Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
Stats

The Quiet Collapse of Surveys: Fewer Humans (and More AI Agents) Are Answering Survey Questions (substack.com) 68

Survey response rates have collapsed from 30-50% in the 1970s to as low as 5% today, while AI agents now account for an estimated 20% of survey responses, according to a new analysis.

The UK's Office for National Statistics has seen response rates drop from 40% to 13%, with some labor market questions receiving only five human responses. The U.S. Current Population Survey hit a record low 12.7% response rate, down from 50% historically.

The Quiet Collapse of Surveys: Fewer Humans (and More AI Agents) Are Answering Survey Questions

Comments Filter:
  • Does this include political surveys? I mean, we all consider political polls to be suspect especially when they don't agree with our own views. Is this charming little stat going to call into question every political poll now? And how do we know that the statistics are accurate? Aiiiiiiiieeeeeeee!!!

    • by rossdee ( 243626 )

      Political Polls should only be sent to people who can actually vote.
      (or will be able to vote at the time of the election (so citizens turning 18 before the next election would be included)

      AI's can't vote.

      • by Sique ( 173459 )
        It's even more complicated than that. There are various types of polls: Polls of people allowed to vote, polls of registered voters, polls of likely voters... Look at the actual poll, they will tell you what type of people they interviewed.
    • Well yeah, that's why when you see political polling get aggregated it gets weighted based on various metrics of quality. Even something like exit polling is going to get tilted somewhat by being non-random of the people leaving the poll.
    • All this about survey responses, according to...a survey?

  • by mccalli ( 323026 ) on Tuesday May 20, 2025 @10:13AM (#65390311) Homepage
    OK - so let's assume the lowest participation rate is also the lowest number of physical answers - that's not a given, but taking this assumption would actually show things in the most optimistic light. Under this assumption if 5 is 13%, then ~39 (rounding up) is 100%. That survey then was sent to only 39 people - that's quite a specialised survey.

    The method of engaging needs to be questioned as well as the statistical fall in responses, I feel. Also the relevancy of the survey - if I'm not interested in the answer, then even if you've found the perfect way to reach me I'm still not going to respond.

    Also - anecdote not data, but I've been working in the UK for 35 years now. Never received any engagement from the ONS other than the standard once every four years census.
  • Other reasons (Score:5, Interesting)

    by dargaud ( 518470 ) <.slashdot2. .at. .gdargaud.net.> on Tuesday May 20, 2025 @10:20AM (#65390323) Homepage
    There are many reasons why people don't want to do surveys anymore:
    • Waste of time
    • Phishing
    • What's in it for you even when it's a valid survey ?
    • Etc...

    I recently got into hot water at work. I received a request for survey and the link was to some external domain so it went straight to spam. Only when there was some debate on the merit of the questions on the work mailing list did I mention that my filter sent it straight to spam and I wasn't certainly the only one (the boss' boss wasn't happy). If people are lazy and make their surveys with external tools/websites that will phish all your info they shouldn't be surprised at the lack of replies...

    • Re:Other reasons (Score:5, Insightful)

      by rahmrh ( 939610 ) on Tuesday May 20, 2025 @10:39AM (#65390385)

      It boggles the mind how stupid companies are (both the employers and say BANKS). They want to train people to look for spam from 3rd parties, and then they send surveys and actual critical information that they want answered using 3rd party sites.

      This is probably the #1 security violation (sending and expecting employees and others) to use/answer a 3rd party site that companies do without a single thought that they are increasing their security risk by their own actions poorly though out actions.

      • by Anonymous Coward

        It boggles the mind how stupid companies are (both the employers and say BANKS). They want to train people to look for spam from 3rd parties, and then they send surveys and actual critical information that they want answered using 3rd party sites.

        This is probably the #1 security violation (sending and expecting employees and others) to use/answer a 3rd party site that companies do without a single thought that they are increasing their security risk by their own actions poorly though out actions.

        Yes! HR wonders why no one is responding to a link from "surveymonkey." Man, you hit one IT phishing link and your work life just went to hell for a while. Side note to IT people, why not put the effort and energy to finding real phising links and put them into span yourselves? I thought you were the experts.

      • Re: (Score:2, Interesting)

        by Anonymous Coward

        It boggles the mind how stupid companies are (both the employers and say BANKS). They want to train people to look for spam from 3rd parties, and then they send surveys and actual critical information that they want answered using 3rd party sites.

        My employer does this and worse, they use "Safe Links" and click counters both of which hide the actual URL!

        • Ugh, safelinks "protection". Yeah, protect me from being able to read the URL so that I can't determine whether it matches the content. Clowns.

          At least our internal surveys are done internally. Well, I say that, but we use Microsoft software, and the EULA says they can exfiltrate any data they want and show it to anyone, so...

      • Just today, I got an email from eBay, which appears to be genuine (DKIM and SPF pass) with a "we changed your password, click this link to reset it" button.

        These big sites are training people to be victims of phishing.

        I say that it appears to be genuine, but with unicode characters in domain names, one can't really be sure.

      • by antdude ( 79039 )

        So annoying. They should host their own surveys.

      • I have a merchant services account through First Data and I get emails from at least 8 different domains, have disparate logins on another 6 or so domains, all of which require password changes every 90 days, there are at least a dozen different numbers to call depending on what kind of help you might need if something isn't working.

        On top of that, they're still sending statements to an old business address I left in 2016 and they can't seem to figure out how to change it to the address I've been at for
    • by Anonymous Coward

      Your boss' boss adds no value to the company that could not be replaced by AI.

      83 out of 100 survey responses agree.

    • Re:Other reasons (Score:4, Insightful)

      by whoever57 ( 658626 ) on Tuesday May 20, 2025 @11:54AM (#65390583) Journal

      I suspect that the biggest reason is simply weariness. So many companies ask you to complete their surveys these days. Who has time for that? I suspect that the surveys don't show real data anyway, because the only time I fill them in is when I am annoyed at the company or service.

      Then, even if you give a detailed reason why you are unhappy, you typically get an email asking why you are dissatisfied. Do they not read their own surveys?

      Yeah, people are simply weary.

    • Exactly.

      The person giving the survey is usually paid. Why shouldn't the people responding be paid?

      If you're not gonna pay me, fuck off with your survey.

    • We have like 10 tools that can do surveys and polls internally, but the company uses an external company anyway. :shrug:
      With the amount of simulated phishing attacks we get, and the consequences of falling for one are not worth the effort of checking whether the external survey link is legit.
      They keep telling us that they use the survey results. They probably use it to evaluate middle management. That's the consensus among the workers anyway.

  • Surveys (Score:4, Insightful)

    by ledow ( 319597 ) on Tuesday May 20, 2025 @10:21AM (#65390325) Homepage

    Why should I give me time for free for the purposes of providing you with data?

    Pay me and I'll take the survey.

    Google pay me a pittance each time I answer a question in their Rewards app. My ex- used to make a living filling out surveys and doing mystery shopping.

    Why should I give you my time and data for nothing?

    Pay me, and I'll fill out surveys for you. It doesn't even need to be a lot. Free money is worth it for clicking a few buttons when I have an idle moment, or can't sleep.

    But this is people valuing their time, probably because their free time is being crushed by current economies and without recompense there's no value in filling out a survey.

    Hell, people will pay you to sit there solving CAPTCHAs...

    • I did get mildly excited once to receive Nielsen poll, which did include some nominal compensation, because it made me feel like maybe I could help make sure my favorite shows got renewed, but yes even stuff like opinion polling seems like it could be useful until you realize the question wording is making it a push poll.
    • Why should I give me time for free for the purposes of providing you with data?

      Why should you comment on Slashdot? The biggest differences between the two is that polls can influence people and Slashdot comments display vanity handles and scores.

      • Why should you comment on Slashdot?

        In exchange for comments, I get to participate in a site where other humans (and some bots) are posting comments.

        What do I get for answering a survey? Fuck-all. Maybe I help someone make better products, which I might but probably won't buy.

        I'd rather shout into the void of cryptocuckery than help corporations directly for free.

    • Years ago I forget the name of the company now, but these people did in person surveys on various things. You got paid, like 100+ to drive over and sit down for an hour or so in a focus group. I used to go if I had time, because as you say, they paid me.
  • Not surprising in the least. For that last 25 years we've been taught that our data is valuable (and it is, Silicon Valley has made billions off it). Not just our email addresses but our opinions and preferences. Why on earth would anyone give away their data for free to a survey? At least Google and Facebook are offering some service of marginal value in return.

  • by nealric ( 3647765 ) on Tuesday May 20, 2025 @10:27AM (#65390347)

    Businesses over the last 10 years got too enamored with giving surveys for every single transaction. If you are handed a survey to fill out every time you buy a $5 item, eventually you are going to tune out. Plus, customers are rarely appropriately compensated for their time. Why should I spend my time helping a for-profit company do market research for free?

    • by rahmrh ( 939610 )

      They seem to be asking for a survey to make everyone "FEEL" better (after support calls and the like), but based on nothing changing and no responses I don't believe ANYONE/ANYTHING processes and/or reads any of those surveys. They surveys are just there to placate everyone.

      It is similar to "your call is important to us, please stay on the line" and everytime I say to myself "if your call was important to me you would have more people to answer".

      • I have to agree. The mind numbing emails to see if you liked the product/transaction whatever. Never fill them out. Imagine my shock when I bought a Motu from sweetwater and they called me with a friendly guy that wanted to know if it was working for me, how I liked it etc. The human element was fantastic. I rarely buy stuff from music places as I'm not a musician, but if I ever need electronics again like the Motu, I'd buy from them just because they are human, even if cost was a bit higher.
        • I don't want a human calling me either. Why should I spend time on the phone to tell a company that I liked their product? If I need help with the product, I will contact them. In any event, I don't answer calls from unknown numbers. If it's important, they leave a voicemail and I will call back. If a company called me about a product I purchased, I would not return that call.

          • It was a short call, so did not bother me. It did tell me they were ready to answer a call if I had one. Unlike lemme see, if I had a problem with google or amazon. An amazon truck actually knocked over my stone mailbox, and the driver would have driven away if I had not seen it happen. Even with that, it was a couple months to get it resolved because contacting them was such a difficult PIA. I applaud good customer service, even if it takes a moment of my time. Good being the operative word in my sentence.
    • Metrics for customer service are valuable to the staff and management, and often point to concrete corrective action. In the old days people managed this stuff by their gut and experience, and that works to a point but it can also be a disaster. The smart move back in the day was to calmly listen to customer complaints, and try to assess the root of the problem when you are hearing similar complaints on a regular basis. It's more ad hoc than a survey, but functions in a similar way.

      The problem with surveys

      • Metrics for customer service are valuable to the staff and management, and often point to concrete corrective action.

        Bull. If surveys had any effect on businesses we would see improvement in service. Instead, everything continues to degrade to crap.
    • Businesses over the last 10 years got too enamored with giving surveys for every single transaction.

      I recently bought some chunky mild steel bar stock cut into lengths, and then a day or two later duly got the survey/review link. Not sure what to say? 4.5* delicious but slightly chewy?

    • Auditors have gotten super bullish on NPS surveys over the last while. If your company is audited (and it probably is if you're not a mom-n-pop), they were probably pushed to routinely survey customers. If your company is audited and more than some threshhold number of employees they want to to survey them too.
  • Like anyone in their right mind wants to or has time to answer all the requests they get to give feedback through online surveys these days, often before the product or service in question are the worst. That said phone surveys that the caller says will only take five or ten minutes and end up taking a half an hour instead, occasionally with the *same* question asked over and over again in different ways are the worst. I had a phone survey about a small pipeline oil spill into a creek in an adjacent count

  • by Kokuyo ( 549451 ) on Tuesday May 20, 2025 @10:29AM (#65390355) Journal

    When I fill out surveys, there's seldom an answer that actually fits my opinion on the thing in question.
    And in science, especially in terms of things like happiness ant contentment, the answers are plagued by so many biases...

    Am I naive in thinking there could be an upside to abandoning this method of information gathering?

    • by cruff ( 171569 ) on Tuesday May 20, 2025 @11:08AM (#65390471)
      I agree with this. Too many surveys fail to actually give you a way to answer how you really feel about something. Either they are too coarse or don't allow you to choose "irrelevant" or "not applicable". There is also a theshold beyond which I will abandon a survey, either for the time spent or when they insist on asking those supremely annoying "variation on a previous question" to try and tease out some distinction about how you really think/feel.
  • by Virtucon ( 127420 ) on Tuesday May 20, 2025 @10:36AM (#65390373)

    IRL: I just bought a Candy Bar and a Coke! I don't need to click your QR code to tell you how I felt the transaction went!

    When everybody wants feedback, nobody gets feedback!

  • 1. You get assailed by requests to fill out surveys. You have to be gullible to wast your valuable time on these.
    2. More often than not, the data is used to maximize profitability instead of providing better value to the customer. In other words, the data is used against you.

    I never fill out surveys. If I have a problem with a product or service, then I'll complain using formal channels. If you don't hear from me about the quality of your product or service, that's good news. Formal complaints cost the prov

  • When we hit the point where you get a survey every time you order a pizza, every time you buy gas, every time you so much as step foot outside your house? Yeah, of course response rates have fallen. People are sick to fucking death of answering leading surveys that try to prod you toward the answer the corporation wants. Fuck 'em. Not everybody needs a pat on the head every other second. Us normies sure as shit don't get one. Fuck the corporations for trying to make us do it for them.

  • I mean really guys, Having you bots answer surveys? If you want to further degrade the quality of data your AIs are consuming, then go ahead.
    We live in the stupidest possible version of a cyberpunk dystopia.

    • by mspohr ( 589790 )

      It's too late.
      AI is already polluted with too much of its own AI generated garbage.
      The internet has become a self-sustaining system that reinforces its own garbage. Enshittification end stage has arrived.

    • Who pays? People who want to influence survey results, of course.

  • by laird ( 2705 ) <lairdp@NoSPAm.gmail.com> on Tuesday May 20, 2025 @11:12AM (#65390481) Journal

    Survey response rates completely depend on context.

    For example, if you're in a paid panel that does high value surveys for real research, response rates are fine.

    If you're sending out fake "push polls" or fundraising appeals using a fake poll as a hook, and there are a flood of those, they've trained people that polls aren't real, they're just scams of one sort or another, so people tune them out. I would not be shocked at all that the scammers have driven people away from all polling. Which is why real pollsters have paid panels of people who opted in.

  • You go to a web page, and 10 second later you get a popup asking you to take a survey. Don't let the user actually do what they came to the page for, just offer the survey so quickly, the user hasn't had a chance to even look around at the web page. So, surveys become useless, because no one wants to be interrupted by these things. An exit survey that would pop up as you are leaving would make far more sense.

  • Roku sent me one recently and I decided to try it to see if I could express my displeasure with their enshitification regarding ads. They did have responses to some questions to indicate I wasn't happy with ads, but follow up questions focused on whether the problem was they weren't relevant enough or other things that assume all is fine if they can make the ads "better" for me.

    There is no "better". Years back they were my choice independent of services like apple and google. They've since made tweaks to c

  • I ENJOY taking surveys. Market research, political, social-- everything. So where's the disconnect?

    I know how to spot scams and everywhere a surveyor would normally seek participants, I've all but cut out of my life. I don't open random spam emails. I don't take random surveys I find on social media. I don't answer phone calls from numbers I don't know or regions I'm not in.

    So what's the solution? I would HAPPILY subscribed to a validate survey service that either emails me digital surveys relevant to my de

  • I got fed up with those asking for opinions on the future of the organisation, service, etc. Then you start reading the questions, and the only answers are the pre-selected options for you to agree/disagree with, and none of them are what I'm hoping the company/service is doing. It's like Netflix asking a loaded question such as "Do you like adverts in your videos?". If you say 'no' they spin this to say "Users don't want adverts so we're putting up the price.". Or they ask, "Do you prefer your adverts be
  • Why oh why, should I answer any survey?

    "We are polling one million people to prove our narrative is right. Please answer the following question: When did you stop beating your wife?"

  • Even before AI bots, surveys only ever measured the opinions of people who answer surveys.

  • I'm done doing surveys for nothing.

    If you want my survey input, give me some output. Otherwise, fuck right off.

    I used to do surveys for loot boxes, at least I kind of got something.

  • Most surveys are businesses/government asking, "Did I do x?" The answer is invariably "yes". The quality, relevance or importance of x, is never examined.

    Surveys have stopped being open-ended solicitations for personal experience, they are work-flow check-lists the resident/customer is required to complete for free. Of course, people are disengaging, there's no benefit to it. Worse, every business is demanding an additional 'pat on the back' for making a sale.

  • Every time I see a physician, I get a survey request from my healthcare provider.
    I get one from my brokerage company also.

    What am I going to do? Complain when I'll be seeing them again. Sometimes, I have no choice. For example, my medical provider will only allow me to see one specialist who isn't much help. If I complain, I'll have to face him on my next visit. Same goes with my stock broker. Yes, I can change but changing organizations or providers is a major hassle with no guarantee it'll be better

  • Most surveys here in the Netherlands always seem to start with "How likely are you to recommend $COMPANY to a friend?" on a scale of 1 to 10. This is such a stupid question that I hit ctrl-W immediately at that point. Please people, be a bit more creative with your questions. As cruff said earlier, not being able to choose "irrelevant" on some questions also means that the data is going to be worthless.

  • Yeah, right, 1000 people in the US make a good survey.

    Reasons surveys are bullshit:
    1. They pick by LEC, so they can choose neighborhoods... or people who used to live in those neighborhoods.
    2. The questions are written such that they will get the results they want. There are no opposites.

Men occasionally stumble over the truth, but most of them pick themselves up and hurry off as if nothing had happened. -- Winston Churchill

Working...