1
00:00:00,270 --> 00:00:07,650
the okay now all looks good thank

2
00:00:07,650 --> 00:00:10,410
you for joining us today frederick the

3
00:00:10,470 --> 00:00:14,640
floor is yours thanks so much everyone

4
00:00:14,640 --> 00:00:16,140
for being here my name is phil

5
00:00:16,530 --> 00:00:19,290
and i'm going to talk about the

6
00:00:19,350 --> 00:00:20,160
matter among convert

7
00:00:20,160 --> 00:00:22,590
option route to musician to date

8
00:00:24,270 --> 00:00:26,970
just to as a reminder you are

9
00:00:27,090 --> 00:00:30,510
able to tune the chatroom onto a

10
00:00:30,510 --> 00:00:33,120
metal wall comes to the department site

11
00:00:33,660 --> 00:00:36,690
just go into the into the talk

12
00:00:37,020 --> 00:00:38,910
card and new ability to join to

13
00:00:38,910 --> 00:00:39,210
there

14
00:00:40,350 --> 00:00:42,810
the chest to two if you if

15
00:00:42,810 --> 00:00:44,160
you ever got some questions i can

16
00:00:44,160 --> 00:00:44,580
answered

17
00:00:46,800 --> 00:00:48,420
okay so here's what we're going to

18
00:00:48,420 --> 00:00:49,770
talk about today first of all i'm

19
00:00:49,770 --> 00:00:51,930
going to introduce myself then we're going

20
00:00:51,930 --> 00:00:54,330
to talk about the prerequisites for me

21
00:00:54,390 --> 00:00:57,390
condition rated tradition with metal then we

22
00:00:57,390 --> 00:00:58,980
will talk about how you can find

23
00:00:58,980 --> 00:01:02,130
your first optimization high to use then

24
00:01:02,130 --> 00:01:04,769
how you can prioritize these harnesses using

25
00:01:05,010 --> 00:01:05,670
some c

26
00:01:05,760 --> 00:01:08,700
part of frameworks and how you can

27
00:01:08,700 --> 00:01:11,160
use matter more buicks analytics features in

28
00:01:11,160 --> 00:01:14,854
the process how to write your testing

29
00:01:14,854 --> 00:01:16,260
that both of his and how to

30
00:01:16,260 --> 00:01:18,090
make your head and he tests using

31
00:01:18,180 --> 00:01:20,100
maternal system features

32
00:01:23,010 --> 00:01:27,480
so with digital analytics and crm consultant

33
00:01:28,620 --> 00:01:31,890
i've been working with several companies and

34
00:01:31,890 --> 00:01:34,020
several organization independent settle

35
00:01:35,250 --> 00:01:38,225
i i am specializing developers open source

36
00:01:38,225 --> 00:01:40,170
for the is for about two years

37
00:01:40,170 --> 00:01:40,560
now

38
00:01:41,460 --> 00:01:45,690
and i also was a peter the

39
00:01:45,694 --> 00:01:50,910
technical director about from of digital marketing

40
00:01:50,910 --> 00:01:54,186
master so i'm quite used to teaching

41
00:01:54,200 --> 00:01:55,650
and building can

42
00:01:57,690 --> 00:02:00,330
teaching programs here you can find lining

43
00:02:00,330 --> 00:02:03,270
up being tiny if you are interested

44
00:02:04,470 --> 00:02:08,400
okay so what do we need to

45
00:02:09,000 --> 00:02:11,610
to do conversion rate optimization with matin

46
00:02:12,960 --> 00:02:14,640
the first thing we need to do

47
00:02:14,670 --> 00:02:17,610
is to define what performance is so

48
00:02:17,610 --> 00:02:20,400
i have a definition here her performance

49
00:02:20,460 --> 00:02:23,760
expresses the level of achievement puppy object

50
00:02:23,760 --> 00:02:24,540
is person

51
00:02:26,250 --> 00:02:30,750
so here's really something important to the

52
00:02:30,750 --> 00:02:32,610
people to to know what we're talking

53
00:02:32,610 --> 00:02:36,030
about because we are talking about something

54
00:02:36,030 --> 00:02:37,950
which is performances it's really important to

55
00:02:37,950 --> 00:02:40,410
know what we are meaning would be

56
00:02:40,410 --> 00:02:41,280
safer from us

57
00:02:42,510 --> 00:02:46,020
so now you need to define the

58
00:02:46,080 --> 00:02:48,120
objectives which means the business goal

59
00:02:49,770 --> 00:02:51,270
first of all going to to give

60
00:02:51,270 --> 00:02:53,490
you an example you may want to

61
00:02:53,490 --> 00:02:55,050
increase the part of people whom you

62
00:02:55,052 --> 00:02:57,780
purchase one of your products sharp and

63
00:02:57,840 --> 00:03:01,680
commerce websites and then you need to

64
00:03:01,680 --> 00:03:03,900
define the level of achievement which means

65
00:03:03,930 --> 00:03:06,390
the indicators the trestles and the seconds

66
00:03:06,750 --> 00:03:08,790
so for instance for the ecommerce

67
00:03:09,390 --> 00:03:11,640
conversion rate the pain de coteau gear

68
00:03:11,700 --> 00:03:14,367
ecommerce conversion rate we could decide that

69
00:03:14,367 --> 00:03:16,530
this result is two persons and we

70
00:03:16,530 --> 00:03:19,050
could decide that distressful is for all

71
00:03:19,080 --> 00:03:23,100
users and to keep going through this

72
00:03:23,370 --> 00:03:26,970
process we will use her two main

73
00:03:26,970 --> 00:03:28,710
talks the first one is the kpi-jci

74
00:03:28,710 --> 00:03:29,340
from work

75
00:03:29,850 --> 00:03:31,230
and the second one was the miserable

76
00:03:31,230 --> 00:03:32,460
clan and the tabletten

77
00:03:34,440 --> 00:03:37,800
when you want to do optimization on

78
00:03:37,890 --> 00:03:40,440
and then it seeks to slight matter

79
00:03:40,440 --> 00:03:42,090
on you you'll need to have

80
00:03:44,130 --> 00:03:47,190
as many data as possible to to

81
00:03:47,190 --> 00:03:49,440
have a granted analysis so it's really

82
00:03:49,440 --> 00:03:52,350
important to have a proper tax planning

83
00:03:52,410 --> 00:03:53,670
proper keep you out from

84
00:03:55,410 --> 00:03:58,530
so what is atp iii framework

85
00:04:00,120 --> 00:04:04,950
basically it's a document that explains how

86
00:04:04,950 --> 00:04:08,520
you're going to measure performers through different

87
00:04:08,520 --> 00:04:11,370
level of engagement from the user point

88
00:04:11,370 --> 00:04:11,610
of view

89
00:04:12,720 --> 00:04:14,580
hope you start you always start with

90
00:04:14,580 --> 00:04:18,089
your business goal and in the scope

91
00:04:18,089 --> 00:04:21,149
of reach the first level of engagement

92
00:04:21,450 --> 00:04:23,160
you want to increase the visibility of

93
00:04:23,160 --> 00:04:26,610
your sights on the level engagement you

94
00:04:26,610 --> 00:04:28,830
want to increase the number of qualified

95
00:04:28,830 --> 00:04:31,890
users on the weapon the platforms you

96
00:04:31,890 --> 00:04:32,790
may also want to

97
00:04:32,880 --> 00:04:35,280
engage the audience and you may want

98
00:04:35,280 --> 00:04:37,560
to divert the visitor base within the

99
00:04:37,560 --> 00:04:39,780
digital ecosystem if you have several websites

100
00:04:39,780 --> 00:04:40,230
or apps

101
00:04:41,310 --> 00:04:43,410
on the conversion level you want to

102
00:04:43,410 --> 00:04:45,900
facilitate to unlock the realization of actions

103
00:04:45,900 --> 00:04:48,330
having a direct impact on the business

104
00:04:49,020 --> 00:04:51,240
and on the advocacy that you may

105
00:04:51,240 --> 00:04:53,190
want to maintain and develop the relationship

106
00:04:53,280 --> 00:04:55,500
create the site's audience increase the number

107
00:04:55,590 --> 00:04:57,270
of contests in this jar empty up

108
00:04:57,420 --> 00:04:59,850
visitors on social networks and then the

109
00:04:59,910 --> 00:05:01,440
interactions on the community

110
00:05:01,440 --> 00:05:05,280
pity her on social networks so for

111
00:05:05,310 --> 00:05:07,378
each of these business goal posts you

112
00:05:07,378 --> 00:05:10,170
will honor the gpl level which means

113
00:05:10,260 --> 00:05:14,250
macro indicators you have you define certain

114
00:05:14,250 --> 00:05:16,620
indicators so for instance for reach and

115
00:05:16,620 --> 00:05:19,260
increase the visibility of websites you want

116
00:05:19,260 --> 00:05:20,970
to increase traffic volume

117
00:05:21,660 --> 00:05:22,980
then you have the description of the

118
00:05:22,980 --> 00:05:25,320
metric you have the matter moon at

119
00:05:25,320 --> 00:05:27,773
which you have the threshold and you

120
00:05:27,780 --> 00:05:28,740
have the source

121
00:05:29,820 --> 00:05:31,740
so here we are going to take

122
00:05:31,740 --> 00:05:33,930
a look at this part here but

123
00:05:33,930 --> 00:05:37,080
it's really it's really important to have

124
00:05:37,800 --> 00:05:41,490
all the levels of customer engagements so

125
00:05:41,490 --> 00:05:43,980
that you know if you have your

126
00:05:43,980 --> 00:05:47,220
preferences is good or nuts for instance

127
00:05:47,220 --> 00:05:49,890
on a monthly basis and surely

128
00:05:49,950 --> 00:05:52,500
also you can see that sometimes you

129
00:05:52,500 --> 00:05:54,210
have a matter more as the main

130
00:05:54,210 --> 00:05:56,340
source of data and you'll also sometimes

131
00:05:56,356 --> 00:05:59,370
have external tools such as as your

132
00:05:59,370 --> 00:06:01,560
tools and sometimes social media

133
00:06:04,860 --> 00:06:08,730
on this document there are some things

134
00:06:08,790 --> 00:06:12,090
that are missing but i just wanted

135
00:06:12,090 --> 00:06:13,410
to show you a big picture of

136
00:06:13,410 --> 00:06:15,510
it what is missing is the segment

137
00:06:15,900 --> 00:06:18,000
and what is missing is also the

138
00:06:18,030 --> 00:06:20,280
calculation of the indicator

139
00:06:23,670 --> 00:06:26,820
what you need after that is to

140
00:06:26,910 --> 00:06:29,250
the measurement plan and the measurement plan

141
00:06:29,250 --> 00:06:31,260
is to document where you explain

142
00:06:32,640 --> 00:06:34,920
what are all the actions you're going

143
00:06:34,920 --> 00:06:38,070
to track with metamorph mainly using

144
00:06:39,120 --> 00:06:41,692
even tax so for instance here you

145
00:06:41,700 --> 00:06:44,130
have the page template and all the

146
00:06:44,130 --> 00:06:46,590
actions that the users can do on

147
00:06:46,710 --> 00:06:49,920
the specific pitch templates the type of

148
00:06:50,010 --> 00:06:52,290
attack he sure you're going to use

149
00:06:53,010 --> 00:06:54,270
in order to move and then the

150
00:06:54,270 --> 00:06:57,270
pocketed tour of your event category action

151
00:06:57,750 --> 00:06:58,290
and name

152
00:06:59,490 --> 00:07:02,280
says this is a really important also

153
00:07:02,280 --> 00:07:05,220
because we are talking about kitty eyes

154
00:07:05,220 --> 00:07:07,138
me which means we're talking about indicators

155
00:07:07,138 --> 00:07:09,030
that we will give to the top

156
00:07:09,030 --> 00:07:11,340
management but sometimes you are also

157
00:07:12,600 --> 00:07:14,520
marketing teams that want to

158
00:07:15,720 --> 00:07:19,230
they increase their level of engagement and

159
00:07:19,235 --> 00:07:22,170
their level of some events some realization

160
00:07:22,170 --> 00:07:24,360
of events so it's important that list

161
00:07:24,360 --> 00:07:26,580
everything here in that you can have

162
00:07:26,580 --> 00:07:28,530
that in your matomo instance

163
00:07:31,410 --> 00:07:33,151
and once you have your permission and

164
00:07:33,151 --> 00:07:35,610
plan then you can go on and

165
00:07:36,030 --> 00:07:38,760
write down your tagging plan and here

166
00:07:39,240 --> 00:07:41,283
this is the technical part where where

167
00:07:41,370 --> 00:07:44,040
you tell her the dev teams what

168
00:07:44,040 --> 00:07:46,320
type of coach to push on the

169
00:07:46,320 --> 00:07:48,870
page so here we have the data

170
00:07:48,875 --> 00:07:49,920
layer for the

171
00:07:50,520 --> 00:07:52,860
events and here you can see the

172
00:07:52,860 --> 00:07:54,780
tracking code and all that the different

173
00:07:54,780 --> 00:07:57,210
town veterans that are pushed to them

174
00:07:58,230 --> 00:07:59,580
to load the data there

175
00:08:01,680 --> 00:08:04,110
and then last but not least you

176
00:08:04,110 --> 00:08:07,290
also have to add otho that different

177
00:08:07,290 --> 00:08:09,360
values for you we will be using

178
00:08:09,360 --> 00:08:11,400
as some custom dimensions

179
00:08:12,630 --> 00:08:15,030
this is really important also to to

180
00:08:15,030 --> 00:08:17,670
add a really granular analysis and two

181
00:08:17,670 --> 00:08:21,420
people to segment all your optimizations

182
00:08:24,810 --> 00:08:27,840
okay now you have sure katie i

183
00:08:27,840 --> 00:08:29,670
from work you have your tagging plan

184
00:08:30,720 --> 00:08:33,480
you can start with mason so we

185
00:08:33,630 --> 00:08:36,900
defined what performance is not what we

186
00:08:36,900 --> 00:08:39,480
are going to do is define what

187
00:08:39,480 --> 00:08:43,289
conversion rate optimization is the name of

188
00:08:43,289 --> 00:08:45,510
what we are doing you are probably

189
00:08:45,510 --> 00:08:50,130
new search engine optimization all kind of

190
00:08:51,420 --> 00:08:55,350
stuff doing optimization but the it's it's

191
00:08:55,350 --> 00:08:58,050
quite specific here because cr one is

192
00:08:58,050 --> 00:09:00,090
sir i have a definitional so it's

193
00:09:00,090 --> 00:09:02,370
the practice of increasing the percentage of

194
00:09:02,370 --> 00:09:04,590
user who perform the desired action on

195
00:09:04,590 --> 00:09:06,780
the website but i have a problem

196
00:09:06,780 --> 00:09:11,010
with this definition because acumen example below

197
00:09:11,880 --> 00:09:14,970
if you increase have if you suddenly

198
00:09:15,210 --> 00:09:17,160
drops of salt if you drop the

199
00:09:17,160 --> 00:09:20,940
prices by power same ninety percent you

200
00:09:20,940 --> 00:09:24,750
will have a really good impact positive

201
00:09:24,750 --> 00:09:27,540
impact on your conversion rates but that's

202
00:09:27,540 --> 00:09:30,390
not what you want because only increasing

203
00:09:30,390 --> 00:09:31,200
the conversion rate

204
00:09:31,290 --> 00:09:32,280
is is not enough

205
00:09:33,441 --> 00:09:35,820
i think we should talk about conversion

206
00:09:35,850 --> 00:09:39,810
volume and value optimization in not only

207
00:09:39,810 --> 00:09:42,060
conversion rate optimization

208
00:09:45,330 --> 00:09:45,840
okay

209
00:09:47,220 --> 00:09:49,110
now we have defined what we will

210
00:09:49,110 --> 00:09:51,750
be doing help how can we find

211
00:09:51,750 --> 00:09:54,930
our first optimization ideas i took the

212
00:09:54,960 --> 00:09:59,040
example of an ecommerce website because this

213
00:09:59,040 --> 00:10:01,770
is really an example where we're talking

214
00:10:01,770 --> 00:10:05,190
about money so performance here is already

215
00:10:05,400 --> 00:10:06,360
simple to understand

216
00:10:07,692 --> 00:10:10,145
when you have your indicators you can

217
00:10:10,145 --> 00:10:13,505
compare the metrics between the different segment

218
00:10:13,565 --> 00:10:16,415
and the audiences for for instance we

219
00:10:16,415 --> 00:10:20,285
decided that our average conversion rate should

220
00:10:20,285 --> 00:10:22,115
be that say two percent

221
00:10:23,225 --> 00:10:25,715
and when we compare the conversion rates

222
00:10:26,315 --> 00:10:30,785
by mobile desktop tablet we see that

223
00:10:30,785 --> 00:10:34,295
there are huge differences so now we

224
00:10:34,295 --> 00:10:38,795
should ask ourselves questions which will be

225
00:10:38,825 --> 00:10:42,515
why is there such differences between devices

226
00:10:43,595 --> 00:10:45,335
being in the real world we knew

227
00:10:45,335 --> 00:10:46,925
that most of the time

228
00:10:47,947 --> 00:10:48,665
people

229
00:10:49,895 --> 00:10:52,925
made their first visits with a mobile

230
00:10:52,925 --> 00:10:55,925
device and then finished the audio on

231
00:10:55,985 --> 00:10:59,195
a laptop or desktop but cheerful the

232
00:10:59,195 --> 00:11:02,105
example we will assume that people start

233
00:11:02,105 --> 00:11:05,015
their session that visits on the device

234
00:11:05,075 --> 00:11:08,165
and finish or don't finish to the

235
00:11:08,615 --> 00:11:09,155
purchase

236
00:11:10,055 --> 00:11:12,065
process so here we see a huge

237
00:11:12,065 --> 00:11:14,615
difference between desktop and mobile

238
00:11:16,235 --> 00:11:18,395
what should we do we should go

239
00:11:18,485 --> 00:11:21,215
and looking to the e commerce says

240
00:11:21,215 --> 00:11:23,375
for them to see if we a

241
00:11:23,375 --> 00:11:27,185
freaking see any difference in the checkout

242
00:11:27,185 --> 00:11:29,615
steps between the different devices

243
00:11:30,965 --> 00:11:34,505
so what you're going to do is

244
00:11:34,595 --> 00:11:36,245
you are going to go in the

245
00:11:36,265 --> 00:11:39,245
funnel records and she got the different

246
00:11:39,245 --> 00:11:42,185
steps so here is her potent application

247
00:11:42,365 --> 00:11:45,935
sign up payment and conversion

248
00:11:47,015 --> 00:11:48,785
and what you want to do here

249
00:11:48,785 --> 00:11:50,975
is to use segments moving matter more

250
00:11:51,305 --> 00:11:53,165
and see if you see any difference

251
00:11:53,315 --> 00:11:56,405
and if you see any difference then

252
00:11:56,495 --> 00:12:00,335
you may want to try to optimize

253
00:12:05,045 --> 00:12:06,995
if you see a huge dude smooth

254
00:12:07,025 --> 00:12:11,555
for instance at the payment information step

255
00:12:12,365 --> 00:12:14,885
let's say for mobile users you may

256
00:12:14,885 --> 00:12:18,695
want to investigate further but this is

257
00:12:18,695 --> 00:12:20,885
not what we're going to do now

258
00:12:21,365 --> 00:12:21,545
bitch

259
00:12:21,545 --> 00:12:23,915
because we're going to see that step

260
00:12:24,005 --> 00:12:26,465
four how to use matter what your

261
00:12:26,465 --> 00:12:29,645
accent features not we're going to to

262
00:12:29,645 --> 00:12:33,695
continue and talk about conversion rate optimization

263
00:12:33,755 --> 00:12:34,475
frameworks

264
00:12:35,735 --> 00:12:38,645
because we need to structure our approach

265
00:12:39,515 --> 00:12:41,525
to ego transitioning we could do what

266
00:12:41,525 --> 00:12:44,525
we call your mystic analysis

267
00:12:45,905 --> 00:12:48,155
just to have two words about what's

268
00:12:48,185 --> 00:12:50,615
yours to candidates is is just you

269
00:12:50,645 --> 00:12:53,345
go through your websites through your congressional

270
00:12:53,345 --> 00:12:56,225
journey and used you take notes on

271
00:12:56,285 --> 00:12:59,135
what's wrong from your point of view

272
00:13:00,545 --> 00:13:02,555
might say on the product page i

273
00:13:02,555 --> 00:13:06,125
find that pictures are too small i

274
00:13:06,155 --> 00:13:08,825
think that the etiquette person is not

275
00:13:08,825 --> 00:13:12,515
printing it enough it's below the fold

276
00:13:12,605 --> 00:13:15,245
of stuff like that you can also

277
00:13:16,325 --> 00:13:18,935
to some kind of benchmark with competitors

278
00:13:19,205 --> 00:13:21,245
saying yes i'm on the product page

279
00:13:21,305 --> 00:13:23,165
me to stuff that you don't and

280
00:13:23,165 --> 00:13:25,655
i think this this could be a

281
00:13:25,655 --> 00:13:27,875
good idea but it's quite risky because

282
00:13:27,875 --> 00:13:30,545
once again this would be your own

283
00:13:30,545 --> 00:13:32,885
opinion and we have been marketing so

284
00:13:33,005 --> 00:13:36,065
our own opinions should should not wait

285
00:13:36,065 --> 00:13:36,425
that long

286
00:13:36,425 --> 00:13:38,495
such in our decision so that's the

287
00:13:38,495 --> 00:13:40,205
reason why we should rely on data

288
00:13:40,625 --> 00:13:42,125
and on the frameworks

289
00:13:44,585 --> 00:13:47,585
so i engage you i give you

290
00:13:47,585 --> 00:13:50,105
here i unnatural three or four temple

291
00:13:50,105 --> 00:13:53,315
of of of cr of are the

292
00:13:53,315 --> 00:13:54,905
first ones t i e

293
00:13:56,315 --> 00:13:58,115
which is a model done by a

294
00:13:58,175 --> 00:13:59,015
wide funnel

295
00:14:00,185 --> 00:14:03,215
in fact you have marks here for

296
00:14:03,785 --> 00:14:04,415
every

297
00:14:05,645 --> 00:14:09,335
lift organization zones and you have three

298
00:14:09,395 --> 00:14:12,665
items the potential of the change the

299
00:14:12,665 --> 00:14:16,205
importance of the top the changes in

300
00:14:16,205 --> 00:14:19,175
the ease of the changes and then

301
00:14:19,295 --> 00:14:21,635
we'll have scored in the high court

302
00:14:21,635 --> 00:14:22,025
miss

303
00:14:23,255 --> 00:14:26,525
the first the optimization should be done

304
00:14:29,675 --> 00:14:33,665
then you have another example of seattle

305
00:14:33,665 --> 00:14:37,355
framework which is the ice model which

306
00:14:37,355 --> 00:14:40,955
was created by sean ennis asher again

307
00:14:40,955 --> 00:14:44,135
you have three items the impacts what

308
00:14:44,135 --> 00:14:46,235
will be the impact if this works

309
00:14:46,625 --> 00:14:47,705
confidence how

310
00:14:47,735 --> 00:14:51,215
confident am i that this will work

311
00:14:51,215 --> 00:14:54,335
in peace what is the ease of

312
00:14:54,395 --> 00:14:55,685
implementation that

313
00:14:56,855 --> 00:14:59,345
this year i have a problem with

314
00:14:59,345 --> 00:15:02,525
it how confident am i since it's

315
00:15:02,525 --> 00:15:04,355
something that i'm going to propose

316
00:15:05,465 --> 00:15:09,875
i think that i will be confident

317
00:15:09,875 --> 00:15:12,455
in it that confidence is not something

318
00:15:13,895 --> 00:15:15,575
really scientific or so

319
00:15:16,985 --> 00:15:21,053
at this this exists so this is

320
00:15:21,145 --> 00:15:22,775
good to know that it exists but

321
00:15:22,835 --> 00:15:25,085
it's it's not something that i would

322
00:15:25,085 --> 00:15:26,525
recommend to to use

323
00:15:28,895 --> 00:15:31,685
you have also another model by done

324
00:15:31,685 --> 00:15:33,275
by had wire

325
00:15:34,535 --> 00:15:36,785
and here you have one points or

326
00:15:36,845 --> 00:15:40,025
zero points for each item so let's

327
00:15:40,025 --> 00:15:40,475
say

328
00:15:42,335 --> 00:15:45,125
this optimization a optimization idea will support

329
00:15:45,126 --> 00:15:46,415
the companion matrix

330
00:15:47,795 --> 00:15:49,805
yes or no one or zero

331
00:15:51,995 --> 00:15:54,905
are when the test has changed on

332
00:15:54,908 --> 00:15:57,635
the results opening pages just for notebook

333
00:15:57,649 --> 00:16:00,905
again on pause or is it a

334
00:16:00,905 --> 00:16:02,825
change above the photo below the four

335
00:16:03,665 --> 00:16:05,645
okay i'm looking to to to explain

336
00:16:05,645 --> 00:16:07,475
everything that you understand the principle of

337
00:16:08,105 --> 00:16:09,965
one or zero and the higher the

338
00:16:09,965 --> 00:16:10,805
note the quicker

339
00:16:11,574 --> 00:16:12,935
the tests should be run

340
00:16:14,975 --> 00:16:17,255
and the last one it's the p

341
00:16:17,255 --> 00:16:20,555
x l framework proposed by conversion xl

342
00:16:21,425 --> 00:16:24,335
and again it's a it's a a

343
00:16:24,335 --> 00:16:27,755
march imitation system with one zero or

344
00:16:27,755 --> 00:16:30,875
two and the higher the note that

345
00:16:31,835 --> 00:16:33,755
the most important to the test

346
00:16:35,585 --> 00:16:37,865
so what's you what you have to

347
00:16:37,865 --> 00:16:40,685
do what he wanted optimization is to

348
00:16:40,685 --> 00:16:43,385
go through the data go through other

349
00:16:43,385 --> 00:16:49,355
websites and look for huge differences in

350
00:16:49,355 --> 00:16:51,305
one you find once you find those

351
00:16:51,305 --> 00:16:54,395
huge differences then you have to investigate

352
00:16:54,425 --> 00:16:54,875
further

353
00:16:57,425 --> 00:16:59,975
and when you want to investigate further

354
00:16:59,975 --> 00:17:02,899
that's when the wax analytics features come

355
00:17:02,915 --> 00:17:03,635
to this

356
00:17:04,744 --> 00:17:10,115
i mean you saw that in the

357
00:17:10,115 --> 00:17:12,035
check out to payment info it's really

358
00:17:12,035 --> 00:17:15,785
lower for mobile users but you you

359
00:17:15,785 --> 00:17:18,395
don't know exactly what the issue is

360
00:17:18,395 --> 00:17:20,704
when you only look at statistical data

361
00:17:20,733 --> 00:17:24,335
with a metal so you may have

362
00:17:24,335 --> 00:17:24,875
an idea

363
00:17:24,905 --> 00:17:29,045
but you have to confirm that your

364
00:17:29,045 --> 00:17:31,085
aunt here is a ghost in the

365
00:17:31,085 --> 00:17:31,865
right direction

366
00:17:34,895 --> 00:17:37,475
so here is what you can use

367
00:17:37,475 --> 00:17:39,155
as features of pinnacle

368
00:17:40,325 --> 00:17:42,875
here it's it is examples taken from

369
00:17:43,265 --> 00:17:45,965
the material the more website

370
00:17:46,985 --> 00:17:49,685
i i think that you all are

371
00:17:49,685 --> 00:17:51,725
aware of what session recordings optum grip

372
00:17:51,755 --> 00:17:54,185
going to just re explain it just

373
00:17:54,185 --> 00:17:56,285
in case what it does is that

374
00:17:56,315 --> 00:17:58,835
it takes video snapshots of

375
00:18:00,635 --> 00:18:02,735
users turn on the website so it

376
00:18:02,825 --> 00:18:06,695
allows you to watch to watch playbacks

377
00:18:06,755 --> 00:18:09,935
video playbacks of users turns on the

378
00:18:09,935 --> 00:18:12,789
website so for instance on the on

379
00:18:12,791 --> 00:18:15,155
the example i took on the payment

380
00:18:15,155 --> 00:18:18,160
info you see that mobile users don't

381
00:18:18,160 --> 00:18:19,895
can go through entered

382
00:18:20,135 --> 00:18:22,895
don't finalize their purchase so you can

383
00:18:22,985 --> 00:18:24,515
ask madam what to record

384
00:18:25,805 --> 00:18:28,715
to curb this behavior and then too

385
00:18:28,745 --> 00:18:31,505
you can watch the playback just to

386
00:18:31,505 --> 00:18:34,535
confirm your your hypothesis that might be

387
00:18:35,105 --> 00:18:37,780
i don't know it's it's it's not

388
00:18:37,780 --> 00:18:39,875
to mobile friendly or i don't know

389
00:18:39,875 --> 00:18:42,215
what you you can watch the videos

390
00:18:42,215 --> 00:18:44,281
just to confirm the idea you have

391
00:18:44,281 --> 00:18:45,785
about why it's not working

392
00:18:46,175 --> 00:18:46,985
and as it should

393
00:18:49,295 --> 00:18:52,835
so he tests again a screenshot of

394
00:18:52,835 --> 00:18:56,255
what it does you have here the

395
00:18:56,255 --> 00:18:57,365
movement of the mouse

396
00:18:58,505 --> 00:19:00,005
on the on the on the page

397
00:19:02,855 --> 00:19:05,314
what you can use also as multiple

398
00:19:05,314 --> 00:19:06,304
features for

399
00:19:07,775 --> 00:19:10,835
just to confirm you on your idea

400
00:19:10,835 --> 00:19:13,805
is to use the it maps feature

401
00:19:14,765 --> 00:19:16,745
so imagine that scene you have this

402
00:19:17,225 --> 00:19:20,765
payment info page and you should click

403
00:19:20,765 --> 00:19:23,105
on the accept the terms of sansa

404
00:19:23,615 --> 00:19:26,585
and you see that people don't click

405
00:19:26,585 --> 00:19:27,545
on it because it's

406
00:19:27,581 --> 00:19:30,635
is not not that visible so here

407
00:19:30,695 --> 00:19:33,425
the the heatmaps will helps you to

408
00:19:33,425 --> 00:19:36,065
confirm the fact that people don't click

409
00:19:36,065 --> 00:19:38,435
on it because it's let's say it's

410
00:19:38,435 --> 00:19:40,745
below the fold it's not visible enough

411
00:19:41,195 --> 00:19:43,265
so that's the reason why people can

412
00:19:43,265 --> 00:19:46,535
go through the purchase process and maybe

413
00:19:46,925 --> 00:19:47,705
they don't see

414
00:19:47,739 --> 00:19:49,055
it and they can click on the

415
00:19:49,083 --> 00:19:51,275
next button so that's the explanation you

416
00:19:51,275 --> 00:19:54,815
have and it's confirmed by the by

417
00:19:54,815 --> 00:19:55,925
the it mats

418
00:20:00,005 --> 00:20:02,195
another interesting feature that you can use

419
00:20:03,425 --> 00:20:06,305
maxima is the form analytics feature

420
00:20:07,385 --> 00:20:10,265
especially when you're talking about

421
00:20:12,445 --> 00:20:15,265
chicken processes just to make sure that

422
00:20:15,295 --> 00:20:17,215
there is no be sure with some

423
00:20:17,335 --> 00:20:18,415
specific fields

424
00:20:20,035 --> 00:20:23,635
for instance you may see that some

425
00:20:23,635 --> 00:20:26,273
fields are left blank or there are

426
00:20:26,275 --> 00:20:29,905
some formats issues with some fields here

427
00:20:29,905 --> 00:20:33,377
it's it's interesting to to have also

428
00:20:33,685 --> 00:20:34,435
information

429
00:20:36,595 --> 00:20:39,775
on on the different horse and what

430
00:20:39,805 --> 00:20:41,395
type of issues can occur on the

431
00:20:41,395 --> 00:20:43,195
defeat to her phones

432
00:20:46,675 --> 00:20:50,515
and last but not least the user

433
00:20:50,515 --> 00:20:54,085
feedback plugin is quite new feature

434
00:20:55,315 --> 00:20:58,255
and it allows you to ask feedback

435
00:20:58,345 --> 00:21:01,405
to users using the tech manager so

436
00:21:01,405 --> 00:21:04,675
if you have that user mobile user

437
00:21:04,675 --> 00:21:07,789
that's in the payment page payment info

438
00:21:08,305 --> 00:21:09,715
from he's about to leave you can

439
00:21:09,715 --> 00:21:12,895
trigger an exit that can use an

440
00:21:12,895 --> 00:21:15,475
exit trigger the trigger and then present

441
00:21:15,505 --> 00:21:16,285
this person

442
00:21:17,425 --> 00:21:20,995
this model saying okay was what went

443
00:21:20,995 --> 00:21:23,574
wrong on this page what could have

444
00:21:23,575 --> 00:21:25,315
been done to make it work

445
00:21:26,755 --> 00:21:28,585
this is quite interesting

446
00:21:31,075 --> 00:21:32,185
there there there was

447
00:21:33,625 --> 00:21:36,115
i anxious it was so yes it

448
00:21:36,205 --> 00:21:39,925
was this morning at ten nine that

449
00:21:39,925 --> 00:21:42,265
was at a conference held by thomas

450
00:21:43,135 --> 00:21:43,735
pearson

451
00:21:44,755 --> 00:21:47,785
i encourage you to watch the playback

452
00:21:47,845 --> 00:21:48,685
of this

453
00:21:50,425 --> 00:21:53,575
a position because this is really interesting

454
00:21:53,605 --> 00:21:56,215
because it's all in matter more it's

455
00:21:56,215 --> 00:21:58,638
not an external tool you have all

456
00:21:58,638 --> 00:22:00,745
the feedbacks in metal you have

457
00:22:02,125 --> 00:22:04,555
the the completion of the feedbacks sent

458
00:22:04,615 --> 00:22:07,225
that can be sent as events so

459
00:22:07,225 --> 00:22:10,015
this is really something he can to

460
00:22:10,315 --> 00:22:12,295
take advantage of to have to better

461
00:22:12,295 --> 00:22:15,565
understand what's going wrong on our pages

462
00:22:18,565 --> 00:22:22,765
okay so now you have your optimization

463
00:22:22,795 --> 00:22:25,045
and tears you have sorted them by

464
00:22:25,045 --> 00:22:28,135
priority you have collected more info with

465
00:22:28,135 --> 00:22:30,295
your ex analytics features and you can

466
00:22:30,295 --> 00:22:33,655
describe the shoot marched at thirteen so

467
00:22:34,015 --> 00:22:36,505
as as said before for mobile users

468
00:22:36,535 --> 00:22:36,815
in the

469
00:22:36,835 --> 00:22:39,295
payments for the acceptance of said button

470
00:22:39,325 --> 00:22:41,755
is not visible enough you saw that

471
00:22:42,085 --> 00:22:44,379
on possession recording because you saw that

472
00:22:44,379 --> 00:22:47,365
the mouse moving and nuts are clicking

473
00:22:47,365 --> 00:22:50,005
the button you saw that also with

474
00:22:50,005 --> 00:22:51,715
the events you prepared on your taking

475
00:22:51,715 --> 00:22:53,837
pen for mobile users see the the

476
00:22:53,928 --> 00:22:54,985
click rate up this

477
00:22:56,155 --> 00:22:58,855
button is is really low you also

478
00:22:58,855 --> 00:23:01,045
see with the hidden obstacles assaults armed

479
00:23:01,135 --> 00:23:03,565
with a maps that people don't click

480
00:23:03,565 --> 00:23:06,055
on it and which makes the next

481
00:23:06,055 --> 00:23:08,695
button inefficient you can click on next

482
00:23:08,965 --> 00:23:11,413
to go through the through through to

483
00:23:11,485 --> 00:23:14,785
go through to the purchase process so

484
00:23:15,115 --> 00:23:16,195
these users do not

485
00:23:16,255 --> 00:23:18,115
stan was wrong and then the pitch

486
00:23:19,345 --> 00:23:21,415
so now you are able to write

487
00:23:21,415 --> 00:23:23,815
down your test of this but this

488
00:23:23,815 --> 00:23:25,345
is really important when you want to

489
00:23:25,369 --> 00:23:28,165
to do ab testing you should use

490
00:23:28,165 --> 00:23:31,465
this pattern if i'm in here describe

491
00:23:31,465 --> 00:23:32,725
what you are going to do on

492
00:23:32,725 --> 00:23:33,445
the page

493
00:23:35,365 --> 00:23:38,095
then users will and then you describe

494
00:23:38,305 --> 00:23:41,725
the change respect the user's behavior

495
00:23:43,225 --> 00:23:44,695
so you're

496
00:23:46,765 --> 00:23:49,705
if i make the accept terms of

497
00:23:49,705 --> 00:23:53,185
sales button more visible in tickets their

498
00:23:53,215 --> 00:23:56,905
payment info vulnerable users then they will

499
00:23:56,905 --> 00:23:59,245
click on it and be able to

500
00:23:59,245 --> 00:24:01,795
go to next steps and complete their

501
00:24:01,795 --> 00:24:04,405
purchase this is what you think is

502
00:24:04,405 --> 00:24:05,875
going to happen when you

503
00:24:05,905 --> 00:24:06,715
make the change

504
00:24:09,625 --> 00:24:14,125
once you decide that syrup i popescu

505
00:24:14,125 --> 00:24:14,875
will design

506
00:24:16,649 --> 00:24:17,515
a page

507
00:24:19,495 --> 00:24:20,035
and then

508
00:24:21,055 --> 00:24:24,175
you will redirect a certain amount of

509
00:24:24,175 --> 00:24:26,935
people to this page and the other

510
00:24:28,045 --> 00:24:29,695
go through the regular page and you

511
00:24:29,715 --> 00:24:31,705
you will want to to check if

512
00:24:31,705 --> 00:24:33,895
there is any difference in the performance

513
00:24:34,015 --> 00:24:38,065
of humans being the the conversion ecommerce

514
00:24:38,065 --> 00:24:41,065
conversion rates and you can start configuring

515
00:24:41,245 --> 00:24:43,405
your or two more ab test feature

516
00:24:43,525 --> 00:24:46,495
but before that you should use a

517
00:24:46,495 --> 00:24:48,115
simple size engine

518
00:24:48,115 --> 00:24:49,885
russian calculators because

519
00:24:51,775 --> 00:24:54,655
if we want the tests to the

520
00:24:55,075 --> 00:24:59,815
statistically significant we need to make sure

521
00:24:59,815 --> 00:25:03,753
that we have some requirements for instance

522
00:25:03,753 --> 00:25:06,805
here with at three percent conversion rate

523
00:25:06,805 --> 00:25:10,165
with eighty ninety five percents confidence level

524
00:25:10,555 --> 00:25:11,155
eighty percent

525
00:25:11,185 --> 00:25:15,955
statistical power and minimum fifty percent conversion

526
00:25:15,955 --> 00:25:18,025
rate lifts and three variants

527
00:25:19,195 --> 00:25:21,265
the calculator says okay if you want

528
00:25:21,265 --> 00:25:23,665
that to be statistically significant you need

529
00:25:23,694 --> 00:25:28,465
to have at least twenty three thousand

530
00:25:28,735 --> 00:25:29,215
had

531
00:25:31,825 --> 00:25:34,735
a sample size per round and also

532
00:25:34,735 --> 00:25:37,165
here it gives you the the lengths

533
00:25:37,555 --> 00:25:41,065
that the tests should should be in

534
00:25:41,065 --> 00:25:45,835
here for how the the minimum detectable

535
00:25:45,895 --> 00:25:47,335
effect you would see

536
00:25:48,385 --> 00:25:49,735
in the duration of the test

537
00:25:50,755 --> 00:25:52,375
so once you have these you know

538
00:25:52,375 --> 00:25:54,235
that you have you need to have

539
00:25:54,595 --> 00:25:55,915
this level of

540
00:25:57,805 --> 00:26:00,115
of for sample size this duration then

541
00:26:00,115 --> 00:26:02,755
you can go on and start configuring

542
00:26:02,755 --> 00:26:06,475
your ab tests so i recommend that

543
00:26:06,475 --> 00:26:09,554
you follow each step on the maternal

544
00:26:09,554 --> 00:26:13,045
interface called for that can you give

545
00:26:13,045 --> 00:26:14,695
a name to your test syria write

546
00:26:14,695 --> 00:26:17,155
down your test hypothesis

547
00:26:17,155 --> 00:26:17,485
this

548
00:26:18,917 --> 00:26:21,475
you're just the description of the test

549
00:26:22,105 --> 00:26:24,675
and the number of variation that you

550
00:26:24,756 --> 00:26:25,165
have

551
00:26:26,455 --> 00:26:28,225
and then again the tablet pages

552
00:26:31,315 --> 00:26:32,875
then you

553
00:26:34,375 --> 00:26:38,065
sorry i guess it's the same no

554
00:26:38,125 --> 00:26:40,465
it's not the same as your you

555
00:26:40,465 --> 00:26:43,945
go through the success metrics matter what

556
00:26:43,945 --> 00:26:46,765
we suggest you some metrics that you

557
00:26:46,765 --> 00:26:48,955
could use as a success metrics that

558
00:26:48,955 --> 00:26:51,595
you can decide to use totally different

559
00:26:51,595 --> 00:26:52,075
metrics

560
00:26:54,355 --> 00:26:57,595
sheer chief the success condition so sheer

561
00:26:57,595 --> 00:27:01,315
the minimal minimum detectable effects you want

562
00:27:01,405 --> 00:27:04,705
to have by default it's fifteen percent

563
00:27:04,705 --> 00:27:07,165
that you can you can write to

564
00:27:07,225 --> 00:27:10,225
defeat number but it's a necessary minimum

565
00:27:10,915 --> 00:27:12,955
the confidence threshold is most of their

566
00:27:12,955 --> 00:27:13,315
time for

567
00:27:13,375 --> 00:27:16,945
all ab tests tools about around ninety

568
00:27:16,945 --> 00:27:19,165
five per cent so just leave it

569
00:27:19,255 --> 00:27:21,235
as it is most of the time

570
00:27:21,235 --> 00:27:23,605
it's okay it's just something that says

571
00:27:24,655 --> 00:27:27,835
it's the probability that it's that the

572
00:27:27,835 --> 00:27:29,305
result will be the same over the

573
00:27:29,335 --> 00:27:32,935
time it says okay you this is

574
00:27:32,935 --> 00:27:33,475
not true

575
00:27:33,475 --> 00:27:35,275
due to a random result

576
00:27:39,145 --> 00:27:42,235
here you have the target pages

577
00:27:43,555 --> 00:27:44,814
i was a target until the ab

578
00:27:44,814 --> 00:27:47,905
test when and vista will not enter

579
00:27:47,905 --> 00:27:49,045
the test when

580
00:27:52,225 --> 00:27:54,235
here you have the traffic allocations so

581
00:27:54,235 --> 00:27:55,552
here you have one hundred percent of

582
00:27:55,555 --> 00:27:58,585
people on going through the desperate search

583
00:27:58,825 --> 00:28:01,795
it's not to a twenty that that

584
00:28:01,885 --> 00:28:04,375
that to that level alone most of

585
00:28:04,375 --> 00:28:08,449
the time here you have that again

586
00:28:08,485 --> 00:28:10,435
the traffic cap allocation

587
00:28:10,495 --> 00:28:12,325
for each variation it's shut

588
00:28:13,795 --> 00:28:14,365
let's say

589
00:28:15,685 --> 00:28:17,935
template patient factor in the test

590
00:28:21,055 --> 00:28:24,745
in here you have the and the

591
00:28:24,814 --> 00:28:28,255
you are a petition to have pitches

592
00:28:30,535 --> 00:28:32,485
here you have to open the start

593
00:28:32,485 --> 00:28:35,310
and the end of the page put

594
00:28:35,310 --> 00:28:39,277
that aside and then the code you

595
00:28:39,295 --> 00:28:41,125
will have to continue pitched to him

596
00:28:41,125 --> 00:28:43,375
to make the the test actor

597
00:28:46,735 --> 00:28:47,725
and yeah

598
00:28:48,625 --> 00:28:52,525
i put a screenshot of an ab

599
00:28:52,525 --> 00:28:54,505
test example coming from

600
00:28:55,585 --> 00:28:59,995
the matter move timo websites so when

601
00:28:59,995 --> 00:29:01,735
you wrote down that i put it

602
00:29:02,365 --> 00:29:04,405
into hypothesis in the

603
00:29:05,665 --> 00:29:07,855
in the interface here it's going to

604
00:29:07,855 --> 00:29:08,815
be reproduced

605
00:29:09,985 --> 00:29:12,528
here you have description and although the

606
00:29:12,565 --> 00:29:14,215
person performing sir

607
00:29:14,815 --> 00:29:15,355
metrics

608
00:29:17,575 --> 00:29:19,615
since hunting the tests has been running

609
00:29:23,095 --> 00:29:25,075
and here you are at the overview

610
00:29:25,165 --> 00:29:28,495
of the tests so you have the

611
00:29:29,665 --> 00:29:32,035
conversion rate for each of our nation

612
00:29:32,395 --> 00:29:35,423
the original impair they apply now so

613
00:29:35,423 --> 00:29:37,435
we have an average conversion rate in

614
00:29:37,465 --> 00:29:41,965
here we have the difference between the

615
00:29:41,965 --> 00:29:44,605
two of them and the detected lived

616
00:29:44,905 --> 00:29:48,325
here since we have a statistical significance

617
00:29:48,325 --> 00:29:49,735
of one hundred per cent we have

618
00:29:50,095 --> 00:29:52,375
a clear winner which means that on

619
00:29:52,375 --> 00:29:53,155
this test

620
00:29:55,375 --> 00:29:56,995
when you say if you change the

621
00:29:56,995 --> 00:29:58,825
top six years on the individual chuck

622
00:29:58,826 --> 00:30:00,715
page to the inaction about the chunk

623
00:30:01,285 --> 00:30:03,145
more users will apply to the job

624
00:30:03,145 --> 00:30:05,995
so the hypothesis was good because we

625
00:30:05,995 --> 00:30:09,115
see here that there is a real

626
00:30:09,415 --> 00:30:09,985
difference

627
00:30:11,215 --> 00:30:13,405
in in the way people acted

628
00:30:15,000 --> 00:30:17,010
add that to a contraction

629
00:30:19,590 --> 00:30:22,170
okay so i'm done

630
00:30:24,090 --> 00:30:27,480
maybe i was a bit too quick

631
00:30:27,570 --> 00:30:30,090
i don't know it's one thirty five

632
00:30:30,750 --> 00:30:33,960
so if you have any question just

633
00:30:33,960 --> 00:30:35,910
feel free to ask emma on the

634
00:30:35,910 --> 00:30:36,330
chat

635
00:30:37,620 --> 00:30:38,310
yeah pigeons

636
00:30:51,240 --> 00:30:52,860
okay so

637
00:30:54,570 --> 00:30:56,940
i don't see any questioning

638
00:31:11,730 --> 00:31:15,155
okay so since there is no question

639
00:31:15,155 --> 00:31:18,720
that i proposed to stop the the

640
00:31:18,720 --> 00:31:19,440
recording

641
00:31:21,180 --> 00:31:23,280
and that soon to end the session

642
00:31:25,650 --> 00:31:29,430
thank you for for this interesting session

643
00:31:29,430 --> 00:31:32,970
frederick i guess people might be feeling

644
00:31:32,970 --> 00:31:34,800
a little bit shy now but of

645
00:31:34,800 --> 00:31:38,160
course the chatroom will be open so

646
00:31:38,160 --> 00:31:41,520
everyone you can also continue the conversation

647
00:31:41,520 --> 00:31:44,520
there if you want to discuss

648
00:31:44,520 --> 00:31:48,720
cast farther with frederick about this session

649
00:31:48,720 --> 00:31:51,330
and what was discussed here so frederick

650
00:31:51,330 --> 00:31:54,030
once again many thanks for attending muthama

651
00:31:54,060 --> 00:31:54,450
camp

652
00:31:55,680 --> 00:31:57,270
thank you bye

653
00:31:58,320 --> 00:31:58,710
nine