Wim:
00:07:26
it should never sit still.
Wim:
00:07:27
It's always being worked on.
Wim:
00:07:29
You don't want it sitting there for, for four hours waiting.
Wim:
00:07:32
It should just keep continuous flow.
Wim:
00:07:34
And so it should be on our tickets as well, as they keep going
Wim:
00:07:37
through our system being worked.
Wim:
00:07:38
They're not waiting for a day or a week or three weeks.
Wim:
00:07:41
So focusing on getting it through the system, out the other end.
Wim:
00:07:46
The other analogy I have is, um, Having a pipeline, oil pipeline, is you don't
Wim:
00:07:50
want that oil sitting still, you don't want gunk building up on the inside,
Wim:
00:07:54
you don't want blockages, and keep it going high velocity, high pressure.
Wim:
00:07:58
So yeah, these daily meetings can really help, like a 10 minute meeting, be looking
Wim:
00:08:01
at, looking at your tickets, what's been waiting, waiting on a client, what's been
Wim:
00:08:06
started and not finished, and we, people get enamored with starting new tickets.
Wim:
00:08:11
Procrastinate on finishing the stuff they're ready to start.
Wim:
00:08:13
Is it creating that culture of completing overstarting?
Wim:
00:08:16
There's a lot of ways to hit it, and it depends on exactly why that
Wim:
00:08:19
work is getting delayed or hung up.
Wim:
00:08:21
But it requires a team effort, like the manager alone can't get it unstuck.
Wim:
00:08:25
It's like working together with the people, bringing in other people,
Wim:
00:08:28
more eyes on it always helps.
Wim:
00:08:30
I think a lot of people are trying hard to improve their performance and throughput
Wim:
00:08:34
and such, but they're optimizing the wrong areas, because they don't understand the
Wim:
00:08:39
real bottleneck, which could be, could be technology, could be a process step,
Wim:
00:08:43
it could be people, could be culture.
Wim:
00:08:45
And as long as you're optimizing something that's not the bottleneck, you're
Wim:
00:08:48
actually not improving the throughput.
Wim:
00:08:50
That's, um, you've got to be scientific about it, is what
Wim:
00:08:54
actually is the issue doing the five whys of why, why did this happen?
Wim:
00:08:59
Okay, and why that and just digging down that way?
Wim:
00:09:02
Because if you're, as long as you're proving them something, not the
Wim:
00:09:05
bottleneck, you're wasting time.
Wim:
00:09:06
You can let the secondary things be inefficient.
Wim:
00:09:09
That's okay.
Wim:
00:09:09
You got to focus on your primary bottleneck.
Wim:
00:09:12
So it's, you can be, you can get carried away with automation,
Wim:
00:09:15
for example, in your RMM.
Wim:
00:09:17
But if it's not your bottleneck, then it's a waste, like maybe it should
Wim:
00:09:20
have been training your technicians, or SOPs, or just the way we onboard clients
Wim:
00:09:25
might be a bigger, a bigger bottleneck.
Wim:
00:09:28
So being a big picture aware of what's going on is key.
Steve:
00:09:31
let's keep talking about that bigger picture.
Steve:
00:09:34
when looking at throughput, there are some other metrics that we
Steve:
00:09:38
should be looking at alongside it.
Steve:
00:09:40
So, for tickets, if we're looking at throughput, we may also want to
Steve:
00:09:44
look at, , overall the number of tickets opened, the number of tickets
Steve:
00:09:48
touched, number of tickets closed.
Steve:
00:09:51
So that way we can get a bigger picture of our service desk.
Steve:
00:09:55
Are there any other metrics that you can think of that go
Steve:
00:09:59
really nicely with throughput?
Wim:
00:10:01
Absolutely.
Wim:
00:10:02
Um, that's the work in progress would be...
Wim:
00:10:06
A very close one to this.
Wim:
00:10:08
Work in progress and lead time and the flow efficiency.
Wim:
00:10:12
So we'll get into those, those next.
Wim:
00:10:14
But work in progress is the number of tasks currently
Wim:
00:10:16
being worked on in the system.
Wim:
00:10:18
So it could be a ticket that's started, and not completed.
Wim:
00:10:21
If you look at manufacturing a car that's been started, it's got one part.
Wim:
00:10:25
Somebody's picked one part, um, from inventory and is starting
Wim:
00:10:28
to do something, but not done.
Wim:
00:10:30
And we want to reduce that because the more work in progress you have,
Wim:
00:10:34
It's more to touch, more to visualize, more to report on, more to talk about.
Wim:
00:10:38
It really bogs down the performance.
Wim:
00:10:41
And I'm really trying to drill that home with MSPs, is reduce the work in progress,
Wim:
00:10:46
visualize it, talk about it, put it on the dashboards, always focus on getting,
Wim:
00:10:50
getting that down, first in, first out, focus on completing overstarting.
Wim:
00:10:55
Because the less you got going on, it's just, there's less distraction,
Wim:
00:10:58
there's less decision fatigue.
Wim:
00:11:00
There's less context switching, less multi tasking, so that's a key one
Wim:
00:11:04
to improve throughput is reduce the work in progress in that system.
Wim:
00:11:09
Whether it's a helpdesk, a project team, a sales team, is get it through, get it out.
Wim:
00:11:14
That's a key one, work in progress.
Wim:
00:11:15
The lead time refers to how long it takes from the time a request
Wim:
00:11:20
is made to the time it's complete.
Wim:
00:11:22
So if you look at a car, how often from the time, let's say if you're doing
Wim:
00:11:25
a custom order on a car, you've On a car, you've gone to the Tesla website,
Wim:
00:11:29
submitted your order, how long until you've got that car on your doorstep?
Wim:
00:11:33
So if you look at where the customer submitted, emailed, support at your MSP.
Wim:
00:11:37
com, how long until it's done done?
Wim:
00:11:39
Or from the time they signed a quote on a 365 migration,
Wim:
00:11:43
how long until it's done done?
Wim:
00:11:45
Like QA's done, no bugs, documented, They're happy.
Wim:
00:11:49
How long is that?
Wim:
00:11:50
So if you can tighten that up, obviously that's going to improve your throughput.
Wim:
00:11:53
And then the third metric supporting throughput is flow efficiency.
Wim:
00:11:58
That's the ratio of time a task is taking between or is actively being worked on to
Wim:
00:12:03
the total time it spends in the system.
Wim:
00:12:05
This is a fascinating one.
Wim:
00:12:06
When you visualize it the right way or see an example, it kind of blows your mind.
Wim:
00:12:09
It's like...
Wim:
00:12:10
Okay, so I got these tickets that are taking a week to get done, but
Wim:
00:12:13
actually the technicians are only putting a half an hour on it, so why
Wim:
00:12:16
is it waiting a week or two weeks?
Wim:
00:12:18
A lot of tickets spent forever just waiting.
Wim:
00:12:20
A lot of projects waiting for people to have time, or some approval from the
Wim:
00:12:24
client, or some little bit of information that most likely we already have.
Wim:
00:12:29
So I mentioned that with the car in the factory is the Teslas
Wim:
00:12:32
is pumping out these cars.
Wim:
00:12:34
They don't want that sitting there at all.
Wim:
00:12:36
It's moving, it's continually moving down the line , and
Wim:
00:12:38
people are moving around it.
Wim:
00:12:40
As these cars come by, the people don't move the work moves and on.
Wim:
00:12:45
Yeah.
Wim:
00:12:45
More, more and more are being done on a multiple people working
Wim:
00:12:48
on it at the same time, the car is never waiting for somebody.
Steve:
00:12:51
Awesome.
Steve:
00:12:52
I really like that.
Steve:
00:12:53
I can't wait to dig into these other metrics with you in future episodes.
Wim:
00:12:59
Likewise.
Steve:
00:13:00
Thank you, Wim.
Wim:
00:13:01
My pleasure.
