Forum Moderators: skibum
If you polled webmasters, I believe, you would find very little tolerance for pop-up ads.
Here is a survery that says that users are far more tolerant than might be suspected.
Does this information have any implication on how you should be marketing your site?
http://www.emarketer.com/analysis/marketing/20020220_mark.html
Your observation a few times that us at WMW are not a good generalization to the web browsing population as a whole is a very good reminder for me anyway...
That said, no info on the sample selection method of the quoted survey either, nor anything to do with methodology. Coming from company that is owned by a market research company, we are always taught that if you dont know the make-up of the sample just ignore the findings..
However as I said, this article is useful in getting the alternative view discussed. Love to know the details of the methodology though...
Here is the url to a list of their white papers available now. The white paper itself is in pdf format.
http://www.dynamiclogic.com/white_papers.php
More information is available from Dynamic Logic about this study. See its last page for Dan Safran's contact information.
I get very annoyed with them. One reason being is the fact that some popups give my computer a cold while it loads up, where i cant get rid of it or click elsewhere
The advertising trend of the future seems to point towards advertisements being even more intrusive, i.e. even more annoying (IMO) than pop ups
The Times of London was serving up a Volvo pop-up the last time I visited that site. The New York Times was popping-up Orbitz Travel.
My guess is that the acceptance by users is really there and that pop-ups are a viable advertising medium.
In the chart above that, only the universally despised telemarketing was thought to be significantly less desirable than pop-up ads.
And a "near majority," to borrow that writer's term, of 46% of the respondants said that it's "appropriate" to see only one or zero pop-up ad per hour, no more.
So... that's "acceptance?" Depends on how you read the numbers.
No one asks the question about whether we like the ads interspersed in a morning newspapers or in our magazines. If we polled people I expect that they would like their TV and radio without commercials too.
So there is a big difference between people liking ads and people finding them intolerable. I am guess that what we are finding is that people find pop-ups tolerable.
If they weren't being done to death and dragged through the mud because of all these stupid and annoying (in your face) ads that nobody asked for ... I would use them on my site. Some webmasters have absolutely no taste and no class. Like everything else in life, too much of a good thing ...
The difference is that most of them don't know they can do anything but accept it. The smarter ones download popup-killers. The use of popup killers has increased exponentially in the last two years as advertisers failed to believe that low click throughs were due to poor targeting or crappy offers and instead decided to force everyone to see their ads.
We recently ran a test, and made a dhtml layer that 'appeared' to be a popup, to the smallest detail, but would not be blocked by popup-killers. Within 3 days we had to remove it due to a massive outcry from the users, and a corresponding drop-off in page views.
Thos are real facts, not predictions based on asking a few people slanted questions. :)
No-one likes popups, but they pay more. Much more. That's the fact.
Ammon Johns
It was a competitor's site, so I figure I really don't have to worry about them at all even if they are 1 position above me in the SERPS. NOBODY in their right mind would go back to that site!
BK could you elaborate on this? It certainly is a creative idea :)
Is there anyway to have the layer behave as a pop-under? This might very well be the future of pop-up advertiseing. I mean people won't go around using dhtml killers will they? :)
Greektomi
Funny you should mention that. I've noticed that sites who were among the first to carry significant numbers of pop-ups, such as Yahoo and ESPN.com, have either severely cut back the number they show or have axed them altogether.
Many who were trying pops are already swarming to the new giant-sized ads that sit right in the middle of content. That in itself speaks volumes on the sustained value of pops.
As for CTRs being more valuable data than user acceptance rates, I find I get more clicks in the long run if my users don't get mad and go elsewhere, if you know what I mean (and I think that you do).
Certainly,
I created a DHTML layer in which I placed a table, and to which I gave a small 3d border so that it appears to be a new window. On the top row of the table, I placed a titlebar-looking text bit, and in the top-right corner, placed an X button that would 'close the window' (make the layer invisible). The second row of the table contained the ad itself, which can be any kind of graphic or text link desired. The third row contained text saying: "This advertisement will self-close in a few moments. Thank you for your support."
The layer began as hidden or invisible, and made itself visible as a javascript OnLoad event, once everything on the page had loaded (including the image that would appear inside it). We set it to vanish automatically 20 seconds later, but during that 20 seconds, it constantly repositioned itself to the center of the screen, even when they scrolled. Mean. ;)
I can tell you that while it caused such uproar that we had to remove it, it certainly got everyone's attention. In such a layer, you can even place standard banners and charge for it as a premium slot (triple rates).
Ammon Johns
I think you a miscasting the Dynamic Logic survey. Here's what they say about their methodology in there white paper.
"The data was gathered through an online survey hosted by Dynamic Logic. Respondents were invited to participate via email and offered a chance to win a sweepstakes as an incentive. The invited list of 5,000 was randomly selected from Dynamic Logic’s database of respondents who have participated in prior research. All respondents in the database were recruited via random sampling on websites. 413 respondents completed the survey, which
is an 8% response rate. The data was gathered from September 28 – 30, 2001."
I don't doubt your results and I don't doubt the Dynamic Logic survey either. The trick is to understand why your results and the survey results are different.
Large sophisticated advertisers on high traffic sites are using pop-ups. They seem to be getting an adequate ROI so that means they work as an advertising medium. This suggests that people are tolerating them just as we tolerate TV commercials. We don't like them but we can live with them and a whole industry can thrive because of them.
If they are a viable medium then perhaps some sites are indeed missing the boat by not using them.
Respondents were invited to participate via email and offered a chance to win a sweepstakes as an incentive. The invited list of 5,000 was randomly selected from Dynamic Logic’s database of respondents who have participated in prior research.
This is, of course, a weakness faced by any polling organization: that participants are always to some extent self-selecting. I know in the case of political polling firms (the only kind I've worked with) they work some magic on the numbers to attempt to overcome the skewed numbers that are the result. So predictions of the votes in an election are not typically precisely the same as the proportion of respondants who answered one way or the other.
My point is, I don't know what the effect might be in this poll, but there probably is one. Maybe people who would respond to this kind of survey are more tolerant of advertising... or maybe less.
Yep, that's called stratified sampling. There is a lot of activity in the market research field now on Web based surveys as a way to replace some other methods. And clients are asking for it. Yet even in countries with large internet penetration like US and Singapore, they create a whole heap of new sampling problems.
Most internet pollsters are using a competition as an incentive. I agree that, straight away, that in itself would bias a sample so much that any stratification 'magic' would be useless. You lose those who are not natural gamblers!
And 8% response rate PLUS 'experienced poll takers'. ??!!?? I wouldnt touch those results with a 10 foot pollster!
Possibly another example of 'a tiny bit of knowledge is a very dangerous thing'.
You're right, in focusing on the point that it's not a random sample I hadn't even thought about how low the response rate is. For that matter, there's no mention of margin of error or confidence level, and a sample of 413 is very small regardless of the response rate or selection method -- considering the size of the parent group.
In other words, while these number spark some interesting discussion, there's not a lot of scientific validity to them.
You could use a javascript OnExit command and effectively turn this into an on exit pop-up am I right?
I have no experience with dhtml but I might could finger it out. I would love to see the source code, sticky it to me if you feel generous :) Thanx for the reply anyhoo.
Greektomi
Here is the url to the companies description outlining their client base and the founders expertise.
http://www.dynamiclogic.com/company_description.php
Does the fact that the webmaster opinions being voiced here mean that the survery is somehow wrong?
Or does it strengthen my original observation that "Because webmasters spend so much time online, their attitudes don't always agree with the more casual users"?
(edited by: JayC at 5:38 am (utc) on Mar. 22, 2002)
Nah.. - they may well have clients and "credentials" but the research design in this case makes the results non-authoritative. If they are well credentialed, as the revious poster wrote, it may not be their responsibility but that of the journalist who summarissed their findings without checking out the full report properly.
Good research companies like these would provide a disclaimer of their results noting limitations due to research design or sampling in their original report. Im guessing that the journalistic reporting of results here (as is almost always the case) doesn't go this far.
It may be that in this case "...our more in-depth questioning and enquiring attitudes don't always agree with the more casual users".
As professionals we should be more enquiring I think than the casual user who can be easily misled by sensationalistic reporting.
And im not saying necessarily that this article is sensational, nor questioning the truth of your general statement. I agree with you that webmasters are a very different group than 'casual surfers', as i have documented already, but in this case it's an advantage, as we cant be so easily convinced by poorly documented research reports.
I will defer judgement till we see further evidence of this admittedly "different" finding, and repeatability in studies -:)
To isolate the impact of advertising exposure on consumer attitudes, two groups of online consumers are sampled at the same time and from the same Web sites on which the campaign is running. As the only difference between the groups is the presence of the advertising, any attitudinal differences between the two groups can be attributed to the exposure to the ad campaign.
This is not recognised as a proper methodology by the body that sets the international standards for Marketing Research, ESOMAR. To be accredited as a 'real' and 'reputable' market research authority, your 'sample' must be representative, and may not be recruited via any means that slew or skew the 'representativeness' of the sample.
For example, you cannot recruit panelists via banner ads, nor any web site built for that purpose. This would skew the panel towards those people who respond to banners and web pages, and away from those who don't, thus invalidating all data.
Strangely, Market Research companies are allowed to recruit panelists from pop-up ads, simply because the click-through ratios are so much higher, but this seems likely to naturally skew all such samples, by excluding those with pop-up blockers, or those who always close pop-ups without reading them, from even the potential to be included in the panel.
Ammon Johns (who has spent the last year marketing for a Market Research Institute)
To accurately judge the Dynamic Logic survey, we'd have to know every detail of the study. How did they create their database of "prior research respondents" (an ad? A pop-up ad)? What was the wording of the title of the recruitment e-mail (are they primarily surveying people who are by nature receptive to things such as unsolicited e-mail)? Also, we'd need the wording of the questions, and a complete demographic breakdown of the respondents. That's how scientists report their findings -- they tell all.
As a professional, without third-party findings I am inclined to be skeptical because this is such a volatile issue.
And as for other sites and advertisers using pop-ups, it definitely does not prove acceptance. It proves only experimentation.
Exactly. These aren't social scientists trying to understand an aspect of human nature, they're market researchers who work for marketers who make their money only if people buy online advertisement. It's to their benefit to promote the idea that the public is all but clamoring for pop-ups.
It's certainly true that webmasters on the whole see this issue differently than does the general public. On the other hand, while speaking here we're speaking in a relatively closed forum and have little motivation to misrepresent what we feel or to put a particular spin on it. But a firm whose profits are built on the successful sales of online advertisement, speaking in a publicly-released document, would have such a motivation.
Dynamic Logic to the best of my knowledge does NOT fit this category. Their business as I understand it is to independently measure the results of a clients online advertising. They audit results and supply business information. They are not in the business of selling advertisements.
