0:05

Welcome to the Productivity Podcast . I'm

0:07

back with Sue and we're going to do some Productivity

0:10

Insights and on this episode we're going to

0:12

focus on frequently

0:14

asked questions . So lots of

0:16

questions we get asked by

0:18

new clients , common clients , so

0:21

we'll work through a few of them and explore

0:23

where we get through . Do you want to kick off , susan

0:25

?

0:29

Yes , so the first question that we get asked

0:31

a lot is do people change their behaviours

0:33

when they're being studied ?

0:36

I think the answer is potentially

0:39

for the first hour or so , but

0:42

I also think that by

0:44

using pace rating , that's normalised

0:47

out of the data . So I think we've talked

0:49

about pace rating before on a different episode , but do you want

0:51

to give people a reminder of what pace

0:53

rating is ?

0:55

Pace rating is where analysts

0:58

are trained to be able to

1:00

assess the effectiveness

1:02

that somebody is working at versus a British standard

1:05

where the British standard is 100 . And

1:07

kind of we call it pace rating . But it's more

1:09

than that . It's also about how

1:11

effective they're being . So , for

1:14

example , if I'm walking at a good pace but

1:16

I'm carrying something and spilling a load of

1:18

it , then I'd be downrated because from

1:20

an effective point of view I might be going quickly

1:22

but I'm not doing a good job of it . So

1:25

pace is then used to normalise

1:28

things . So I think sometimes people

1:30

have concerns that . Do people perhaps slow down

1:33

when they're being timed to make

1:35

things look like they take longer ? Well

1:37

, again , the pace rating would pick that up

1:39

, because if somebody's working at 80

1:41

, then actually when

1:44

we then do the analysis , it's then normalised

1:46

back as if it was 100 . The

1:49

only people that you perhaps wouldn't want to study

1:51

in that way would be if they're going slower because

1:53

they're new in the training . So really

1:55

we should only be studying qualified

1:58

operatives , so that people that are

2:00

competent at the role . So if you've got somebody

2:02

that's brand new and in training and they're slow because

2:04

of that , then they aren't a good subject

2:06

to be studying .

2:08

So that answers the question kind of what

2:11

if you go slow , speed up , which is

2:13

good One's

2:15

, I kind of get . So how

2:17

many stores , locations should

2:19

I study ? What's the sample size ?

2:22

It's always a tricky one , isn't it ? Because it depends

2:24

. So the

2:28

more variability there

2:30

is in whatever you're measuring , then

2:32

the bigger sample size you need . So

2:35

if I've got 10

2:37

restaurants but each of them is a different

2:39

format , I'll need to spend longer

2:41

in each one than if they

2:43

were all the same . So it can be

2:45

variability in terms of kind

2:48

of the outlet type that you're measuring

2:50

. It can also

2:52

be variability in the processes . So

2:55

if a process is always the same , so

2:58

a production line might be kind

3:00

of an obvious example where it's a standard

3:02

one , then you don't need to measure

3:04

it very often because it's always the same . But

3:07

things that are more variable so it might

3:09

be different menu items in a

3:11

restaurant , anywhere where people

3:13

and conversations are involved , always

3:15

has a lot of variability in it . So , for

3:18

example , customers in a shop they

3:20

might be chatty , they might be , not all

3:22

that sort of thing . So it depends on the variability

3:25

. I think often we say it

3:27

kind of in a smaller

3:29

number of outlets you might

3:31

be looking to do 10% of the estate

3:33

, but obviously if you've got 2,000

3:36

shops , then you wouldn't be looking to do 10%

3:38

. So it varies . It's

3:40

not just the number of stores , it's also

3:42

how many days you spend .

3:45

There's no magic formula . I don't think it's on

3:47

a case by case basis , along

3:49

with , as you say , that bit around

3:52

the variability

3:54

is a key bit , and

3:56

I think that that's probably

3:58

a casing point around the

4:01

number of days that you say . So

4:03

there's some organisations out there that are

4:05

kind of hooked on there , measuring all the

4:07

time and measuring everything . We're not massive

4:09

advocates of that . I think every

4:11

time you measure it changes the number . So you've

4:14

then got a bigger data set , clearly more

4:16

robust data , but you've got to explain the variance

4:18

. So kind of leads me

4:20

on to how often should you re-measure

4:22

?

4:24

Again , it's that it depends question . So

4:27

if you're in a business

4:29

that doesn't change at all , why

4:31

would you bother re-measuring ? But the

4:33

reality is that businesses do

4:35

change all the time . So if

4:38

you assume that somebody's done a sort

4:40

of a big , wholesale measure of

4:42

most of the processes , you

4:44

can then just follow it up . So if you change

4:46

one part of your operation , you can

4:48

just follow it up by measuring just

4:50

that one part , and that can be a very

4:52

good way to see how change is working

4:55

and identifying other ways to optimise

4:57

it . Generally

4:59

, most people would want to rebase

5:01

their numbers , a maximum

5:03

of kind of three to four years

5:06

, because

5:08

things do just change . Customers

5:10

and people change , if nothing else .

5:12

Yeah , and if you think what happened pre and post pandemic

5:15

, there's a massive change . There wasn't there so interesting

5:19

. So more than just times

5:22

again I speak to people and

5:24

they say I got handed this spreadsheet or I got

5:26

given this pie chart . That's

5:28

great . There must be more to it . And my answer is always

5:30

absolutely there is . So just talk

5:32

to us a bit about insight around

5:34

some of the studies .

5:38

So we always like to do more

5:40

than just give over the data

5:42

to people . So , yes , people can have the

5:44

raw data and go

5:46

through it as kind of anybody

5:48

else would provide to them . What we then

5:50

try to do is then say , well , from

5:52

that data , this is what it's telling you , this

5:55

is why this is the data that supports

5:57

that , and , as a result of that , here

5:59

are the quantified opportunities of things that you

6:02

could do , and here are some ideas

6:04

that you might like to look at . So we'll

6:06

always try and take it further . Partly

6:08

. There's lots of richness in the data

6:10

, so there's the different study types , but all

6:12

of them have got a degree of richness in that we

6:14

can get to partly through benchmarking , because

6:17

we've got some great data sets that we can

6:19

benchmark against , but we also

6:22

like to combine it with the observations

6:24

that our analysts make on site . You know

6:26

they're all trained observers and they spot

6:28

things that perhaps you wouldn't show up on the data

6:30

.

6:31

And I think it's like kind of some

6:33

of the data that you may get presented . It's

6:36

like having the book in the chapters but then no words in

6:38

each chapter . The insight gives you the richness and

6:40

all the detail underneath .

6:42

Yeah , unless

6:44

it's made actionable for you , then

6:46

actually it's always quite a challenge and

6:49

although we deal with this sort of data all

6:51

the time , the majority of people don't . You

6:53

know it's something that's different in you

6:55

, and anything that we can do to help people

6:57

get the most out of it is a positive thing

7:00

.

7:01

Brilliant One that's cropping up

7:03

more and more , I think , is people

7:05

are struggling with economics

7:07

, shrink wage inflation . Why

7:10

do I need a workload model ? Why do I need

7:12

to know how long things take and then build

7:14

from the bottom up to kind of suppress

7:17

from the top down to meet the financial

7:19

demands ? Why shouldn't I just go back to

7:21

Costa Cell ?

7:24

Costa Cell is just such a blunt tool , isn't it

7:26

? For one

7:28

time in my career I was running

7:30

shops that were low productivity

7:32

and it was in a tough economic

7:34

area , so my average basket

7:37

sale was pretty low , so the average

7:39

transaction value was low . I'd

7:42

got colleagues that were kind of in

7:44

much more affluent areas , that

7:46

people would buy more expensive

7:48

items and

7:52

we'd still have to put the same

7:54

number of items to the shelf . We'd have to serve

7:56

the same number of customers through the till , but

7:59

actually the value of the sales that I

8:01

was getting were lower than what some

8:03

of my colleagues would be in more well

8:05

off areas . So that's a good example

8:07

of why it needs to be different , because if you just weren't

8:09

with a cost to sell , then my

8:12

stores would have been under resourced

8:14

and potentially their stores would have been over resourced

8:16

.

8:17

Well , I think again in the current economic climate

8:19

it's an interesting debate because

8:21

there's a divergence of cost and

8:23

volume . So I could

8:25

be , if I take a

8:27

million , if I sell a million things for a pound

8:30

or one thing

8:32

for a million pounds , in a cost to

8:34

sell scenario it's

8:36

the same , but actually I've got a million times

8:38

more workload in one than two . But

8:41

ultimately at the moment , with price inflation

8:43

, prices are going up , so my sales

8:45

are going up through nothing I'm doing but

8:48

volumes probably dropping . So

8:50

actually , again in a cost to sell world , you're

8:53

masking the true impact of work needed

8:56

. So the reality , probably for most

8:58

people at the moment , is sales

9:00

are higher but there's less work

9:02

that needs doing , so I therefore need less budget

9:04

. It can't be intuitive , I get , but actually

9:06

the price increase isn't volume

9:08

, it's item

9:11

price . That's going through Any

9:14

other questions you can think of that . People

9:16

are often asking you when you're presenting

9:19

about data or in conversation

9:21

.

9:22

Perhaps one thing is about how . What's the best way

9:24

to engage their teams when they're doing these

9:27

things ? Because obviously , what

9:29

we do isn't secret . There's a person

9:32

that turns up and is observing processes

9:34

happening , so making

9:37

sure that works well by having teams

9:39

that are expecting us know what's happening

9:41

, know there aren't any secrets , is usually

9:44

the best way to go .

9:46

Yeah , and it's a balance , isn't it ? Because

9:49

sometimes the initiative is around cost-saving

9:51

, which is sensitive because as consequences

9:53

, clearly depending on the results , sometimes

9:56

it's around just actually understanding

9:58

what's happening in that business and

10:00

then making decisions off the back of the data . Sometimes

10:03

it's about putting more people in front

10:05

of customers . Sometimes it's a mix of all three

10:07

. So treading carefully

10:10

is important , but I think what

10:12

I've seen , being as transparent as you can be

10:14

at the initial briefing , is also really

10:16

important .

10:17

Yeah , I think there's . There's sort

10:19

of three steps that we'd say

10:21

is the best practice . So one is , if

10:24

there's a phone call

10:26

, then so

10:29

a phone call with perhaps the line managers and

10:31

that sort of people , so they've got a chance to ask

10:33

any questions , that's

10:35

led by the

10:38

central team , so it's their own people

10:40

, and then we're there to answer any questions . That they've

10:42

got is a great way to do it . Follow that

10:44

up with some written comms , because not everybody's

10:46

going to get to that brief , so follow

10:48

it up with some written comms . That

10:50

again sex out . We're interesting

10:53

people , not processors . We don't capture

10:55

people's names . It's not secret

10:57

. We'll happily show you the tablet . We want

10:59

to know your thoughts , that's . That's

11:02

a good way to do it . And then when the analyst

11:04

arrives , location to start

11:06

studying . If there's a team huddle or

11:08

team briefing , it's great if

11:10

our analysts can join

11:12

that say hello to everybody and again , it

11:15

just reassures everybody . So you know

11:17

, everybody knows what's happening and

11:19

why .

11:19

Yeah , and the benefits of getting

11:22

information from colleagues . So we're

11:24

independent . They can share their

11:26

gripes or their golden nuggets with

11:28

us and we can build that into the deck as well .

11:31

Yeah and then a final

11:33

question for you is are the things that we can't

11:35

measure ? So , is there anything

11:38

that you can't measure ?

11:40

I think practically you can measure anything that

11:42

happens from a . Can

11:44

I go and watch somebody doing it ? I think

11:46

there's a couple

11:49

of things that always stand out one frequency

11:51

. So you could spend a lot of time trying to capture

11:53

something that doesn't happen very often or is

11:55

weekly or monthly , or Dunning

11:57

the dead of night , whatever . And I think there's

11:59

also a bunch of stuff , and training is always

12:02

the one that is a conversation . Can

12:04

you measure training Absolutely ? How

12:06

much did you see in the study ? None , yeah

12:09

. So why is that ? Well , because people

12:11

are busy and it's one of the first things that's just

12:13

dropped or he's done it Home

12:15

, or just not happening or batched up

12:17

. And I think training is a great example

12:20

of from a workload model point of view . It's

12:22

a policy decision . So what

12:25

do you want to fund per head

12:27

, per employee , per week , month , year

12:29

, for training , and then build that into the model

12:31

? Measuring

12:33

what happens probably tells

12:35

you and I'd say 99.9

12:38

percent of times what you don't want to know , that there's

12:40

not enough of it happening or none of it . So

12:43

you've got to turn , turn it

12:45

on its head for that one and say , well , so how do we create

12:47

the funding to give the time and

12:50

then locally , how do people use and plan

12:52

that time effectively ? Sometimes

12:55

things like emails , they're tricky

12:57

. You know how many come in . An

13:00

email response could be a line , it could be ten

13:02

lines . So again , efficiency study and looking

13:04

at it proportionally rather than in absolute

13:06

decimal minutes , they're

13:09

probably there . The two , those admin

13:11

bits and certainly trainings are recurring conversation

13:13

of can you measure training ?

13:15

Yeah , and I suppose actually with the range

13:17

of techniques that we've got , we can measure

13:19

everything from things that take Small

13:22

fractions of seconds through

13:24

to kind of as long as you want to

13:26

go . I guess some of the things

13:28

that you pass and you

13:30

know even production lines and that sort of thing , there's

13:32

really easy ways video in and

13:34

then looking at those . So I think most

13:36

things , unless it was something that you know Happened over

13:38

a year or something like that , like you say , low

13:40

frequency things that don't happen very often

13:43

over a really prolonged period of time

13:45

Then they can be trickier to do .

13:47

Yeah , I think there's some other tricky bits around

13:49

processes that . So

13:51

if you think of a sales process , sometimes

13:54

you might speak to the customer and I'm talking in a

13:56

in a high-end and

13:58

furniture world . You might speak to the customer

14:00

in January , they might then decide

14:03

to buy it in April and they

14:05

might then have delivery in August

14:07

. So there's some things where you're not

14:09

going to see end-to-end of the

14:11

same customer but you can see

14:13

representative the end-to-end of the component

14:15

parts of the process because you you

14:17

would be there physically too long .

14:19

Yes , if you see every step of the process , you can

14:21

then put those bits together , even if

14:23

it wasn't one single custom .

14:24

Yeah , and you want to see that a number of times

14:26

, so you get a nice average . Yeah . And back

14:29

to your point . You know things like pizza making

14:31

, car production , that whole MTM

14:33

world then comes into play , that we've talked about another

14:35

podcast around Video

14:37

breaking down those human movements to

14:40

move the kind of time . These three I'm

14:44

sure there's plenty more . I think those are the key

14:46

ones . Maybe we do another one of these Early

14:49

in 2024 , but those are the key

14:51

ones . Hope you find that helpful

14:54

and thanks again , sue , for your time .

14:56

Thanks bye .