0:04

Hello,

0:04

listeners. Welcome to Inside

0:06

Tech Comm with your host Zohra

0:06

Mutabanna. In season three, we

0:12

shift our focus to shed light on

0:12

why Technical Communication is a

0:17

core business asset. In this

0:17

regard, we will speak with

0:21

guests who are our stakeholders,

0:21

such as product managers,

0:25

marketing professionals, UX

0:25

designers, QA and customer

0:30

support, who engage with writers

0:30

to create a seamless experience

0:34

for the customer and meet

0:34

business goals together. Let's

0:38

get started. Hi, peeps. Welcome to another

0:41

episode of Inside Tech Comm

0:45

with your host Zohra Mutabanna.

0:45

Our guest today is Lisa Zarfl.

0:50

She is a project manager and an

0:50

in-house translator at

0:53

MadTranslations based in Graz,

0:53

Austria. She has several years

0:57

of experience in the translation

0:57

industry, specifically in

1:00

managing the translation of

1:00

Madcap Flare projects. She

1:04

follows her passion by combining

1:04

the technological aspects of

1:07

documentation with the demands

1:07

of multilingual environments.

1:11

Hey, Lisa, welcome to the show.

1:13

Hi Zohra. Good to be

1:13

your guest today.

1:16

Awesome. And I'm honored to have you. You're my first guest from

1:17

Austria. Thank you so much. Yes.

1:21

And I'm excited. Excited to be

1:21

talking to you.

1:24

I'm excited too.

1:24

Yeah, I'm looking forward to

1:26

this today.

1:27

Oh, me too. So thank you for being on my show. And by the way, guys,

1:28

Lisa and I met at MadWorld 2022.

1:34

Back in June of 2022, obviously,

1:34

in Austin, and she had a

1:39

presentation on translation. And

1:39

as she was talking, I wasn't

1:42

focused as much on what she was

1:42

presenting, which I'm sorry to

1:45

admit, because I'm like, I want

1:45

to have her on my show. I

1:47

promise. I did go back and check

1:47

out her brand presentation and

1:51

her material. And it was

1:51

fantastic. And I'm just thrilled

1:54

that she decided to be here

1:54

because then I can pick on her

1:56

brains and get to know more

1:56

about how the magic happens.

1:59

Because I do not work in the

1:59

translation industry. So with

2:02

that, Lisa, take it away. Tell

2:02

us a little about yourself or as

2:05

much as you want.

2:07

Yeah, like Zohra has

2:07

already said my name is Lisa

2:10

Zarfl.. I'm in Graz right now.

2:10

So in Austria Graz is the second

2:15

biggest city of Austria next to

2:15

Vienna. I'm a localization

2:18

project manager and in-house

2:18

translator at MadTanslations,

2:23

and MadTranslations is, well

2:23

it's affiliated to Madcap

2:28

Software. So basically the

2:28

company that's developing and

2:33

distributing Madcap Flare,

2:33

Madcap Lingo and so on. I have

2:38

been with MadTranslations for

2:38

the past couple of years. And

2:43

before that, I did a degree,

2:43

basically in translation here at

2:48

the University of Graz, for

2:48

German, English and French. So

2:53

German is my mother tongue. And

2:53

I'm also working with English

2:57

and French. And yeah, we met at

2:57

MadWorld, which was such a great

3:01

experience here in Austin, I

3:01

really enjoyed going there, it

3:03

was my first time in the US. And

3:03

yeah, I also gave a presentation

3:08

on Machine Translation, which

3:08

was really, really good to do.

3:12

Absolutely.

3:12

You know it was so awesome to

3:15

run into somebody from Austria.

3:15

And to see such a diverse panel

3:20

of presenters at MadWorld. It

3:20

was definitely a great

3:23

experience, a great immersion on

3:23

all the different things that

3:26

are happening outside of

3:26

traditional technical

3:29

communication. And since I do

3:29

not work in translation, it was

3:32

good to see how machine language

3:32

is being utilized in

3:36

translation. And I think we're

3:36

gonna sort of touch upon that,

3:39

as well. Right, Lisa?

3:40

Okay. Sure.

3:40

Artificial intelligence in the

3:42

form of machine translation is

3:42

gaining more and more importance

3:43

is machine language translation,

3:43

the same as leveraging AI. I

3:47

in the field of translation.

3:47

There really have been huge

3:50

advances in the past couple of

3:50

years, especially, well, no,

3:54

maybe I should start with a

3:54

little bit of an explanation

3:58

first, for those who are not as

3:58

familiar with the machine

4:01

mean, is it an another word for

4:01

AI? Or is it using AI to do

4:02

translation process. So what

4:02

even is machine translation?

4:06

Machine Translation is a process

4:06

when a computer software

4:10

translates text from one

4:10

language to another without any

4:11

machine language translation,

4:14

human involvement. So that's

4:14

what we call a raw MT output.

4:17

And machine translation works by

4:17

comparing large amounts of

4:21

Well, in our world machine

4:21

translation is based on AI.

4:21

source and target language data.

4:21

And then they are compared,

4:26

matched against each other by a

4:26

machine translation engine. And

4:30

there are lots of machine

4:30

translation engine providers on

4:34

the market. The best known ones

4:34

are probably Google Translate or

4:38

Amazon, translate Microsoft

4:38

Translator, something along

4:42

these lines. And as I said,

4:42

there really have been huge

4:45

advances because 10 years ago,

4:45

machine translation was only

4:49

available for a couple of

4:49

languages. But as of today, most

4:53

of the world's bigger ones and

4:53

also many middle sized and

4:57

smaller ones are supported by

4:57

machine translation engines,

5:13

Something that also always comes

5:13

up when talking about machine

5:17

translation is post editing. So

5:17

post editing describes the

5:23

process by which human

5:23

linguists, so what we call the

5:26

post editor, in this case,

5:26

reviews the machine translated

5:30

output. So the MT output the

5:30

raw, MT output, and makes

5:34

corrections to it in order to

5:34

improve the outcome. So in most

5:39

cases, there will be some form

5:39

of post editing, because raw MT

5:45

output, even though there have

5:45

been huge advances still can be

5:48

pretty unreliable, especially

5:48

when you are trying to use

5:53

machine translation for any

5:53

texts that your customers can

5:56

see. Or even something like a

5:56

safety instruction or a

5:59

contract. So it's really

5:59

important that real life human

6:02

being is thoroughly checking

6:02

these texts after the machine

6:07

has pre translated.

6:08

Oh,

6:08

that's good to know. So I was at

6:10

an interview recently, Lisa, I

6:10

was asked if artificial

6:14

intelligence AI replaced

6:14

technical writers, in the

6:17

context of this conversation.

6:17

May I ask - will machine has

6:22

machine translation replaced

6:22

translators, since there has

6:26

been such significant advances?

6:29

No, it has not yet I

6:29

don't think that it will replace

6:33

human translators. I don't think

6:33

that it will fully replace human

6:37

translators, let's put it like

6:37

that. I mean, with all the

6:40

advances in the field, of

6:40

course, there are being more and

6:44

more text types that are fully

6:44

translated by machine

6:47

translation, like anything

6:47

that's maybe I don't know,

6:51

internal company letters or

6:51

emails, basically, everything

6:55

that's not customer facing can

6:55

already be well translated with

6:58

machine translation, and most of

6:58

the time, you will get a pretty

7:01

decent output. And the engines

7:01

are definitely going to get

7:05

better and better, maybe even to

7:05

the point where there's not so

7:09

much of human post editing

7:09

needed. Anyway, I, or however, I

7:13

still think that there will

7:13

always be humans involved. Just

7:18

because they are, I don't know,

7:18

it will always need humans to

7:23

really cater the texts to target

7:23

audience. So also, for example,

7:29

whenever there is some creative

7:29

writing involved, for example,

7:33

let's take marketing texts or

7:33

something like that anything

7:36

with pans with idioms,

7:36

allusions, irony, something like

7:40

that. The machine translation

7:40

engine, as good as they are,

7:43

they don't understand what they

7:43

translate. They just can't read

7:47

between the lines. Yeah, I think

7:47

humans will remain the experts

7:51

for that for quite a while. And

7:51

also, there's so much around

7:54

translation that can't easily be

7:54

replaced by machine translation.

7:59

I mean, when we're talking about

7:59

the translation process, it's

8:02

not only getting the text from

8:02

one language to another, there's

8:06

so much more trips, there's this

8:06

whole localization management

8:09

around it, when it comes to

8:09

well, taking care of the

8:13

translation memories and

8:13

organizing terminology, just

8:18

keeping everything in place,

8:18

allocating the translations,

8:21

managing everything. So the

8:21

think humans will continue to

8:25

play an important part.

8:27

I love

8:27

how you said, reading between

8:29

the lines, that machine

8:29

translation is not there. And

8:32

that was one of my I did not,

8:32

you know, argue about in such an

8:37

articulate manner? I think you

8:37

did it beautifully. It's sort of

8:41

I think it aligns with what my

8:41

theory is that it's going to be

8:45

some ways off before that can

8:45

happen. And I think the one of

8:49

the things that I said was, it

8:49

is emotion unaware, culture

8:54

unaware.

8:55

Culture is also a very important thing.

8:57

Right?

8:57

And contextualizing

8:57

to culture,

8:59

Context

8:59

all those things are so

9:02

important. And for that to

9:02

happen, is some ways of, and I'm

9:06

talking in the next three to

9:06

five years, we obviously don't

9:09

have a crystal ball that looks

9:09

into the next 10 years. And it

9:12

is advancing pretty fast. But

9:12

it's advancing

9:15

Really fast. Really,

9:15

it's exciting to see what we

9:18

will experience in our lifetime.

9:21

Yeah, and I think I mean, of course, this is not the scope of our

9:22

discussion, but I am I'm

9:25

interested to see where this

9:25

technology will take us. And how

9:28

we as humans can adapt because

9:28

this is going to happen in every

9:31

other field. It's not just

9:31

within translation or technical

9:33

writing. It's happening across

9:33

all fields. It's getting faster

9:38

and faster. It's getting faster

9:38

and faster. So as humans, we

9:40

have to stay ahead of the game.

9:40

But I think since we jumped

9:44

right into what is machine

9:44

translation, which is fantastic.

9:47

I want you to sort of take a

9:47

step back and talk about how,

9:53

what the process is at your

9:53

company.

9:56

Most of the time, it

9:56

starts with the company. Are you

10:00

wanting to expand their business

10:00

or to go global? And that's when

10:05

they first approached us because

10:05

they need translations or other

10:09

language services to do so. So

10:09

basically, it is really it's

10:14

often technical writers who are

10:14

approaching us first. And it's

10:19

while there are project managers

10:19

like myself, who are guiding the

10:23

technical communicator through

10:23

the whole process. So for us,

10:28

the first step is always to

10:28

evaluate our potential new

10:33

partner's needs. So we are

10:33

definitely going to have a look

10:38

at the source material. And

10:38

we're also going to discuss the

10:42

service levels with well, the

10:42

technical writer in this case.

10:47

So is translation maybe

10:47

sufficient? Or do we rather

10:51

recommend translation and

10:51

revision? Or is the text suited

10:55

for machine translation, we also

10:55

always evaluate that something

10:59

that's also important for us to

10:59

figure out at first is if there

11:03

are any additional needs. So for

11:03

example, are there any images

11:08

that need to be localized or

11:08

screenshots or something like

11:11

that? Is there going to be DTP

11:11

work, or anything, basically, so

11:16

we're just evaluating the needs

11:16

and discussing everything with

11:21

the technical writer in this

11:21

case. And well, during this

11:25

whole process, it's very

11:25

important for us that our

11:29

contact persons, so the

11:29

technical writer just reaches

11:32

out to us if there are any

11:32

questions coming up. Because as

11:36

much as we are the experts when

11:36

it comes to the localization of

11:39

content, the technical writers

11:39

are definitely the experts when

11:43

it comes to their material. And

11:43

I really think that both parties

11:46

have to well work together,

11:46

that's really indispensable in

11:51

my opinion, just to achieve the

11:51

best possible result. So if

11:56

there are any questions during

11:56

this phase, please always feel

11:59

free to ask your language

11:59

service provider any questions

12:02

because you're potentially

12:02

looking for long term partner.

12:05

So it's important to figure all

12:05

these things out. And after

12:09

we've evaluated the need and

12:09

needs an answer to the

12:12

questions, we're going to

12:12

prepare a code for potential new

12:18

client. And if the quote gets

12:18

accepted, we get the translation

12:23

project started right away. We

12:23

give the translations to our

12:29

experienced translators, either

12:29

in house translators like

12:32

myself, or we're also working

12:32

with lots of freelance

12:36

translators around the globe,

12:36

but they are definitely all of

12:40

them are experienced

12:40

translators. They are

12:42

translators, according to the

12:42

ISO 17100 standard. So this

12:47

means they either have formal

12:47

education and some experience or

12:50

they have a certain amount of

12:50

years of experience, which also

12:53

qualifies them to do

12:53

translations for us. And we also

12:57

make sure that the translators

12:57

are specialized in the fields of

13:01

the prospective client. And we

13:01

always encourage our translators

13:06

to ask questions, if they don't

13:06

understand something, or if they

13:11

need more context, to fully

13:11

grasp the meaning of something

13:15

in the source text. I personally

13:15

think that's really important.

13:20

Because, of course, they are not

13:20

the experts for exactly this

13:25

product or software. But I think

13:25

it's very important that they

13:28

get an understanding of it, that

13:28

they understand what they're

13:32

translating, do the research. So

13:32

whenever there are any

13:35

questions, I definitely

13:35

encourage that they send them to

13:38

me, and I get in touch with the

13:38

technical writers, hey, they

13:42

just sent me some questions,

13:42

would you be so kind as to have

13:45

a look and try to help them out?

13:45

And we really appreciated it, if

13:50

technical writers take some time

13:50

out of their day and try to help

13:54

out translators, because that's

13:54

really well, it's, it's

13:57

improving the overall outcome, I

13:57

think. And as soon as the

14:01

translations are ready, we do

14:01

all of our texts in our

14:06

translation software. And we

14:06

re-export the files to the

14:11

original source file format. So

14:11

this can be anything from well,

14:16

in our case, it's often the

14:16

Madcap flare project or

14:18

Microsoft Word, Excel files,

14:18

JSON properties, really

14:22

basically anything. We're doing

14:22

our final checks there and then

14:25

we are delivering the project to

14:25

the client. We do appreciate

14:29

feedback if anything is coming

14:29

up. And some clients also like

14:34

to do well a client review

14:34

cycle. So what we often call SM

14:39

e review, subject matter expert

14:39

review. What I'd like to say

14:43

about this subject matter expert

14:43

reviews. If you intend to do

14:48

something like this, please just

14:48

get in touch with your language

14:53

service provider. Tell them

14:53

about it, because there are good

14:57

ways to do such a review. For

14:57

example, we offer either during

15:01

the review in our web-based

15:01

translation service, or in

15:05

bilingual RTF files that allow

15:05

us to update the translation

15:09

with your changes. Just if you

15:09

just correct the translation,

15:13

let's say in the in the Word

15:13

file we're delivering or if you

15:16

make comments in PDF files, it's

15:16

much more difficult and

15:21

sometimes impossible to update

15:21

the translations in our

15:24

software. So please, just if you

15:24

want to do an SME review, talk

15:28

to the language service provider

15:28

about it, there are right ways

15:31

to do it. That makes life for

15:31

both parties much easier.

15:35

So you

15:35

talked about the whole process,

15:37

which is fantastic. A lot of

15:37

questions that have come out of

15:40

that. One of the questions that

15:40

I want to sort of get out of the

15:43

Human translation.

15:43

Human translator does human

15:44

way is, you mentioned as you are

15:44

evaluating the needs, you look

15:48

at the content, and you decide

15:48

whether it is good for

15:51

translation or machine

15:51

translation. So I'm, I'm

15:54

assuming that the translation

15:54

you mean to say, manual translation?

15:59

translation and machine

15:59

translation basically,

16:02

How do

16:02

you discern whether the content

16:04

is suitable for one or the other?

16:06

All contents are

16:06

suited for human translation.

16:10

That's just how it has been for

16:10

many, many years. But when it

16:14

comes to machine translation,

16:14

well, there are certain text

16:18

types that are better suited for

16:18

machine translation than others.

16:22

Like I've already stated before,

16:22

everything that's non customer

16:26

facing is definitely good for MT

16:26

like internal company letters,

16:31

news articles, intranet posts,

16:31

they mostly have simple contents

16:35

and are ideal candidates for MT,

16:35

but also everything in the field

16:40

of technical documentation. So

16:40

most operating instructions or

16:45

software manuals I'm

16:45

encountering are well suited for

16:49

MT, except for projects that are

16:49

very technically complex. So I

16:56

can just talk from my experience

16:56

with Madcap Flare projects, if

16:59

there are lots of very complex

16:59

conditions and variables in

17:03

there, that would probably not

17:03

recommend using machine

17:06

translation, just because then

17:06

the engine doesn't really

17:10

understand these concepts and

17:10

can't make much of it. And text

17:14

types that are not so well

17:14

suited for machine translation

17:17

in general is, like we've

17:17

already said anything that's

17:19

creative for contains cultural

17:19

references. Also, complex legal

17:24

texts are not so well suited for

17:24

machine translation, just

17:28

because they normally have

17:28

really highly specialized

17:31

terminology and long and complex

17:31

sentences that are really

17:34

important. So we rather do human

17:34

translation for those. And also

17:39

something that is nearly always

17:39

translated by human translators,

17:44

at least in our company, UI

17:44

strings. So the user interface

17:49

texts of Office software, for

17:49

example, because MT engines work

17:55

best with longer sentences and

17:55

larger textual context. And UI

18:01

strings. They are often

18:01

submitted, for example, in

18:03

Microsoft Excel format. And

18:03

there's mostly unrelated terms.

18:09

And often there's only one word

18:09

per cell. And that word might

18:13

even have several meanings.

18:13

Let's, for example, take open,

18:17

that's a very common English to

18:17

string. But it can be a verb to

18:22

open something, or it can also

18:22

be an adjective to be open. So

18:26

that's really, really difficult

18:26

for an MT engine, it's already

18:30

difficult for human translator,

18:30

they often ask questions when it

18:33

comes to things like that.

18:35

I would

18:35

have never thought that there's

18:37

so much that goes on with

18:37

translation behind the scenes.

18:41

It's pretty complex.

18:41

But I think that also applies to

18:44

technical writing. I mean, all

18:44

of these processes are so much

18:47

more complex if you're in the middle of it.

18:49

Absolutely,

18:49

yes. Just the thought pure. In

18:53

your opinion,

18:53

internationalization versus

18:56

localization, I've always

18:56

struggled with that. Maybe you

19:00

can be my SME and tell me what

19:00

the difference is.

19:03

So localization

19:03

basically refers to localizing a

19:08

certain product or software to a

19:08

specific country to a specific

19:12

market. So really localizing to

19:12

the local culture and the laws

19:18

and the concepts that exist in

19:18

this country. And

19:21

internationalization more refers

19:21

to well, kind of streamlining

19:27

the original product or

19:27

software. So to remove or make

19:35

easier, everything that would

19:35

need to be localized afterwards.

19:39

So basically, to just make the

19:39

source product easier, and

19:45

taking out contents that would

19:45

need localization afterwards.

19:49

I think

19:49

that makes sense. You know, as

19:52

you're sort of sharing your perspective on internationalization, from my

19:54

perspective, I'm thinking okay,

19:58

we're talking about how does

19:58

content in the context of

20:02

technical communication

20:02

contribute to business value?

20:06

So, as a technical writer, if

20:06

you're writing content you want

20:09

to be, even if you're not sure

20:09

if this content is going to be

20:12

translated, you want to be aware

20:12

about how your content is being

20:16

written so that it is devoid of

20:16

cultural nuances. I think the

20:20

cool things that you said

20:20

earlier in the interview puns,

20:23

idioms, allusions, ironies, that

20:23

kind of stuff, so that it is it

20:27

lends itself better to

20:27

localization down the road.

20:30

That's a good

20:30

explanation. Thank you.

20:32

Thanks

20:32

to you. I mean, you brought it

20:33

And also content

20:33

that's consistent and precise.

20:35

all together for me. So just

20:35

like as much as this is, I'm

20:39

synthesizing information, as

20:39

you're giving it to me, because

20:43

I'm like, the bottom line is how

20:43

are we? How is content

20:46

contributing to business value?

20:46

That is my focus for season

20:50

three. So far, all the things

20:50

that you've talked about right?

20:54

Machine Translation, the open

20:54

dialogue between technical

20:58

writers and translators,

20:58

important, all that sort of

21:01

starts contributing to okay, if

21:01

you're keeping all these things

21:05

in mind ahead of time, then the

21:05

work that needs to happen

21:09

downstream, can save you costs.

21:09

So I don't create content that

21:13

is going to be translated, but

21:13

being aware of that. And

21:17

sometimes it's just as part of

21:17

our training, you know, it says

21:21

in the styleguide don't do this.

21:21

And I don't pay attention to

21:25

that. Why, but it makes sense

21:25

now, right? It's like, oh, that

21:29

aha moment? Yes. In any case, I

21:29

need to create content that is

21:33

that is devoid of all these

21:33

references and nuances,

21:33

And

21:33

precise, important, very

21:41

important. I realized, at some

21:41

point, I said, manual

21:44

translation. And I know that as

21:44

we move towards more inclusive

21:48

language, human translation, so

21:48

I had to stop and correct

21:52

myself, it's important that I

21:52

also become aware of how the

21:55

language is also changing in my

21:55

industry, so thank you for that.

22:00

I'm really having fun. We've

22:00

talked about a lot of stuff

22:02

here. Lisa, you know, now down

22:02

to the meat of the question.

22:06

You've touched upon how

22:06

translation can save time. But

22:10

in terms of saving cost, apart

22:10

from the things that you have

22:14

already mentioned, what can you

22:14

think of that technical writers

22:17

can do to save costs downstream,

22:19

I'd like to talk

22:19

about three topics, if possible,

22:22

please, starting out with

22:22

translation. Well, what we call

22:26

basically translation oriented

22:26

writing. So I think this one

22:30

especially applies to technical

22:30

writers, because you can already

22:35

keep a potential translation of

22:35

the material in mind from the

22:39

start, basically, while doing

22:39

the technical documentation.

22:43

Well, there are just some ground

22:43

rules that you can easily follow

22:48

that will help to save cost in

22:48

the long term. So starting out

22:52

with, try to formulate your

22:52

sentences, simply and precisely.

22:56

So basically, when you're

22:56

writing in English, this means

23:00

like Subject, Verb, Object, try

23:00

to really keep them concise. So

23:05

I'm not saying that you should

23:05

write your sentences, overly

23:09

simplistic, but just try to

23:09

avoid traces that go on and on

23:13

and on and on. And something

23:13

that's also really important is,

23:18

well, terminology. Let's say you

23:18

are talking about a laser

23:22

scanner, for example, in your

23:22

documentation, then call it a

23:26

laser scanner, and not device

23:26

for example. And after you have

23:30

chosen your term, try to stick

23:30

to the same term throughout. So

23:35

it's really important to be

23:35

specific and to use unambiguous

23:39

terminology. Well, I'd also say,

23:39

be careful with abbreviations,

23:43

especially the non official

23:43

ones. Try to explain them

23:47

somewhere because, well, the

23:47

translator will not know your

23:51

company internal translate

23:51

abbreviations. Also try to avoid

23:55

filler words, try to avoid

23:55

incomplete sentences, it's

23:59

really important to try to stay

23:59

as consistent as possible,

24:03

because consistent texts are

24:03

what is really important. For

24:07

the second topic I'd like to

24:07

discuss in these are translation

24:12

memories. So I just quickly like

24:12

to explain the whole process

24:16

when you're sending us a text to

24:16

translate. After evaluating

24:20

everything, we import that text

24:20

into our translation software.

24:25

And this translation software

24:25

must not be confused with

24:29

automated machine translation.

24:29

So that's not the same thing.

24:33

Automated machine translation is

24:33

really, as we've said, no human

24:37

involvement machine translates

24:37

text and a translation software.

24:42

It's just a software where

24:42

translators and agencies will

24:46

translate the text basically, we

24:46

don't translate it in Word or

24:50

something. We do translate them

24:50

in a translation software. Well

24:55

known translation softwares, for

24:55

example, our MemoQ or SDL,

24:59

Trados Studio, for example. And

24:59

in this translation software,

25:03

there are two columns. There's a

25:03

source column and the target

25:08

column. So there's the source

25:08

text on the left, and

25:11

translation goes on the right.

25:11

And the text is split up into

25:15

different segments. And

25:15

normally, one sentence

25:19

constitutes one segment. That's

25:19

how the text looks for a

25:23

translator, and then translator

25:23

starts to translate, translate

25:27

all the segments, and they're

25:27

kind of segment pairs they match

25:31

together. And after the

25:31

translation, after we did all of

25:35

the checks, we save all these

25:35

translation pairs in a

25:39

translation memory. That's

25:39

basically a database where all

25:43

the translations for this

25:43

specific projects are saved. And

25:47

now let's say the client is

25:47

sending the same manual again,

25:51

six months later, of course, he

25:51

has continued working on the

25:56

manual, and it's an updated

25:56

version. Now, they are sending

26:00

the manual again, we are

26:00

importing it again into our

26:04

translation software. And then

26:04

we run the text against the

26:08

translation memory to see what

26:08

has already been translated

26:12

before. So everything that has

26:12

been translated before is what

26:16

we call 100% match, or even 101%

26:16

match if the context is also the

26:21

same. And the segment's are pre

26:21

translated by our translation

26:25

memory. And we do not charge for

26:25

them anymore. So the client is

26:30

only paying for anything that's

26:30

new, or has changed. And of

26:34

course, over time, more and more

26:34

translations are being saved

26:38

into the database into the

26:38

translation memory. And this

26:42

helps to tremendously reduce the

26:42

cost over time, because there's

26:47

more and more within the

26:47

database. This just leads me to

26:50

the third and last point, to

26:50

save money with translations.

26:55

This is basically centralizing

26:55

your translation business. So I

26:59

do understand that often, you

26:59

will have different language

27:03

service providers, different

27:03

vendors you send translations

27:07

to. But in the long term, it's

27:07

really best to form a long term

27:12

partnership with one translation

27:12

service provider, just because

27:16

of all the translation memories.

27:16

And well, it's just a

27:20

partnership is developing. And

27:20

of course, if you are asking us

27:24

for your translation memories,

27:24

we will send them to you this is

27:29

your content, you own these

27:29

translation memories, you can

27:33

send them to different vendor,

27:33

for example. But we also often

27:37

receive translation memories

27:37

from the clients that they have

27:41

received from their previous

27:41

vendors, for example. And it's

27:46

almost inevitably leading to

27:46

inconsistencies just because you

27:50

don't know what quality these

27:50

teams are having. And they have

27:54

different translators maybe

27:54

working on them. So in the long

27:59

term, you should really try to

27:59

centralize your translation

28:03

business with one vendor just to

28:03

make sure that everything stays

28:07

as consistent as possible. So

28:07

those are my three tips for

28:11

saving money, basically,

28:11

translation oriented writing,

28:15

keeping a translation in mind

28:15

from the get go, profiting off

28:19

translation memories and

28:19

centralizing all the translations.

28:24

I mean,

28:24

it sounds easy to do, but I'm

28:26

sure it's not right. Because

28:26

you're doing this over time. And

28:29

if especially if you're doing it

28:29

for the first time, you really

28:32

have to put in that effort. Have

28:32

there in your experience, has

28:36

there been a point in time where

28:36

something was so like, where you

28:40

had to reject and say we cannot

28:40

translate, or they had to just

28:43

go back to the drawing board and

28:43

just start from scratch?

28:45

We've had some

28:45

pretty difficult projects in the

28:49

past that posed a lot of

28:49

challenges for us, and for the

28:54

translators. And for the client.

28:54

I think we have never said, we

28:59

can't help you at all. Don't

28:59

think that we've ever done that

29:02

before. We are always trying to

29:02

find the solution with the

29:06

client. Maybe it's not exactly

29:06

what they have been imagining in

29:10

the first place. But we're

29:10

always trying to explain where

29:13

the issues are, and why some

29:13

things might not be possible as

29:17

they're imagining them. But then

29:17

we try to work together with

29:20

them to find solutions, for

29:20

example, well, we have already

29:25

helped client fans to just

29:25

rework the source text, for

29:30

example. So have some of our

29:30

linguists help them to improve

29:35

on their source texts. We often

29:35

help clients with their Madcap

29:38

Flare projects, if they're

29:38

struggling with the technical

29:41

issues there. Yeah, I don't

29:41

think we've ever said we can't

29:44

do anything for you. We're

29:44

always trying to.

29:47

Yeah, I think that was probably an extreme question, but I just

29:49

wanted to kind of see, you know,

29:52

for example, if there have been

29:52

real challenges, but I think I'm

29:56

going to sort of flip what you

29:56

gave us and say everything As

30:00

you said, right, if you can take

30:00

care of these things upstream,

30:03

then it doesn't become a problem

30:03

later. So you're gonna say

30:05

extract with translation, and

30:05

then you're going to reach a

30:09

larger target audience. So the

30:09

company reach their content is

30:13

reaching a larger audience, it

30:13

is localized. And if you're

30:16

taking care of all these things,

30:16

you're saving cost with

30:19

translation, you are expanding

30:19

your business. And eventually,

30:23

my point of this season is

30:23

content, technical communication

30:28

is bringing value to the bottom

30:28

line. So there is a lot of

30:32

thought and process that needs

30:32

to go into product

30:35

documentation, technical content

30:35

that is being created, be

30:38

internal, anything that is going

30:38

to reach a target audience has

30:41

to be well thought out. And that

30:41

investment has to be made. And I

30:46

think there are many companies

30:46

where there is this challenge

30:50

with justifying why, why you

30:50

need a technical writer on the

30:53

team. And I think this sort of

30:53

lends to that, you need to think

30:57

the why I think this probably

30:57

answers the why and more.

31:01

Totally understand

31:01

it's the same for translators.

31:03

Yes. And so that's the thing that I wanted to sort of bring to light

31:05

what, what LSPs, dual language

31:09

service providers do, am I

31:09

right? Is that the right term?

31:12

Lisa? LSP Right. Okay. And, and

31:12

sort of elevating all these

31:16

adjacent disciplines that sort

31:16

of come together to create

31:20

content. It's not just, oh, I

31:20

can go write a sentence. And I'm

31:23

done with writing, there is a

31:23

lot of thought and strategy that

31:26

goes into bringing that content

31:26

to audience. And yes, nobody, I

31:30

think will agree that it's fun

31:30

to read a manual, or...

31:33

Content can be more

31:33

fun if it's well written.

31:37

Exactly,

31:37

and experienced professionals

31:41

have are doing a good job. And

31:41

there are a lot of success

31:44

stories. But those are

31:44

overshadowed by content that may

31:48

not be well written. So we need

31:48

to focus on why that may be

31:52

happening, and what can we do to

31:52

address it, and we need the

31:54

support of the higher ups to

31:54

sort of because at the end of

31:58

the day, it is going to be

31:58

something that is going to

32:00

contribute to business value

32:00

over the long term. And taking a

32:03

short term perspective on things

32:03

is detrimental to your business.

32:07

I think you make an absolutely

32:07

fantastic case for this. We've

32:10

had a great conversation so far.

32:10

Lisa, I want to make sure that

32:13

I've covered all the questions

32:13

that I had in mind. But is there

32:15

anything else that you would

32:15

like to add an expert insight

32:19

that I may not have touched upon?

32:21

More, we've talked

32:21

about localization process, but

32:24

machine translation, cost

32:24

saving? Very important. I'd like

32:30

to contribute something to what

32:30

you just said, with the business

32:33

value? You're absolutely right.

32:33

And I think it's, I mean, I

32:38

think the technical writing is,

32:38

is a pivotal part of all the

32:43

little things that have to play

32:43

together and that are essential

32:47

to make a product or process

32:47

work. As is translation, if you

32:51

are going global with your

32:51

product. For technical writers,

32:54

I think it's really important to

32:54

well to highlight their

32:58

expertise, but also to openly

32:58

communicate their needs. But I

33:03

think it's also important to

33:03

keep in mind that the people

33:05

they are talking to at their

33:05

companies often don't have the

33:09

same well field of expertise.

33:09

And what's most importantly,

33:13

probably they have different

33:13

expectations. So I think it's

33:17

it's important for technical

33:17

writers as well as for

33:19

translators, basically,

33:19

everybody communicating with

33:23

other departments. Just explain

33:23

why you need something, why it

33:28

has to be done like that. Yeah.

33:28

If nothing helps anymore, it's

33:32

always a good idea to visualize

33:32

costs, maybe that could result

33:36

from bad documentation, because

33:36

people underestimated bad

33:40

documentation really can well,

33:40

as you said, can be detrimental.

33:45

Yeah. And I think the critical thing that you touched upon was the

33:46

why, as much as we say that we

33:50

don't have the buy in technical

33:50

writers, I think I've said this

33:53

before, is that we have to step

33:53

up, technical communicators have

33:57

to step up and take these other

33:57

professions that sort of work in

34:01

tandem with us to elevate all of

34:01

that, and to and to put it in a

34:04

business sense and say why this

34:04

is important. This is what I'm

34:07

going to do. Please help me out.

34:07

But no, why is this important to

34:11

the bottom line? I think that is

34:11

something that I myself am

34:14

grappling with, how do I sort of

34:14

like you said, we are pivotal.

34:17

But how do we communicate that,

34:17

in a business sense, is equally

34:21

important? And I'm trying to I

34:21

don't have an MBA. So I think

34:25

I'm still trying to figure that

34:25

out. But yeah, you make a good

34:27

point that talking in numbers

34:27

probably will make sense.

34:31

I mean, that really

34:31

speaks to managers in general,

34:35

In general, I agree. And that's, that's something that I have to

34:37

kind of grow myself to grow

34:41

into. And probably all of us are

34:41

on that journey, where we are

34:44

trying to figure out how do we get there, how do we start speaking in those terms, so that

34:46

we are visible and we kind of

34:51

make our value known in a value

34:51

sense.

34:58

Language people in

34:58

general have to figure out for

35:01

the future. Thank you.

35:02

I think we all have to put our heads together to figure that

35:05

Same issues when.

35:07

Exactly,

35:07

exactly. So this has just been

35:10

an amazing, amazing

35:10

conversation. I say this every

35:13

single time but every single

35:13

time I walk away learning so

35:15

much, thank you for sharing all

35:15

your insights with us. I hope I

35:19

get to visit Austria someday,

35:19

and I would love to come hang

35:23

out with the you.

35:24

Please come visit.

35:24

It would be amazing. Yeah, I can

35:27

show you the Alps and the lakes

35:27

and forests. And we have been

35:32

talking about Vienna before the

35:32

podcast so you can come and

35:34

visit Vienna and Graz.

35:36

I want

35:36

to. I envy you. I see those.

35:39

Which movie is that?

35:41

The Sound of Music,

35:41

maybe. It's very popular among

35:45

Americans.

35:46

I love

35:46

it. I'm like, I want to be

35:48

there. So someday, I'm going to

35:48

come and live my moment.

35:51

Sure,

35:52

In Austria with you. Thank you so much.

35:54

Thanks so much for

35:54

having me. It was a really great

35:56

session.

35:57

Absolutely. Subscribe to the podcast on your favorite app, such as Apple, Google, or Spotify. For the latest on my show, follow me on LinkedIn, Instagram, or visit us at www.insidetecomm.show. Catch you on another

35:57

episode.