1 1 00:00:00,330 --> 00:00:02,760 Impact Analysis. 2 2 00:00:02,760 --> 00:00:04,680 When we talk about impact analysis, 3 3 00:00:04,680 --> 00:00:06,480 this is a really important concept 4 4 00:00:06,480 --> 00:00:08,610 as part of our triage function. 5 5 00:00:08,610 --> 00:00:10,380 Now, when we talk about triage, 6 6 00:00:10,380 --> 00:00:11,850 really what we're focused on 7 7 00:00:11,850 --> 00:00:14,970 is how do you look at an event and decide how severe it is, 8 8 00:00:14,970 --> 00:00:16,980 and how much priority you need to give it? 9 9 00:00:16,980 --> 00:00:19,950 After all, you only have so many resources. 10 10 00:00:19,950 --> 00:00:22,440 Now, triage is a concept that originally started 11 11 00:00:22,440 --> 00:00:23,820 with the military. 12 12 00:00:23,820 --> 00:00:25,920 Military doctors have limited resources 13 13 00:00:25,920 --> 00:00:27,240 when they're on the battlefield. 14 14 00:00:27,240 --> 00:00:29,970 And so, if they have 10 different injuries come into them 15 15 00:00:29,970 --> 00:00:32,640 and they only have enough resources to help three people, 16 16 00:00:32,640 --> 00:00:34,050 they need to go through very quickly 17 17 00:00:34,050 --> 00:00:36,870 and figure out who is the most severe case? 18 18 00:00:36,870 --> 00:00:38,760 Who needs assistance right now, 19 19 00:00:38,760 --> 00:00:40,290 and who can wait a couple hours, 20 20 00:00:40,290 --> 00:00:43,320 or who can wait a couple of days, or a couple of weeks? 21 21 00:00:43,320 --> 00:00:45,150 That's the idea of triage. 22 22 00:00:45,150 --> 00:00:46,320 Now, this is really important, 23 23 00:00:46,320 --> 00:00:48,330 and hospitals do this all the time. 24 24 00:00:48,330 --> 00:00:50,490 If you go to the emergency room right now, 25 25 00:00:50,490 --> 00:00:52,830 they're going to look at you when you first check in. 26 26 00:00:52,830 --> 00:00:55,740 And if you have a big gaping wound that's gushing blood, 27 27 00:00:55,740 --> 00:00:57,540 you're going to get seen right away. 28 28 00:00:57,540 --> 00:00:58,860 If you're having a major heart attack, 29 29 00:00:58,860 --> 00:01:00,327 they're going to help you right away. 30 30 00:01:00,327 --> 00:01:03,090 But if you go in and you have a high fever and a temperature 31 31 00:01:03,090 --> 00:01:04,590 and you think you might have the flu, 32 32 00:01:04,590 --> 00:01:07,020 well, you might be sitting there for several hours 33 33 00:01:07,020 --> 00:01:08,730 until a doctor has time to see you 34 34 00:01:08,730 --> 00:01:10,950 because your case is a lower priority 35 35 00:01:10,950 --> 00:01:13,260 than the person who needs urgent medical assistance 36 36 00:01:13,260 --> 00:01:14,790 because they have a gushing wound 37 37 00:01:14,790 --> 00:01:16,320 or they're having a heart attack, 38 38 00:01:16,320 --> 00:01:18,180 and that's the idea of triage. 39 39 00:01:18,180 --> 00:01:21,120 Now, as we go through and we start thinking about triage, 40 40 00:01:21,120 --> 00:01:23,340 we're going to consider different categories. 41 41 00:01:23,340 --> 00:01:26,340 For example, is there damage to data integrity? 42 42 00:01:26,340 --> 00:01:28,200 Is there unauthorized changes? 43 43 00:01:28,200 --> 00:01:30,660 Is there a theft of data or resources? 44 44 00:01:30,660 --> 00:01:32,790 Is there disclosure of confidential data? 45 45 00:01:32,790 --> 00:01:34,890 Is there going to be an interruption of services 46 46 00:01:34,890 --> 00:01:37,170 or will there be system downtime? 47 47 00:01:37,170 --> 00:01:39,300 All of these are things that we have to consider 48 48 00:01:39,300 --> 00:01:41,220 as we are doing our triage. 49 49 00:01:41,220 --> 00:01:42,600 Now, as I went through this list, 50 50 00:01:42,600 --> 00:01:44,970 you might've saw that last one with system downtime 51 51 00:01:44,970 --> 00:01:47,100 and thought, "That doesn't seem so bad. 52 52 00:01:47,100 --> 00:01:48,540 What's the big deal if the system is down 53 53 00:01:48,540 --> 00:01:50,340 for a few minutes or a few hours?" 54 54 00:01:50,340 --> 00:01:53,400 Well, in reality, system downtime is very expensive 55 55 00:01:53,400 --> 00:01:56,340 and costs our companies a lot of money. 56 56 00:01:56,340 --> 00:01:59,130 Back in 2019, there was a survey done, 57 57 00:01:59,130 --> 00:01:59,963 and they started looking 58 58 00:01:59,963 --> 00:02:02,310 at the average cost per hour of downtime. 59 59 00:02:02,310 --> 00:02:03,900 As you can see here on the chart, 60 60 00:02:03,900 --> 00:02:06,330 the average with about 25% 61 61 00:02:06,330 --> 00:02:09,780 was somewhere around 300 to $400,000. 62 62 00:02:09,780 --> 00:02:13,380 Now, that is a lot of money for one hour of downtime. 63 63 00:02:13,380 --> 00:02:16,830 Some companies though were as high as $5 million or more 64 64 00:02:16,830 --> 00:02:18,690 for just one hour of downtime 65 65 00:02:18,690 --> 00:02:21,270 with 15% of the respondents saying that. 66 66 00:02:21,270 --> 00:02:24,120 Now, if you look down at the up to $10,000, 67 67 00:02:24,120 --> 00:02:27,120 zero people said it cost them $10,000 or less 68 68 00:02:27,120 --> 00:02:30,690 for one hour of downtime of those larger companies surveyed. 69 69 00:02:30,690 --> 00:02:33,270 So, you can see that downtime is a costly factor, 70 70 00:02:33,270 --> 00:02:35,730 and as instant responders, we have to keep that in mind 71 71 00:02:35,730 --> 00:02:38,070 because if we have a system that is completely down 72 72 00:02:38,070 --> 00:02:40,320 and it stops us from being able to take in orders, 73 73 00:02:40,320 --> 00:02:41,250 that might be something 74 74 00:02:41,250 --> 00:02:43,650 that we need to prioritize resources to quickly 75 75 00:02:43,650 --> 00:02:45,210 to get that system back online 76 76 00:02:45,210 --> 00:02:47,370 and into a safe configuration. 77 77 00:02:47,370 --> 00:02:49,290 Now, triaging categorization 78 78 00:02:49,290 --> 00:02:51,690 is going to be done based on one of two things, 79 79 00:02:51,690 --> 00:02:55,050 it's either going to be impact based or taxonomy based. 80 80 00:02:55,050 --> 00:02:56,760 And these two approaches are different ways 81 81 00:02:56,760 --> 00:02:59,400 for us to consider how we want to prioritize things. 82 82 00:02:59,400 --> 00:03:01,440 If we're using an impact-based approach, 83 83 00:03:01,440 --> 00:03:03,570 this categorization approach is going to focus 84 84 00:03:03,570 --> 00:03:06,510 on the severity of an incident such as an emergency, 85 85 00:03:06,510 --> 00:03:09,540 significant, moderate, or low priority. 86 86 00:03:09,540 --> 00:03:11,820 If we're using a taxonomy-based approach, 87 87 00:03:11,820 --> 00:03:13,020 this approach is going to define 88 88 00:03:13,020 --> 00:03:15,330 the instant categories at the top level. 89 89 00:03:15,330 --> 00:03:18,120 For example, we might call them a worm outbreak, 90 90 00:03:18,120 --> 00:03:20,310 or a phishing attempt, or a DDoS, 91 91 00:03:20,310 --> 00:03:22,560 or an external host, or account compromise, 92 92 00:03:22,560 --> 00:03:24,570 or internal privilege abuse. 93 93 00:03:24,570 --> 00:03:26,220 Based on those different categories, 94 94 00:03:26,220 --> 00:03:28,980 we would have different procedures and different priorities 95 95 00:03:28,980 --> 00:03:31,350 based on the different categories we're using. 96 96 00:03:31,350 --> 00:03:33,720 Now, when you're using an impact-based approach, 97 97 00:03:33,720 --> 00:03:35,040 you're going to categorize the incident 98 98 00:03:35,040 --> 00:03:36,990 based on the scope and the cost, 99 99 00:03:36,990 --> 00:03:38,250 and this is usually the preferred way 100 100 00:03:38,250 --> 00:03:40,080 of doing this inside the industry. 101 101 00:03:40,080 --> 00:03:42,270 If it's a large scope and a very large cost, 102 102 00:03:42,270 --> 00:03:43,710 it's going to have a higher priority. 103 103 00:03:43,710 --> 00:03:46,440 If it's a smaller scope and it's going to have a smaller cost 104 104 00:03:46,440 --> 00:03:48,510 then we might delay our response a little bit 105 105 00:03:48,510 --> 00:03:49,343 because we're going to deal 106 106 00:03:49,343 --> 00:03:51,120 with other things that are more important. 107 107 00:03:51,120 --> 00:03:52,740 Now, in my real-world experience, 108 108 00:03:52,740 --> 00:03:56,220 most organizations seem to use the impact-based approach. 109 109 00:03:56,220 --> 00:03:57,150 And when they do that, 110 110 00:03:57,150 --> 00:03:59,310 they're going to use four main categories 111 111 00:03:59,310 --> 00:04:00,960 of impact that you can look at. 112 112 00:04:00,960 --> 00:04:04,680 These are the organizational impact, the localized impact, 113 113 00:04:04,680 --> 00:04:07,260 the immediate impact, and the total impact. 114 114 00:04:07,260 --> 00:04:09,090 Let's take a look at these four areas. 115 115 00:04:09,090 --> 00:04:11,280 First, organizational impact. 116 116 00:04:11,280 --> 00:04:12,960 Now, this occurs when you have an incident 117 117 00:04:12,960 --> 00:04:15,090 that affects mission essential functions, 118 118 00:04:15,090 --> 00:04:17,880 and therefore the organization can't operate 119 119 00:04:17,880 --> 00:04:19,320 as it's intended to. 120 120 00:04:19,320 --> 00:04:21,330 When you're considering an organizational impact, 121 121 00:04:21,330 --> 00:04:24,300 this is something that affects a wide range of users 122 122 00:04:24,300 --> 00:04:26,190 across your organization. 123 123 00:04:26,190 --> 00:04:28,500 For example, if my email server went down, 124 124 00:04:28,500 --> 00:04:31,260 that affects pretty much everybody in the organization 125 125 00:04:31,260 --> 00:04:34,800 so it has an organizational impact, that's a big deal. 126 126 00:04:34,800 --> 00:04:37,260 Then, we go and look at localized impact. 127 127 00:04:37,260 --> 00:04:38,700 This is an incident that is limited 128 128 00:04:38,700 --> 00:04:41,460 in scope to a single department, a small user group, 129 129 00:04:41,460 --> 00:04:43,050 or just a few systems. 130 130 00:04:43,050 --> 00:04:45,570 For example, if my customer service manager's computer 131 131 00:04:45,570 --> 00:04:47,730 has been hacked, that is one system 132 132 00:04:47,730 --> 00:04:50,010 that is localized to her issues. 133 133 00:04:50,010 --> 00:04:51,870 And so, while it's a bad thing for us 134 134 00:04:51,870 --> 00:04:54,390 and we need to solve it, it's not nearly as bad 135 135 00:04:54,390 --> 00:04:56,340 as something as large as all of the email 136 136 00:04:56,340 --> 00:04:58,230 across the company is down. 137 137 00:04:58,230 --> 00:05:00,450 Now, one thing you have to be careful of, 138 138 00:05:00,450 --> 00:05:03,120 don't always assume that a localized impact 139 139 00:05:03,120 --> 00:05:05,040 is necessarily less important 140 140 00:05:05,040 --> 00:05:08,220 or less costly than an organizational impact. 141 141 00:05:08,220 --> 00:05:09,480 Now, the reason why I say this 142 142 00:05:09,480 --> 00:05:11,580 is if I have a single computer 143 143 00:05:11,580 --> 00:05:13,440 or a single server that's affected, 144 144 00:05:13,440 --> 00:05:15,480 that may be a localized impact, 145 145 00:05:15,480 --> 00:05:17,880 but it may have less of an impact 146 146 00:05:17,880 --> 00:05:20,340 than a larger organizational impact. 147 147 00:05:20,340 --> 00:05:23,310 A good example of this might be our payroll system. 148 148 00:05:23,310 --> 00:05:25,650 We have one computer that runs our payroll, 149 149 00:05:25,650 --> 00:05:27,630 and every Friday we send out paychecks 150 150 00:05:27,630 --> 00:05:29,040 for all of our employees. 151 151 00:05:29,040 --> 00:05:31,680 Well, if that system was hacked on Thursday, 152 152 00:05:31,680 --> 00:05:33,540 that even though it's a localized impact 153 153 00:05:33,540 --> 00:05:36,090 and has one function that is being affected, 154 154 00:05:36,090 --> 00:05:39,000 one small part of the business, in our case payroll, 155 155 00:05:39,000 --> 00:05:40,830 that would be a bigger issue for us 156 156 00:05:40,830 --> 00:05:43,620 than the entire email server being down right now 157 157 00:05:43,620 --> 00:05:46,170 because we have to get those paychecks out on Friday. 158 158 00:05:46,170 --> 00:05:47,850 That's the kind of things you have to think about 159 159 00:05:47,850 --> 00:05:49,890 and why you need a thinking human 160 160 00:05:49,890 --> 00:05:52,170 being put in place to make these decisions 161 161 00:05:52,170 --> 00:05:53,707 because it's not as easy as saying, 162 162 00:05:53,707 --> 00:05:56,460 "This affects 10 computers, that affects one computer 163 163 00:05:56,460 --> 00:05:59,310 therefore 10 is more than one, that's a bigger issue." 164 164 00:05:59,310 --> 00:06:00,900 It's not necessarily that easy. 165 165 00:06:00,900 --> 00:06:03,240 It's always more important to think about what functions 166 166 00:06:03,240 --> 00:06:05,370 do those machines and systems serve 167 167 00:06:05,370 --> 00:06:09,270 inside of the organization or inside that local area? 168 168 00:06:09,270 --> 00:06:11,970 The next thing we want to talk about is immediate impact. 169 169 00:06:11,970 --> 00:06:14,490 Now, immediate impact is when an incident is measured 170 170 00:06:14,490 --> 00:06:17,640 based on the direct cost incurred because of the incident, 171 171 00:06:17,640 --> 00:06:22,020 such as downtime, asset damage, penalties, and fees. 172 172 00:06:22,020 --> 00:06:23,010 Based on these things, 173 173 00:06:23,010 --> 00:06:25,200 you're going to have some kind of dollar value, 174 174 00:06:25,200 --> 00:06:27,480 and that allows you to quantitatively assess 175 175 00:06:27,480 --> 00:06:29,490 how important this thing is to get fixed 176 176 00:06:29,490 --> 00:06:31,830 and how much time you can delay. 177 177 00:06:31,830 --> 00:06:34,380 The next one we're going to talk about is total impact. 178 178 00:06:34,380 --> 00:06:37,080 Now, a total impact is going to measure that incident 179 179 00:06:37,080 --> 00:06:38,340 based on the cost that arise 180 180 00:06:38,340 --> 00:06:40,590 both during and following the incident, 181 181 00:06:40,590 --> 00:06:42,930 including the damage to the company's reputation. 182 182 00:06:42,930 --> 00:06:45,030 So, we take all that immediate impact stuff 183 183 00:06:45,030 --> 00:06:47,880 and we add all the long-term stuff as well. 184 184 00:06:47,880 --> 00:06:49,770 Let me give you a good example of this. 185 185 00:06:49,770 --> 00:06:52,410 Now, a lot of companies have had data breaches, 186 186 00:06:52,410 --> 00:06:53,700 and it's bad 187 187 00:06:53,700 --> 00:06:55,560 and there's going to be a long-term impact to them, 188 188 00:06:55,560 --> 00:06:57,150 but it's not necessarily as severe 189 189 00:06:57,150 --> 00:06:58,950 depending on what industry they're in. 190 190 00:06:58,950 --> 00:07:02,250 But in my company, we're an online education company, 191 191 00:07:02,250 --> 00:07:03,270 and so if we look at all 192 192 00:07:03,270 --> 00:07:05,310 the online education companies out there 193 193 00:07:05,310 --> 00:07:06,600 and they had a data breach, 194 194 00:07:06,600 --> 00:07:08,730 for them it probably wouldn't be that bad. 195 195 00:07:08,730 --> 00:07:10,380 They would say they're sorry to their customers, 196 196 00:07:10,380 --> 00:07:12,687 they pay some fines, they would do some remediation, 197 197 00:07:12,687 --> 00:07:14,700 and they would work on it, and they would move on, 198 198 00:07:14,700 --> 00:07:16,170 and there would be no big issue. 199 199 00:07:16,170 --> 00:07:18,690 But if my company had a big data breach, 200 200 00:07:18,690 --> 00:07:20,430 that would have longer lasting effects 201 201 00:07:20,430 --> 00:07:23,100 and have a higher total impact, why? 202 202 00:07:23,100 --> 00:07:24,930 Because we teach cybersecurity, 203 203 00:07:24,930 --> 00:07:27,360 and if the cybersecurity training company gets hacked, 204 204 00:07:27,360 --> 00:07:28,950 that's going to look really bad for us, 205 205 00:07:28,950 --> 00:07:30,840 that's going to tarnish our reputation. 206 206 00:07:30,840 --> 00:07:33,990 And so, that is going to have longer total impact effects to us 207 207 00:07:33,990 --> 00:07:36,030 than a similar company who teaches something 208 208 00:07:36,030 --> 00:07:38,730 like how to draw, or how to play piano, 209 209 00:07:38,730 --> 00:07:40,170 or something like that. 210 210 00:07:40,170 --> 00:07:42,030 And so, you have to think about your industry 211 211 00:07:42,030 --> 00:07:43,980 and what that total impact will be, 212 212 00:07:43,980 --> 00:07:46,630 and that is going to impact your categorization as well.