User research and usability testing with Maze | IT's time to talk #2

00:00 → 00:03

No digital product or service can be built

00:03 → 00:06

without users' feedback.

00:06 → 00:09

We need to know what kind of features can be used by

00:09 → 00:13

particular types of users in order to know what will

00:13 → 00:17

be needed and what will gain popularity among different

00:17 → 00:18

kinds of audiences.

00:18 → 00:20

We need to take into consideration their tech

00:20 → 00:23

savviness, their age, their cultural background,

00:23 → 00:26

their education, and thousands of different factors.

00:26 → 00:30

Today, we will be discussing this topic with Oskar Czubacki and

00:30 → 00:31

Hanna Klich.

00:31 → 00:34

Ladies and gentlemen, this is Merixstudio series.

00:34 → 00:35

It's time to talk.

00:41 → 00:43

Hello, guys. How are you doing?

00:43 → 00:44

Hello, everyone.

00:44 → 00:46

Hi. We're good.

00:46 → 00:47

Alright then.

00:47 → 00:51

As seasoned product designers, you could

00:51 → 00:54

share some fundamental principles

00:54 → 00:58

with me about gathering users' feedback.

00:58 → 01:01

What is the most important here?

01:01 → 01:04

What kind of factors we need to take into consideration?

01:04 → 01:06

How we should prepare for this process?

01:08 → 01:12

So maybe I can start this one.

01:12 → 01:16

Well, if you are talking about fundamental principles in

01:16 → 01:18

user research,

01:18 → 01:23

first, I think you need to define your clear goals.

01:23 → 01:24

I your clear goal.

01:24 → 01:25

So

01:26 → 01:30

you need to know what you want to learn from the test.

01:30 → 01:34

Clear goals definitely help in designing that test

01:34 → 01:37

and also in interpreting the results.

01:37 → 01:41

And, also, if we're looking about user research,

01:41 → 01:46

it's obvious that we also need to recruit the right participants.

01:46 → 01:50

So you need to ensure that your participants represents

01:50 → 01:52

your target audience.

01:52 → 01:57

So this makes your the the feedback

01:57 → 02:02

more relevant and more, let's say, actionable.

02:02 → 02:03

Yeah.

02:03 → 02:06

I think also it depends on the type of testing you're doing.

02:06 → 02:10

So for example, for moderated testing where you have the

02:10 → 02:14

researcher present with the users,

02:14 → 02:16

you have to focus on the users.

02:16 → 02:19

So don't create a so called researcher effect.

02:19 → 02:23

Don't try to lead them or influence researcher effect?

02:23 → 02:27

So it it's a it's an effect where the the candidate or the

02:27 → 02:31

the user kind of responds to please you.

02:31 → 02:35

So what they do is just to make you feel satisfied as the

02:35 → 02:40

researcher, and it can happen when you are maybe

02:40 → 02:43

making yourself too much of a star during the process.

02:43 → 02:46

So you you really should be

02:46 → 02:48

attentive to what they are doing,

02:48 → 02:53

but try not to hover too much over, you know, their process.

02:53 → 02:55

Remind them to think aloud.

02:55 → 02:59

Obviously, follow on their, like, hesitation moments.

02:59 → 03:03

So if they do those or oh,

03:03 → 03:05

try to gather more insight on this,

03:05 → 03:07

but don't try to be the star yourself.

03:07 → 03:11

Like, their feedback matters, more than, like,

03:11 → 03:14

you being con too confident or overly confident.

03:14 → 03:18

It can be also discouraging to the participants if you are

03:18 → 03:20

too loud maybe.

03:20 → 03:22

Some of them may feel discouraged,

03:22 → 03:25

and definitely this will be damaging to the results.

03:26 → 03:28

Okay then. Thank you for this.

03:28 → 03:31

And what is

03:32 → 03:35

crucial when, for example, selecting the participants?

03:35 → 03:39

What kind of factors we should take into consideration while,

03:39 → 03:43

for example, choosing the right group of people?

03:43 → 03:46

I think it's our favorite answer. It depends.

03:46 → 03:49

So the demographics, I would say,

03:49 → 03:52

would always be different and depend on your type of projects

03:52 → 03:54

and on the objectives that you have.

03:55 → 03:58

So you can't really say they will be the same, like,

03:58 → 04:01

age group or or target audience each time.

04:01 → 04:05

That'll be very individual and tailored to the the business

04:05 → 04:07

objective, for example.

04:07 → 04:11

But I think it's important to make this cast of users

04:11 → 04:15

diverse and also diverse from yourself because only then you

04:15 → 04:18

will get unbiased kind of opinions.

04:18 → 04:21

Otherwise, you'll you'll kind of just confirm you'll have this

04:21 → 04:23

converse confirmation bias that, you know,

04:23 → 04:26

the same group of people that's very similar to you will

04:26 → 04:28

confirm your assumptions.

04:28 → 04:30

So Okay. Just be mindful of that.

04:31 → 04:32

Yeah.

04:32 → 04:35

Also, if you have to talk factors

04:37 → 04:39

of the target audience, it's also, for example,

04:39 → 04:40

tech savviness.

04:40 → 04:43

That's also important, for example,

04:43 → 04:47

in terms of just conducting data research because maybe

04:48 → 04:51

with the people that are, let's say, less tech savvy,

04:51 → 04:55

then you need to prepare yourself to explain something small

04:56 → 04:58

more detailed or in more, let's say,

04:58 → 05:01

easier way for them.

05:02 → 05:02

Yeah.

05:02 → 05:04

And and it's useful to have, you know,

05:04 → 05:06

a proper sample of these users.

05:06 → 05:10

Typically, six to eight people give you some

05:10 → 05:15

results already, so some patterns will be emerging in their behavior.

05:15 → 05:17

But there are recommendations even that you test with

05:17 → 05:20

fifteen, so three sessions with five users.

05:20 → 05:23

That's recommendation that some time ago Jacob

05:23 → 05:27

Nielsen made may lead to more cost effective,

05:27 → 05:31

you know, changes just because the you you can conduct free

05:31 → 05:35

research session with with more participants, really.

05:35 → 05:39

So it can be giving you better results in the end.

05:40 → 05:43

So ladies and gentlemen, your stakeholders,

05:43 → 05:45

people who want to build a product or service,

05:45 → 05:49

please keep in mind, eight, maybe a minimum.

05:49 → 05:53

In many cases, group of fifteen people should give us some answers.

05:53 → 05:56

What is needed and what is not needed? Yes?

05:56 → 05:57

Very often, as Merixstudio,

05:57 → 05:59

we are working with a different groups of people,

05:59 → 06:03

with facility managers, with warehouse managers,

06:03 → 06:06

with electricians, with

06:07 → 06:11

with people who are responsible for maintaining conveyor

06:11 → 06:14

belts in manufacturing on manufacturing sites.

06:14 → 06:16

Yes? These people have different backgrounds.

06:16 → 06:18

Those people have different knowledge,

06:18 → 06:21

and it's extremely important to get to know what is needed and

06:21 → 06:24

how do they understand what they see when they're

06:24 → 06:27

interacting with a product.

06:27 → 06:29

Okay then.

06:29 → 06:32

For a while, we are working with a software

06:32 → 06:34

which is called Maze.

06:34 → 06:38

And I'm wondering how much amazing this Maze

06:38 → 06:42

is and if you could tell me a little bit more what is,

06:42 → 06:45

first of all, the benefit from using it and what is the

06:45 → 06:48

difference between, let's say, doing traditional tests and

06:48 → 06:50

tests doing?

06:50 → 06:51

Well,

06:53 → 06:55

so if we're talking about the

06:55 → 06:58

maze, the let's say,

06:58 → 07:00

one of the biggest advantage of maze is that you can

07:00 → 07:03

run underrated tests.

07:03 → 07:07

And, also, you still can get a

07:07 → 07:11

lot of feedback and a lot of analysis from the from this

07:11 → 07:17

unmoderated tests basing on the features

07:17 → 07:21

offer, which may offer, like heat maps,

07:21 → 07:24

like test recordings,

07:24 → 07:26

and and and others.

07:26 → 07:27

So

07:29 → 07:34

that's the you using Maze to for your test

07:34 → 07:39

for your research for user research allows you to access,

07:39 → 07:44

let's say, more diverse and a group of participants

07:44 → 07:47

without, for example, geographical construct

07:47 → 07:48

constraints, for example.

07:48 → 07:49

Right? Okay.

07:49 → 07:54

So if you can easily run remote testing

07:54 → 07:58

at any time, reducing also the time needed for

07:58 → 08:02

scheduling meetings with moderate tests and so on.

08:02 → 08:03

Yeah.

08:03 → 08:05

Testing outside your time zone,

08:05 → 08:09

that definitely can be streamlined with Maze.

08:09 → 08:13

But I feel think also just creating the tasks,

08:13 → 08:16

it comes with a lot of modules as Oscar said.

08:16 → 08:19

You know, you can test heat maps and other elements,

08:19 → 08:21

but even building the scenario for the task,

08:21 → 08:24

it's it's almost like putting Lego together.

08:24 → 08:26

So it's it's quite fast.

08:26 → 08:30

You know, you can do it in a in in a couple of minutes even or

08:30 → 08:33

change the, let's say, the order of tasks that you

08:33 → 08:36

schedule for the participants.

08:36 → 08:40

So in that way, it's easier to build than, you know,

08:40 → 08:43

redoing the scenarios every time.

08:43 → 08:44

That's great.

08:44 → 08:45

One one more thing,

08:45 → 08:48

I I think is great also is aggregated results.

08:48 → 08:53

So you have, obviously, everyone everyone has AI now.

08:53 → 08:56

There is some AI features there, but even on this, like,

08:56 → 08:58

very, very basic algorithmic level,

08:58 → 09:00

it gives a good aggregate of the data.

09:00 → 09:02

So you have, like,

09:02 → 09:05

almost like an overview of what has happened even before diving

09:05 → 09:09

into the recordings, and that's that's really nice.

09:10 → 09:10

Yeah.

09:10 → 09:13

And one one more thing that is also important before I think

09:13 → 09:17

about Maze is that possibilities of Maze

09:17 → 09:20

integration with other tools.

09:20 → 09:24

Tools like Figma or Adobe,

09:24 → 09:26

you can easily integrate your, for example,

09:26 → 09:31

prototype from Figma directly into Maze, and, you know,

09:31 → 09:34

all the research magic is happening inside Maze.

09:34 → 09:38

You don't have to go outside and run, for example,

09:38 → 09:42

Figma prototype or or whatever. Everything is in one place.

09:42 → 09:44

Heatmaps are based on your prototype,

09:44 → 09:47

so that's that's great.

09:48 → 09:49

Alright then.

09:49 → 09:52

You mentioned something about different time zones.

09:52 → 09:54

Does it mean that you can perform these tests

09:54 → 09:56

asynchronously?

09:57 → 09:57

Yeah.

09:57 → 10:00

For For unmoderated testing, you just share a link and

10:00 → 10:02

can share it to whoever you like.

10:02 → 10:03

Right?

10:03 → 10:06

So if we are in Poland and we are running a test in Japan,

10:06 → 10:07

that's still fine.

10:07 → 10:11

The candidates can do it at their convenient time.

10:11 → 10:14

We don't have to be present for an unmoderated tests,

10:14 → 10:16

but the recording will be available.

10:16 → 10:20

And we can also share the same link to the session to actual

10:20 → 10:22

stakeholders in Japan.

10:22 → 10:23

So, you know,

10:23 → 10:28

you you will have an option to view wherever you're from.

10:28 → 10:29

Okay.

10:29 → 10:32

I think this is very convenient given the fact that

10:32 → 10:36

when we are talking with our clients about the necessity of

10:36 → 10:39

performing tests, they usually have a

10:39 → 10:42

problem with gathering a representative group at the

10:42 → 10:46

same time or similar, let's say, time slot.

10:46 → 10:46

Yeah?

10:46 → 10:49

And it sounds like a major

10:49 → 10:52

convenience to be in a situation when the stakeholder

10:52 → 10:54

can, for example, ask, I don't know,

10:54 → 10:58

their employees or their clients, and this can be done,

10:58 → 11:03

let's say, in a distance of two or three weeks with different people.

11:03 → 11:05

I think this is a great benefit.

11:05 → 11:07

Okay then.

11:10 → 11:13

Maze software offers advanced analytics and insights into

11:13 → 11:14

user behavior.

11:14 → 11:18

Can you share some examples of particular, let's say,

11:18 → 11:22

activities that you discover using this tool?

11:22 → 11:25

Testing which does not really require prototypes.

11:25 → 11:27

You can test information architecture, for example.

11:27 → 11:32

I did that when we were redesigning our own Merix page.

11:32 → 11:36

So we've had people card sorts,

11:36 → 11:40

so do an activity where they associate categories

11:40 → 11:45

between and the cards of service names

11:45 → 11:47

to different categories.

11:47 → 11:51

And in that way, we try to see what's the the level

11:51 → 11:53

of understanding of those services.

11:53 → 11:57

So that helped to define the the

11:57 → 11:59

order of items on the page.

11:59 → 12:02

It helped to define the navigation menu.

12:02 → 12:05

And that led to also the the changes that are as simple as

12:05 → 12:08

copy changes because some labels were just not very

12:08 → 12:12

distinguishable and are not being understood well by the participants.

12:12 → 12:16

So even without building, a full blown prototype,

12:16 → 12:20

you can use the building blocks of Maze to your testing

12:20 → 12:23

on on an even smaller scale.

12:24 → 12:25

Yeah.

12:26 → 12:26

Yeah.

12:26 → 12:31

And I also had a chance to to test some time ago

12:31 → 12:34

one of the prototype of the of the product that I was

12:34 → 12:37

responsible for for the design.

12:37 → 12:39

And and yeah.

12:39 → 12:42

Especially in the reports from testing,

12:42 → 12:46

I was focused on the heat map and video recordings.

12:46 → 12:49

And thanks to that, I was able to, for example,

12:49 → 12:53

identify the place where user, for example,

12:53 → 12:55

tried to click on element,

12:55 → 12:59

but it was never applicable.

12:59 → 13:00

So,

13:01 → 13:02

yeah, I you know,

13:02 → 13:05

I was able to identify the places or elements in the

13:05 → 13:10

interface that are maybe are not designed properly or are

13:10 → 13:14

not intuitive as I thought they could be.

13:14 → 13:15

So,

13:16 → 13:21

yeah, that's that's really helpful in terms of analyzing

13:21 → 13:22

and, you know,

13:24 → 13:26

adjusting your designs and

13:27 → 13:31

add some improvements to your to your designs.

13:31 → 13:32

Yeah.

13:32 → 13:35

It's a great way to kind of test your thinking in practice.

13:35 → 13:39

If you are stuck with a solution, if you are not sure,

13:39 → 13:40

everyone will understand it.

13:40 → 13:43

Or sometimes if you're overly sure,

13:43 → 13:45

everyone knows what's happening.

13:45 → 13:47

A test, it can help, you know,

13:47 → 13:51

uncover any problems without even building the code for us.

13:51 → 13:53

So just do it early.

13:56 → 13:59

Okay. Thank you for this.

13:59 → 14:00

And what about the integrations?

14:00 → 14:02

You mentioned something about Figma,

14:02 → 14:05

but what other integrations are available,

14:05 → 14:10

and how does the integration with Figma actually work like?

14:12 → 14:14

Well, so

14:14 → 14:18

especially, I had a chance to, in most of the cases,

14:18 → 14:22

use Figma integration just because I was testing

14:22 → 14:24

the, you know, the prototypes.

14:27 → 14:30

And well, in fact, it works quite well.

14:30 → 14:33

I mean, the thing my integration works quite smooth,

14:33 → 14:36

and it's quite intuitive, let's say,

14:36 → 14:39

to Implement your prototype

14:39 → 14:41

directly into Maze.

14:41 → 14:44

However, I had last time,

14:44 → 14:49

when I was preparing the Maze test for user research and

14:49 → 14:51

For prototype testing,

14:53 → 14:56

I had a I had to say some problems

14:56 → 14:59

with the with this with this integration Okay.

14:59 → 15:04

Because I was creating prototype right after the

15:04 → 15:09

latest Figma releases about advanced prototyping and variables.

15:09 → 15:13

And one thing that I didn't notice

15:14 → 15:20

that time was that Maze does not support

15:21 → 15:24

variables and advanced prototype Right.

15:24 → 15:25

Prototype features.

15:25 → 15:29

So if you want to, for example, in Figma prototype,

15:29 → 15:31

if you want to avoid those, you know,

15:31 → 15:34

a lot of connections between frames and

15:34 → 15:36

create, let's say,

15:37 → 15:41

let's say, you're prototyping more advanced way and more clear way

15:41 → 15:45

without all of those connections using variables, for example.

15:47 → 15:52

It's it it won't work with Maze just because

15:52 → 15:56

they Maze is not able to identify

15:56 → 15:58

the changes from the one frame.

15:58 → 16:01

So when you are building the prototype for Maze,

16:01 → 16:03

it's the top totally better.

16:03 → 16:07

It's definitely better to build them frame by frame,

16:07 → 16:11

step by step instead of using, for instance, variables.

16:11 → 16:15

As a result, after this testing, I get

16:16 → 16:17

honestly nothing.

16:17 → 16:22

I mean, I get the one only one recording with for one frame,

16:22 → 16:25

and nothing was happened during all time of the,

16:25 → 16:28

you know, of the of the user tests.

16:28 → 16:29

So so yeah.

16:29 → 16:33

And to take care on that when you will be creating the

16:33 → 16:36

prototype for your mice testing.

16:37 → 16:37

Yeah.

16:37 → 16:42

So so now you have a heads up for what to do. Right?

16:42 → 16:45

But I think the integrations, they may progress.

16:45 → 16:48

You know, Figma is, like, moving on the fast speed,

16:48 → 16:51

and they are making those adjustments really,

16:51 → 16:54

really quickly to our workflow as designers.

16:54 → 16:58

So I think the other software just needs to catch up at some

16:58 → 17:00

point, and I I think that will make

17:00 → 17:04

everyone's life easier in this regard.

17:04 → 17:06

But there are still some good features.

17:06 → 17:08

You can also have, like, a Slack integration,

17:08 → 17:12

and that notifies you when someone is filling out the test

17:12 → 17:15

and, you know, a result is ready.

17:15 → 17:18

So that that helps too if if you're looking very or very

17:18 → 17:20

curious about the results.

17:20 → 17:23

So, yeah, I think a lot of companies are using Slack.

17:23 → 17:26

We definitely use it as a communication channel.

17:26 → 17:30

So that that is something that that can help improve your workflow too.

17:32 → 17:32

Alright.

17:32 → 17:36

It sounds to me like what is

17:38 → 17:42

that there is a thing that we could call

17:42 → 17:45

Figma oriented working environment.

17:45 → 17:46

Like,

17:47 → 17:50

let's say a stakeholder is coming to us,

17:50 → 17:54

and this person doesn't need to be necessarily, let's say,

17:54 → 17:59

tech savvy, but we are able from almost day one to present, let's say,

17:59 → 18:03

some prototyping work in order to verify some assumptions,

18:03 → 18:07

Then we can integrate the Figma prototype that we are building, let's say,

18:07 → 18:10

of a web application or mobile application with Maze in order

18:10 → 18:13

to conduct the

18:14 → 18:19

conducted research with the with the potential groups of users.

18:19 → 18:23

And very dynamically, the same Figma can be used

18:23 → 18:26

in order to gather feedback also from the stakeholder or

18:26 → 18:29

stakeholders by leaving, for example, comments.

18:29 → 18:32

And, also, almost at the same time, it can be directed,

18:32 → 18:35

pointed to development team.

18:35 → 18:35

Yeah?

18:35 → 18:39

So so so the front end or mobile people who will be

18:39 → 18:41

working on this prototype.

18:41 → 18:41

Yeah?

18:41 → 18:46

So Figma as an ultimate source of truth solving all our problems.

18:46 → 18:48

Yeah. I think it's an industry standard.

18:48 → 18:51

Everyone appreciates it, I think.

18:51 → 18:54

We've No more email attachments.

18:54 → 18:57

No more email attachments. The collaboration is good always.

18:57 → 19:00

In May, you also have ways like, interact with people.

19:00 → 19:04

They can leave comments when you're building a scenario or task.

19:04 → 19:07

They can leave their, maybe,

19:07 → 19:10

feedback on what else should be tested or, you know,

19:10 → 19:12

that they don't want this element included.

19:12 → 19:15

So the collaboration element is there as well.

19:15 → 19:17

Everyone appreciates this.

19:18 → 19:20

Alright then.

19:20 → 19:21

So

19:21 → 19:26

Maze enables to collect qualitative feedback alongside

19:26 → 19:28

quantitative data.

19:28 → 19:30

What are the features within the Maze that

19:30 → 19:34

facilitate this process, like efficient gathering

19:34 → 19:36

of that data?

19:36 → 19:40

Yeah. What which features help you doing that?

19:41 → 19:46

So during the creation of the whole scenario of the

19:46 → 19:49

user tests, user research in Maze,

19:49 → 19:54

you can create it very intuitive intuitive way and

19:54 → 19:57

very easily, like, step by step process.

19:57 → 20:02

So first, you need to do this and then to do another task and so on.

20:02 → 20:05

So and on each step,

20:05 → 20:08

you can add different elements,

20:08 → 20:10

and it could be feedback form.

20:10 → 20:12

It could be rating scale.

20:12 → 20:18

So, for example, you can ask user for feedback.

20:19 → 20:23

So user are able to provide qualitative

20:23 → 20:26

insights and comments during the test session.

20:26 → 20:30

Or, for example, you can in implement and add to your

20:30 → 20:32

scenario rating scale.

20:32 → 20:36

So you can also measure, let's say,

20:36 → 20:39

take feedback on kind of, let's say,

20:39 → 20:41

user preferences or Right.

20:41 → 20:43

Level, something like that.

20:43 → 20:46

You also have, like, more

20:46 → 20:49

specialized weights that the results are displayed.

20:49 → 20:52

I've mentioned the card sorting test that I've done.

20:52 → 20:56

So for this, you will get results in the typical way that

20:56 → 20:59

you get or should create the results for,

20:59 → 21:00

card sorting.

21:00 → 21:04

So you'll get get a graph with associations of how many cards

21:04 → 21:06

landed in the same categories.

21:06 → 21:09

So it does a lot of work to for you. Right?

21:09 → 21:13

Typically, you do you'd either use an another type of card sorting

21:13 → 21:18

software, and you have to manually put it in an Excel and then count.

21:18 → 21:20

That definitely helps to aggregate the results,

21:20 → 21:23

and and you don't have to almost spend time.

21:23 → 21:26

Like, most most of the features are in build.

21:26 → 21:30

You don't have to spend time on reporting for that long.

21:32 → 21:36

Okay then. And I think that we have the very last question.

21:38 → 21:41

In the context of AI capabilities integrated into

21:41 → 21:43

May software, what opportunities and challenges do

21:43 → 21:46

you foresee of leveraging

21:46 → 21:49

AI driven tools to enhance

21:49 → 21:50

testing process?

21:54 → 21:55

Well,

21:57 → 21:59

if we are thinking about AI and

21:59 → 22:03

Integration of Maze, first, I think

22:04 → 22:08

the first thing that I I I need to say is that the Maze,

22:08 → 22:10

of course, is a great tool

22:10 → 22:15

But never replace, for some traditional methods of

22:15 → 22:16

for some moderated testing.

22:16 → 22:16

Right?

22:16 → 22:17

Correct.

22:17 → 22:20

Unless AI replaced us as

22:20 → 22:22

a moderators.

22:22 → 22:24

Don't say that.

22:24 → 22:25

Yeah.

22:25 → 22:27

So if

22:28 → 22:32

we want to think about the future and AI integration

22:32 → 22:36

of Mace, it's, of course, the automation.

22:36 → 22:39

So AI could automate some repeated

22:39 → 22:44

tasks, could help us in data analysis,

22:44 → 22:48

and you know, aggregate all of the

22:48 → 22:53

results of the reports and trans transform them

22:53 → 22:57

into some actionable items or

22:58 → 23:00

or or decisions.

23:02 → 23:02

Yeah.

23:02 → 23:07

I think that's that's that's all from my side.

23:07 → 23:10

I would say we treat it as another pencil in our pencil case.

23:10 → 23:13

It's not a replacement. It's an assistant.

23:13 → 23:16

So aggregating results, yes, they can do it.

23:16 → 23:19

You know, the the algorithms,

23:19 → 23:24

they can tell you to rephrase certain questions not to be leading.

23:24 → 23:26

So it is helpful.

23:26 → 23:28

It has it has this, like,

23:28 → 23:29

another eye.

23:29 → 23:33

So typically, it can replace if you're a single UX designer

23:33 → 23:34

for the company.

23:34 → 23:38

It can be that second pair of eyes in a way for you.

23:38 → 23:41

So that that's how you should probably treat it.

23:41 → 23:45

It can probably do some,

23:45 → 23:46

more advanced desk research.

23:46 → 23:48

So on the results, I think in the future,

23:48 → 23:49

we'll be able to compare.

23:49 → 23:51

Is it a trend?

23:51 → 23:53

Has this changed over time? Right?

23:53 → 23:55

So do the manual work that you'd have to do it,

23:55 → 23:58

yourself or maybe compare it with, like, I don't know,

23:58 → 24:01

overall worldwide statistics.

24:01 → 24:05

Like, if if people didn't click on that link, typically,

24:05 → 24:08

it's better than, you know, thirty other percent of people,

24:08 → 24:10

you know, in that region.

24:10 → 24:13

So maybe maybe those comparisons can be made as well.

24:13 → 24:17

But I don't think it has the or it will not have in the next

24:17 → 24:20

upcoming years the cognitive capabilities.

24:20 → 24:24

You have to have the human skill, right, to to do it.

24:24 → 24:27

You you still have to be human, and we as designers,

24:27 → 24:30

we have backgrounds in different things.

24:30 → 24:34

So cognitive knowledge, we have we have the psychological knowledge.

24:34 → 24:37

Some of us have, you know, background in sociology.

24:37 → 24:40

So all that knowledge is the human part.

24:40 → 24:43

So I don't think we're going anywhere anytime soon.

24:44 → 24:47

Humans know how to human humans.

24:47 → 24:48

Correct.

24:48 → 24:52

Ladies and gentlemen, in the last two or three years,

24:52 → 24:54

also with Oscar and Hannah,

24:54 → 24:57

I had the pleasure to work on several projects which were

24:57 → 25:01

dedicated to fleet managers based in Asia and Africa,

25:01 → 25:05

sales field representatives selling

25:05 → 25:09

photovoltaics in North America, people representing

25:09 → 25:12

ages between thirty to fifty,

25:12 → 25:15

people who had very deep educational background,

25:15 → 25:19

but also people who have been just graduating schools

25:20 → 25:23

in, for example, central or west Western Europe.

25:23 → 25:27

People digital natives, but also people who remember

25:27 → 25:29

Commodore sixty fours. Yes?

25:29 → 25:33

So in order to build products that can be widely

25:33 → 25:36

used by different specific groups of users,

25:36 → 25:41

we consider as a very valuable thing to perform test research.

25:41 → 25:45

This can help you save a lot of assets,

25:45 → 25:49

save a lot of money during the development phase,

25:49 → 25:51

which according to our manager, Maria Borove,

25:51 → 25:56

can one dollar spend on design and

25:56 → 26:00

research can save ten dollars or euros or pounds

26:00 → 26:03

or Polish water spend on development.

26:03 → 26:05

So please keep this in mind.

26:05 → 26:08

And, guys, any ending words from your side?

26:10 → 26:14

Just go and test. It's easier than it ever was. Just do it.

26:14 → 26:16

Yes.

26:16 → 26:19

Yes. Testing is good.

26:19 → 26:22

Getting to know your users is good.

26:22 → 26:25

Building software that is beautiful and useless is not good.

26:25 → 26:28

So, ladies and gentlemen, thank you all.

26:28 → 26:29

Have a great day.

26:29 → 26:32

This was another show coming from Merixstudio.

26:32 → 26:33

It's time to talk.

26:33 → 26:36

Please stay tuned for further episodes. Thank you.

26:36 → 26:38

Have a great day.

26:38 → 26:38

Thank you.

26:38 → 26:40

Thank you very much. See you.