Say you sit down to finish something and find another security reminder. You already have a list of things to check, update or remember. Adding one more sounds responsible, but you can’t see what it changes.
You can take security seriously and still ask what each task does for you. Some steps deal with a real risk. Others may ask for more time and effort than the protection is worth.
This module is for adults making their own everyday security choices. It assumes nobody is known to be targeting you, and that no likely loss would threaten the way you earn a living.
If you know someone is targeting you, including stalking or abuse, you need advice fitted to your case. The same goes for very exposed public roles, such as journalists, activists and public officials. It also goes for anyone who looks after other people’s money or private information. If a loss that could really happen would threaten your living or the savings you depend on, this everyday review isn’t enough.
What you’ll get from this module
In this module
By the end of this module, you’ll have a clearer way to decide which security tasks deserve your effort. We’ll look at an argument from economics and a set of interviews. We’ll look at studies of payment warnings, sign-in checks and browser updates, and at technical guidance on passwords. You’ll think about what the findings mean in daily life, then review one security demand. You’ll finish with a reason to keep it, change it or find out more, and you’ll name what would make you think again.
Ask Jing
Become a Jingius.
What does protective effort actually buy?
Economic analysis and interviews: effort has a cost.
Cormac Herley’s 2009 paper asks a practical question. Is the extra protection from a security task worth the extra effort it takes? This is an argument from economics. It isn’t an experiment showing that most security advice fails.
A loss can sound frightening. That doesn’t tell you how much of it one step would prevent. To judge whether a step is worth doing, weigh two things: the harm it could really prevent, and the effort it takes. Herley also warns against treating his argument as a licence to ignore security rules or advice.
Brian Stanton and colleagues’ 2016 study shows what that effort can feel like. They studied 40 interviews held in early 2011 with everyday users around Washington DC and central Pennsylvania. In a later look at the interviews, fatigue came up in 25 of them. It hadn’t been what the study first set out to find.
People described giving in, feeling they had lost control, and avoiding decisions. The interviews describe what these people went through. They don’t tell us how common security fatigue is across the country. Nor do they show that everyone gives up on protection.
Why it matters: Together, these accounts show why adding more tasks doesn’t by itself make a security list more useful. What a task asks of you matters too, next to the protection it can give. Feeling tired doesn’t make a task worthless. But it gives you a reason to look hard at the demands on your attention.
Read the economic analysis of security advice. Read the interview study of security fatigue.
Platform field experiment: a warning before a risky payment.
Xin Wen and colleagues’ 2022 study tested new warning designs during Alipay payments in China. The platform had already flagged these payments as possibly risky. Over two months, the researchers compared two new designs with the warning already in use. The paper reports that users were put into the groups at random.
The reported groups held 43,015 and 153,470 users, with about four million in the group that saw the old warning. The researchers measured a case rate. It counted fraud cases after a warning, set against the number of warnings shown.
That rate was 0.24% with the old warning. It was 0.13% and 0.17% with the new designs. These are shares of warning events. They aren’t shares of all users, or amounts of money saved.
The researchers also counted safe payments that went ahead despite a warning, again set against all warnings shown. This rate fell with the new designs. That points to a cost next to the benefit: honest payments may get interrupted too. Because the count was set against all warnings, it isn’t an exact rate of disruption among safe payments.
Why it matters: The result shows that the design of a warning can matter when a payment has been flagged as risky. It gives people a chance to think again just as they’re about to act. It doesn’t tell us how well a person’s own checking habits will work.
Read the field experiment on payment warnings.
Account-access study: protection must also be usable.
Periwinkle Doerfler and colleagues’ 2019 study looked at fourteen kinds of sign-in challenge at Google. A challenge asks for extra proof when the provider thinks a sign-in looks suspicious. The security part of the study covered more than 350,000 real attempts to get into accounts without permission. A separate part looked at about 1.2 million challenges, to see how real account owners coped with them.
Phishing means tricking someone into handing over private information. Challenges based on a device blocked more than 94% of the studied attempts that began with phishing. Some challenges based on knowledge blocked as few as 10% of those attempts. These figures describe blocked attempts in the data the provider chose. They don’t tell you what share of your own losses the same checks would prevent.
The findings also show why ease of use matters. The study’s summary says 52% of real owners failed to sign in at first, while 97% got into their accounts in the end. The follow-up lasted one week. Trouble at first and losing access for good are different outcomes.
An extra check can block an intruder and still make signing in harder for the owner. So counting the checks isn’t enough to judge the protection. You also need to know what proof they ask for, and whether the person using the account can give it.
Why it matters: The security comparison used challenges picked by the provider’s risk system. So its challenges and users can’t be treated as like-for-like groups in an experiment.
Read the study of account-access challenges.
Browser-update observations: routine protection can require fewer decisions.
Thomas Duebendorfer and Stefan Frei’s 2009 report compared four ways that browsers update, using Google’s web logs from around the world. It tracked whether active users had moved to particular releases.
By day 21 after a release, 97% of active users had the latest version of one browser that updated silently. For a browser that updated with one click, the figure was 85%. Each figure covered users within that browser’s own major version, not all browser users.
The comparison only watched what happened. The browsers, releases and users all differed, so the gap can’t be read as a clean test of automatic updating. What was measured was take-up, not money saved or time saved.
Why it matters: The comparison shows that routine upkeep can ask more or less of the user. Some updates need another decision from you. Others can happen without one. It’s worth asking which tasks really need your attention each time.
Read the report on silent browser updates.
Technical guidance: a calendar ritual and a meaningful trigger are different.
The National Institute of Standards and Technology’s July 2025 guidance on sign-in security is written for providers that check people’s access. It says they must not make people change passwords on a schedule. It also says they must require a change when there is evidence that a password has been compromised.
Why it matters: Time passing is one reason to change a password. Evidence that it has been compromised is a different one. Dropping a rule based on the calendar doesn’t remove the need to act when there’s evidence of compromise. This is guidance for providers. It isn’t an experiment that measures what households save by doing less.
Read the authentication guidance on password requirements.
What this means for you
Start by asking what harm the task guards against, and how it could help prevent it. Then think about what it asks of you: setting it up, deciding again and again, and any trouble using your own accounts. You don’t need to invent a money figure when you lack the facts to make one.
Check an unusual, important request through a separate route
When an unusual request involves a lot of money or sensitive access, pause. Check it through a contact or details you find yourself, apart from the request. The separate route matters. Carrying on inside the same request doesn’t give you an independent check.
Say you get an unexpected request for a large payment. Before you act, make contact yourself, using details you got apart from that request. The aim is to find out whether the request is genuine before you commit to it.
If you can’t find that out, leave the decision open while you look for what’s missing. A pause helps only if it makes room for a real check. Waiting, then following the same unchecked request, hasn’t answered the question.
This puts your attention where the decision has consequences. It doesn’t mean treating every ordinary activity as a new investigation.
Make important access both stronger and usable
When misuse of an account could do you serious harm, look for proof that doesn’t just ask you to remember one more fact. The study’s device-based challenges show the difference: they need you to have a device. Think about whether you can give that proof in normal use, and what setting it up and using it will involve.
The study shows why both questions matter. A check may block someone trying to get into your account, and make it harder for you to get in too. An easy check isn’t always protective. A hard one isn’t always better.
If an option is hard to use, work out what the problem is before you treat the setup as settled. You need protection you can use, or a clear question to take to someone who can help with your situation.
Let suitable routine protection run with less repeated judgment
Think about automating a task when three things are true. You understand what the process does. You can use it normally. And you can notice and deal with failures. Those conditions belong together. An automatic process that quietly stops working doesn’t meet the same practical need as one you can understand and keep an eye on.
The browser study makes automation worth a look, but it doesn’t choose your setup for you. Ask whether a routine task really needs your decision each time. If the process is confusing, or its failures would be hard to deal with, it may be worth keeping watch yourself.
Reduce a ritual only for a reason
For a repeated task that feels like a burden, name the harm it’s meant to guard against and how it could help prevent it. Then describe what doing it costs you. Being used to a task doesn’t mean it should stay on your list for ever.
The password guidance shows how a reason can matter more than a calendar. When you rethink a routine, keep the triggers that mean something and the protection that helps. Changing your protection because of evidence is different from repeating a duty just because time has passed.
Keep useful steps that ask little of you. Cut back or drop a repeated task when the effort it takes is out of proportion to the protection it can fairly give. If you don’t know the benefit, write that down. Don’t count it as zero and throw out an important step.
Say a repeated task feels like a burden, but you can’t explain what it prevents. For now, that gives you a question to answer. It isn’t proof that stopping is justified. If information you can trust supports an easier way to keep the protection, that gives you a reason to change the routine.
Revisit the decision when the situation changes
Write down what would change your answer: what you could lose, the threats you face, the protection on offer, or your ability to use it. That gives you a clear reason to look at the decision again.
If known targeting, responsibility for others or a loss that could ruin you comes into the picture, go beyond this everyday review and get a tailored assessment. Age alone doesn’t tell you what protection you need.
How about your daily life?
Which protective demand keeps asking for your attention? Is there an important account whose extra proof is awkward to use? Or routine upkeep that still needs you to decide again and again?
You could also think about how you would check an unusual request with a lot riding on it. Choose a situation you can look into before you commit money or access.
Write down some possibilities. One good idea is enough; more ideas can help you choose.
| A situation in my daily life | What I want to understand or improve |
|---|---|
| ____________________ | ____________________ |
| ____________________ | ____________________ |
My chosen situation: ____________________
Choose an everyday personal decision that fits this module’s scope. Picking it doesn’t mean a change in protection is justified. Take it into the exercise below.
Review one security demand
Use your chosen situation. If you started here, pick one everyday protective demand of your own that you want to understand before you change it. This exercise ends with a reasoned decision, or with a question you have stated plainly and not yet answered.
If you know someone is targeting you, including stalking or abuse, you need a tailored assessment. So do very exposed public roles, such as journalists, activists and public officials, and anyone responsible for other people’s money or sensitive information. If a loss that could really happen threatens your living or the savings you depend on, seek that assessment. Don’t use this everyday review to cut back protection.
- Name the demand and the harm. Describe what you’re being asked to do and what harm to you it guards against. Say why that harm matters in your situation.
Demand, harm and stakes: ____________________ - Identify how it could help. What part of that harm could this step prevent? Use what you can back up. Where information is missing, write that down; don’t invent odds or a savings figure.
Supported benefit and what remains unknown: ____________________ - Describe the effort and what needs to work. Include setup, repeated effort and ease of use. For automation, check that you understand the process and can notice and deal with a failure. For proof of access that doesn’t rest on memory alone, think about whether you can give it in normal use.
Burden and conditions: ____________________ - Choose the next step. Keep useful protection that asks little of you. Change or drop a routine only when its effort is out of proportion to the protection it can fairly give. Not knowing the benefit isn’t, on its own, enough reason to drop an important step.
For an unusual request involving a lot of money or sensitive access, pause. Check through a route you start yourself before you act. If an important answer is still uncertain, name the information or tailored help you need.
Keep, change or seek information — and why: ____________________ - Name a reason to reconsider. Write down a change in what you could lose, the threats you face, the protection on offer or your ability to use it.
Revisit if: ____________________
Ask Jing
Become a Jingius.
Your next step
You have finished when you can explain your choice and its conditions, or state the open question and a next step for answering it. You don’t need to remove a task for this review to be useful.
A note on the evidence
The sources look at different things: what users go through, what happens after warnings, blocked access attempts and take-up of updates. They don’t give one right amount of daily effort for everyone, or a reliable guess at what you’d save. The warning study is about payments a platform’s risk system had already picked out. The security comparison in the account study, and the browser-update comparison, only watched what happened.
Keep those differences in mind when you judge what a step can add. A useful result in one setting doesn’t fill every gap about your own burden or circumstances.
How we prepared this module
This module links an argument from economics and interview research with studies of warnings, account access and routine updates, plus guidance on sign-in security. Made-up everyday situations make the choices clearer. The written exercise helps you look at one protective demand and find a justified next step, or the information you’re missing.
Sources
- Cormac Herley. (2009). So Long, And No Thanks for the Externalities: The Rational Rejection of Security Advice by Users. Proceedings of NSPW 2009, pp. 133–144, conference 8–11 September 2009; ACM publication copyright 2010. Sections 7.4–7.6 and conclusion, PDF pp. 10–11.
- Brian Stanton, Mary F. Theofanos, Sandra Spickard Prettyman, and Susanne Furman. (September–October 2016). Security Fatigue. IT Professional, 18(5), 26–32. Author manuscript deposited in PMC, available 2 April 2024. Methods and Results.
- Xin Wen, Liang Xu, Jie Wang, Yuan Gao, Jiaming Shi, Ke Zhao, Fuyang Tao, and Xiuying Qian. (7 July 2022). Mental States: A Key Point in Scam Compliance and Warning Compliance in Real Life. International Journal of Environmental Research and Public Health, 19(14), 8294. 2022 published PDF, Study 2, sections 3.1–3.2 and Figure 4, pp. 10–12.
- Periwinkle Doerfler, Maija Marincenko, Juri Ranieri, Yu Jiang, Angelika Moscicki, Damon McCoy, and Kurt Thomas. (13–17 May 2019). Evaluating Login Challenges as a Defense Against Account Takeover. The Web Conference 2019, pp. 372–382. Abstract; sections 3.1–3.4, 4.1 and 5.1–5.3; Tables 2, 3, 5 and 7, PDF pp. 1, 4–10.
- Thomas Duebendorfer and Stefan Frei. (2009). Why Silent Updates Boost Security. ETH Zurich Technical Report TIK 302. Report describing April 2009 conditions; methods/results p. 5, discussion p. 7 and conclusions p. 8.
- National Institute of Standards and Technology (NIST). (July 2025). Digital Identity Guidelines: Authentication and Authenticator Management. NIST SP 800-63B-4, final. Section 3.1.1.2, Password Verifiers, periodic-change rule and compromise exception.
