1
00:00:00,680 --> 00:00:06,020
The last video we spoke about how we were going to take our instance of a survey and an e-mail template

2
00:00:06,020 --> 00:00:11,490
that we're about to create and merge the two together inside of a Meyler object at the very end I had

3
00:00:11,510 --> 00:00:17,540
said that somehow communicating this mailer off to the email provider is a rather complicated process

4
00:00:17,750 --> 00:00:21,590
particularly for the application that you and I are going to build.

5
00:00:21,620 --> 00:00:23,210
So let's dive into this.

6
00:00:23,210 --> 00:00:26,720
We're going to figure out why it's challenging and why this is all such a big deal.

7
00:00:27,260 --> 00:00:27,580
OK.

8
00:00:27,620 --> 00:00:29,600
So again we're going to look at a diagram here.

9
00:00:29,680 --> 00:00:35,660
That's going to help you understand exactly how this mailer gets sent off to the e-mail provider.

10
00:00:35,720 --> 00:00:37,640
We're going to look at two different diagrams.

11
00:00:37,640 --> 00:00:43,000
The first diagram is going to show you a bad way of communicating the mailer to the provider.

12
00:00:43,190 --> 00:00:46,930
And the second one is going to be the good way the way in which we are going to do it.

13
00:00:46,940 --> 00:00:48,730
So let's take a look.

14
00:00:49,500 --> 00:00:49,820
OK.

15
00:00:49,850 --> 00:00:51,440
So this is the bad approach here.

16
00:00:51,470 --> 00:00:56,070
This is what you and I do not want to do in this type of approach.

17
00:00:56,090 --> 00:01:03,440
I'm saying that for every single recipient inside of our survey list we could create a separate mailer

18
00:01:03,490 --> 00:01:04,570
object.

19
00:01:04,880 --> 00:01:10,160
We could then send that mailer object off to our e-mail provider and that e-mail provider would that

20
00:01:10,220 --> 00:01:13,550
send the e-mail off to the individual recipient.

21
00:01:13,550 --> 00:01:16,120
Now why might this be about approach right here.

22
00:01:16,490 --> 00:01:23,300
Well when you think about it if we say that for every single recipient we are going to create one mailer

23
00:01:23,360 --> 00:01:24,230
object.

24
00:01:24,380 --> 00:01:31,790
And then in one single HTP request we're going to send each and every Mailer or excuse me one separate

25
00:01:31,790 --> 00:01:33,320
request for every mailer.

26
00:01:33,320 --> 00:01:41,590
We would then have ten thousand or a thousand or 100000 separate requests required to handle one survey.

27
00:01:41,600 --> 00:01:49,850
In other words for every single recipient who is inside of one survey be it 1000 10000 or 100000.

28
00:01:49,890 --> 00:01:50,260
We will.

29
00:01:50,270 --> 00:01:55,310
We are saying that we're going to create a separate Mailer and communicate it to our e-mail provider

30
00:01:55,490 --> 00:01:58,570
in a separate HTP request.

31
00:01:58,610 --> 00:02:04,640
Now obviously off the bat we probably don't want to make a separate request over to this e-mail provider

32
00:02:04,670 --> 00:02:06,670
for every single recipient.

33
00:02:06,740 --> 00:02:14,180
It would be a lot better if we could somehow take all of these recipients create one mailer object and

34
00:02:14,180 --> 00:02:20,570
then just take that one single e-mail or that one single Mailer and communicate off to our e-mail provider.

35
00:02:20,570 --> 00:02:26,450
In other words I want to do a sort of batch operation and say for all the different recipients who are

36
00:02:26,450 --> 00:02:31,190
attached to a single survey just make one mailer and send it off.

37
00:02:31,190 --> 00:02:33,740
And so that's exactly what we're going to do.

38
00:02:33,800 --> 00:02:39,070
We're going to take all of the different recipients who are included inside of our survey.

39
00:02:39,140 --> 00:02:44,750
We're going to create one single mailer object and then we're going to pass that mailer off to our e-mail

40
00:02:44,750 --> 00:02:51,890
provider and one Cingular network request will then leave it to our e-mail provider to figure out exactly

41
00:02:51,890 --> 00:02:56,270
how to handle the e-mail to each individual recipient.

42
00:02:56,270 --> 00:02:56,910
OK.

43
00:02:57,200 --> 00:03:01,240
So this is the approach we're going to take right here for all these different recipients.

44
00:03:01,250 --> 00:03:06,120
Will make one mailer we'll send the one mailer off to the e-mail provider.

45
00:03:06,620 --> 00:03:13,720
Now how does this approach affect our application and it really does in a big way.

46
00:03:13,730 --> 00:03:19,400
So let's kind of break this down let's consider some of the cons of this approach right here and figure

47
00:03:19,400 --> 00:03:23,600
out exactly how we're going to address them inside of our application.

48
00:03:23,630 --> 00:03:29,000
First thing I want to do is look back at a diagram that we were just glancing at in the last video.

49
00:03:29,000 --> 00:03:36,140
Remember we had said that for a given mailer we have one survey instance and we have one e-mail templates

50
00:03:36,650 --> 00:03:43,400
and this e-mail template right here provides the kind of HVM L or kind of the body of the actual e-mail

51
00:03:43,400 --> 00:03:44,820
that a user receives.

52
00:03:45,080 --> 00:03:49,400
So essentially that e-mail template is creating this stuff right here.

53
00:03:49,400 --> 00:03:50,730
The message to the user.

54
00:03:50,810 --> 00:03:55,910
And it also is going to set up the two possible links that a user can click to provide feedback to our

55
00:03:55,910 --> 00:03:57,280
application.

56
00:03:57,320 --> 00:04:04,010
So if we are using the exact same templates the exact same Like instance of the template for all these

57
00:04:04,010 --> 00:04:06,490
people what's the downside to that.

58
00:04:06,690 --> 00:04:09,560
Let's think it through for a minute.

59
00:04:09,560 --> 00:04:16,890
So if everyone is getting the exact same e-mail like I mean every last character 100 percent identical

60
00:04:17,450 --> 00:04:20,230
even the links inside the e-mail are identical.

61
00:04:20,330 --> 00:04:21,780
And that's the important part here.

62
00:04:21,800 --> 00:04:26,000
So let's imagine that everyone is getting the exact same links and the e-mail.

63
00:04:26,090 --> 00:04:31,040
How are we going to somehow understand who clicked on a link.

64
00:04:31,040 --> 00:04:31,650
Whenever they do.

65
00:04:31,650 --> 00:04:37,010
Like if you get this e-mail and you click on No right now how am I going to know that you clicked on

66
00:04:37,010 --> 00:04:39,780
this e-mail as opposed to someone else.

67
00:04:39,820 --> 00:04:41,300
And why do we care about that at all.

68
00:04:41,300 --> 00:04:47,480
Well remember that was the entire purpose of recording that list of recipients inside of our survey

69
00:04:47,480 --> 00:04:53,390
schema we had said that we want to store the list of recipients so that for every different email we

70
00:04:53,390 --> 00:04:58,730
send out or every single recipient we have indicated by this e-mail property right here we want to make

71
00:04:58,730 --> 00:05:02,860
sure that we record whether or not they have responded to the e-mail previously.

72
00:05:03,240 --> 00:05:08,170
And that we were going to do that to make sure that we kind of put a curb on any sort of duplicate votes

73
00:05:08,170 --> 00:05:09,330
inside of our application.

74
00:05:09,340 --> 00:05:12,010
We don't want people to be able to vote multiple times.

75
00:05:12,130 --> 00:05:16,160
So we were going to start off everyone with a responded Of false.

76
00:05:16,270 --> 00:05:20,060
And then after you provide us feedback one time we flip this over to true.

77
00:05:20,080 --> 00:05:22,700
And then after that we no longer record your votes.

78
00:05:22,960 --> 00:05:28,540
So remember we went through this entire process of ensuring that we could somehow uniquely identify

79
00:05:28,540 --> 00:05:33,990
people in order to combat possible fraudulent votes or duplicate votes.

80
00:05:34,000 --> 00:05:39,610
And so when you look back at the actual e-mail right here it obviously becomes very pinnacle to make

81
00:05:39,610 --> 00:05:45,210
sure that you and I understand exactly who is clicking on this link.

82
00:05:45,250 --> 00:05:46,880
Now why is this such a challenge.

83
00:05:46,930 --> 00:05:53,980
Remember again you and I we do not or we are not going to have the ability to create a separate link

84
00:05:53,980 --> 00:06:00,310
right here for every individual user because everyone is going to be getting the same exact mailer so

85
00:06:00,370 --> 00:06:08,240
everyone sees the exact same link so just to kind of drill this point home because I want to make sure

86
00:06:08,240 --> 00:06:11,480
that it's really clear really you know very very clear.

87
00:06:11,480 --> 00:06:17,000
The last thing I want to say here is in a perfect world we would have been able to set up this route

88
00:06:17,000 --> 00:06:25,010
right here or this link right here as send you to somewhere like email dot com surveys slash feedback.

89
00:06:25,280 --> 00:06:30,420
And then maybe we indicate the idea of the survey with like you know 1 2 3 right here.

90
00:06:30,530 --> 00:06:33,800
So maybe this email corresponds to survey 1 2 3.

91
00:06:34,010 --> 00:06:38,840
And then maybe if the user clicked on the no button we would do something like Slash no to say save

92
00:06:38,840 --> 00:06:40,080
the user clicked on no.

93
00:06:40,280 --> 00:06:45,260
And then in the link right here Ideally we would have been able to put the user's e-mail address so

94
00:06:45,260 --> 00:06:50,770
something like email or user out example dot com.

95
00:06:50,810 --> 00:06:56,810
So inside of just the URL of this link right here or if someone clicked on this link we could say OK

96
00:06:56,810 --> 00:07:00,090
someone's trying to give us feedback for survey 1 2 3.

97
00:07:00,140 --> 00:07:01,390
They want to vote no.

98
00:07:01,430 --> 00:07:04,610
And here is the e-mail address of this particular user.

99
00:07:04,640 --> 00:07:08,570
So if we've been able to do something like this it would have solved all of our problems and we could

100
00:07:08,570 --> 00:07:14,870
say that we could customize every single link here when the user clicks it just by the path that the

101
00:07:14,870 --> 00:07:18,070
user linked on we can figure out exactly what to do.

102
00:07:18,280 --> 00:07:21,770
And so what I'm saying very clearly here I'm trying to make sure this is really clear.

103
00:07:21,770 --> 00:07:28,340
I'm telling you that everyone is going to get the exact same link so we cannot somehow communicate the

104
00:07:28,340 --> 00:07:30,340
email inside the link.

105
00:07:30,350 --> 00:07:31,790
This is the best we can do right here.

106
00:07:31,820 --> 00:07:38,630
At best we can say hey if you click on this link it'll take you to feedback for survey 1 2 3 and you

107
00:07:38,630 --> 00:07:40,360
are attempting to vote yes or no.

108
00:07:40,640 --> 00:07:47,630
We cannot put any identifying token into the email because everyone gets the exact same one from a single

109
00:07:47,690 --> 00:07:48,600
mailer.

110
00:07:49,130 --> 00:07:49,570
OK.

111
00:07:49,610 --> 00:07:52,550
So I feel like it probably drilled at home enough.

112
00:07:52,790 --> 00:07:57,830
Hopefully you understand it by now hopefully understand that we need some mechanism to uniquely identify

113
00:07:57,830 --> 00:07:59,880
users who click on these links.

114
00:07:59,960 --> 00:08:01,850
So obviously there's a solution to this.

115
00:08:01,850 --> 00:08:03,430
Let's talk about what we're going to do.

116
00:08:03,880 --> 00:08:09,740
So to solve this problem this is going to go into a little bit of the behind the scenes stuff with our

117
00:08:09,800 --> 00:08:12,950
e-mail provider and the e-mail provider that we're going to use.

118
00:08:12,980 --> 00:08:17,000
So I can tell you right now is a provider called cent..

119
00:08:17,330 --> 00:08:19,870
So we'll sign up for this provider in just a minute.

120
00:08:19,880 --> 00:08:26,600
But if you want to you can open up a new browser tab and go to send grid dot com and check out a little

121
00:08:26,600 --> 00:08:28,960
bit there documentation.

122
00:08:29,060 --> 00:08:33,050
So we're going to talk a little bit about what send grids does behind the scenes.

123
00:08:33,050 --> 00:08:38,870
That's going to solve this problem that we have been able to identify our users uniquely.

124
00:08:38,870 --> 00:08:39,440
All right.

125
00:08:39,470 --> 00:08:40,860
So let's get to it.

126
00:08:41,480 --> 00:08:47,900
So whenever we send grid to send out an e-mail to one of our users send grid does something a little

127
00:08:47,900 --> 00:08:50,890
bit automatic behind the scenes for us.

128
00:08:50,960 --> 00:08:57,740
Here's what they do for each user that centigrade is going to send an e-mail to they look at the body

129
00:08:57,740 --> 00:09:03,180
of the e-mail and they look to see if any links exist in that e-mail.

130
00:09:03,530 --> 00:09:08,090
And in our case there will be links right where we have the yes link and then no link.

131
00:09:08,300 --> 00:09:12,610
We never send grid sees a link inside of our e-mail.

132
00:09:12,710 --> 00:09:20,600
They automatically replace the link with a customized link that sends the user to their own servers.

133
00:09:20,600 --> 00:09:26,240
So in other words regardless of what you are l you and I put inside these two links right here regardless

134
00:09:26,240 --> 00:09:31,880
of what you and I do when censored sends out this e-mail they are going to replace the destination links

135
00:09:31,880 --> 00:09:37,960
inside of here with links that send the user to their own personal send grid server.

136
00:09:38,300 --> 00:09:43,820
And they do this to build a collect metrics on what links people are clicking inside of all these different

137
00:09:43,820 --> 00:09:45,660
e-mails they're sending out.

138
00:09:45,680 --> 00:09:52,460
So whenever a user clicks on Aslund great here they will be sent off to send grid's server grade is

139
00:09:52,460 --> 00:09:58,370
going to record some analytic information about the user and then send good sense off the user to the

140
00:09:58,370 --> 00:10:03,380
actual destination of the link like the actual original destination.

141
00:10:03,380 --> 00:10:05,180
So how is that going to help us out here.

142
00:10:05,180 --> 00:10:07,670
Well here's what happens.

143
00:10:07,670 --> 00:10:13,060
So when we tell grid to send an e-mail out to people send it is going to scan the e-mail.

144
00:10:13,220 --> 00:10:17,510
It's going to replace every link with their own very individual special one.

145
00:10:17,720 --> 00:10:20,830
And then here is the very important part here.

146
00:10:20,990 --> 00:10:27,560
When when when grid replaces these links with their own special ones they customize the links for each

147
00:10:27,650 --> 00:10:31,850
individual person that the e-mail is going to be sent out to.

148
00:10:31,850 --> 00:10:38,060
So even though you and I are communicating this email over to send grade with a single mailer object

149
00:10:38,300 --> 00:10:44,390
which room which means that we only have like exactly one unique copy the e-mail sent grid goes back

150
00:10:44,390 --> 00:10:51,140
through the e-mail in that customizes each e-mail to be very particular to every specific recipient

151
00:10:51,230 --> 00:10:54,000
every specific person who actually gets the e-mail.

152
00:10:54,380 --> 00:11:01,100
So when Segretti replaces these links they put some token into the link that uniquely identifies the

153
00:11:01,100 --> 00:11:01,880
user.

154
00:11:01,880 --> 00:11:06,510
So send Gred in fact knows what user is clicking on a link.

155
00:11:06,550 --> 00:11:12,850
We don't send grid does because they go through this customization process just for the purpose of collecting

156
00:11:12,850 --> 00:11:14,600
some type of metrics.

157
00:11:14,620 --> 00:11:20,020
So what we really want to do is somehow get censored to tell us whenever someone clicks a link on an

158
00:11:20,020 --> 00:11:20,510
e-mail.

159
00:11:20,530 --> 00:11:22,730
And that's exactly what we're going to do.

160
00:11:22,750 --> 00:11:29,080
So after sending semi-rigid replaces all these links they send out the e-mail a user clicks on the link

161
00:11:29,440 --> 00:11:32,450
send grid sees that the user clicked on the link.

162
00:11:32,470 --> 00:11:35,260
They know exactly who clicked on it.

163
00:11:35,440 --> 00:11:38,800
The user gets sent off to the destination and the user is happy.

164
00:11:39,070 --> 00:11:45,820
But then more importantly Saenger then sends a little message to our server informing us that hey someone

165
00:11:45,820 --> 00:11:47,720
just clicked on one of your links.

166
00:11:47,800 --> 00:11:48,970
Here's the link it was.

167
00:11:49,120 --> 00:11:54,760
And here is some information about that particular user who clicked on the link specifically the e-mail

168
00:11:54,760 --> 00:11:57,100
address of the person who clicked on it.

169
00:11:57,220 --> 00:12:03,670
And so this whole process right here is what is going to save us from not knowing who clicks on the

170
00:12:03,670 --> 00:12:05,230
link.

171
00:12:05,770 --> 00:12:06,510
So one more time.

172
00:12:06,520 --> 00:12:07,560
I know it's a long section.

173
00:12:07,570 --> 00:12:12,660
I know that I'm repeating this stuff about why the links you know why we can't really use this effectively.

174
00:12:12,670 --> 00:12:17,140
But I again I really want to make sure that this stuff is very clear because we are definitely in a

175
00:12:17,140 --> 00:12:21,040
jump through a couple of hoops to make sure that all this stuff works correctly and I want to understand

176
00:12:21,310 --> 00:12:23,200
why we're taking this extra trouble.

177
00:12:23,200 --> 00:12:24,070
So from the top.

178
00:12:24,070 --> 00:12:32,230
One more time you and I are creating a single mailer object that mailer contains exactly identical content

179
00:12:32,290 --> 00:12:35,600
for every single user that we are sending the survey out to.

180
00:12:35,620 --> 00:12:40,200
So we have no ability to customize these links on a per user basis.

181
00:12:40,360 --> 00:12:44,710
So we do not know who is clicking on any individual link.

182
00:12:44,890 --> 00:12:50,800
Thankfully when we tell it to send out the e-mail send great is going to replace every link with one

183
00:12:50,800 --> 00:12:57,010
of their own and the link that says great replaces it with is a very special link that is customized

184
00:12:57,010 --> 00:13:03,700
for each individual particular recipient of the email after the user clicks on the link send grid then

185
00:13:03,700 --> 00:13:09,520
knows exactly who clicked on it because they customized the link for every individual person said Great

186
00:13:09,550 --> 00:13:16,410
Well then send a message to our server telling us about the link now this very last step of the process

187
00:13:16,410 --> 00:13:20,760
here where second grade eventually sends a message back to our server telling us that someone clicked

188
00:13:20,760 --> 00:13:27,690
on something just you know this process is referred to as a web hook a web poke is anything where some

189
00:13:27,750 --> 00:13:34,050
outside API is kind of facilitating some process and then gives our application some type of callback

190
00:13:34,050 --> 00:13:37,410
or some type of little notice that some events just occurred.

191
00:13:37,620 --> 00:13:39,860
So we refer to this as a web book.

192
00:13:40,040 --> 00:13:46,650
And so maybe now if we go back to our route definitions Baqir the route we had spoken about to record

193
00:13:46,650 --> 00:13:51,750
feedback from a user we had named it API slash survey slash web hooks.

194
00:13:51,750 --> 00:13:58,770
So this right here this route is going to be a route that only send grid accesses and send it at some

195
00:13:58,770 --> 00:14:03,210
intervals are going to be sending us a little notification saying hey someone just clicked on one of

196
00:14:03,210 --> 00:14:03,660
your links.

197
00:14:03,660 --> 00:14:05,090
Here's the information about it.

198
00:14:05,130 --> 00:14:07,960
So that's why we call this thing Web hook back here.

199
00:14:08,490 --> 00:14:09,340
OK.

200
00:14:09,750 --> 00:14:11,160
So that's the entire process.

201
00:14:11,170 --> 00:14:15,360
It is the I think that we've gone through it in great deal detail at this point.

202
00:14:15,360 --> 00:14:16,500
So let's take a quick break.

203
00:14:16,500 --> 00:14:18,180
We're going to come back in the next section.

204
00:14:18,210 --> 00:14:23,580
We're going to start to sign up with send grid and integrate center grid into our application in order

205
00:14:23,580 --> 00:14:25,210
to send out all these e-mails.

206
00:14:25,360 --> 00:14:28,020
So see in the next section.