Does anyone know who is running this ad? It was sent to me via email from a well-meaning friend who is clueless about what I do. It absolutely screams SCAM...but I have to wonder who is behind it..
Ah...used a trash email and found this: www.letsgetsocial.com/video/
Does anyone know if the stats in teh video are true?
I don't know... I couldn't get past the first 30 seconds. I will say this:
I don't think it's unfair of my to consider this to absolute bullshit based on:
http://www.warriorforum.com/warrior-forum-classified-ads/351138-become-social-media-manager-i-am-lets-get-social-2011-a.html
It's a $97 product (why do all of these people insist on prices ending in 7 even though tons of research has shown this to be utter BS?).
Quote from: ergophobe on January 09, 2012, 05:11:33 PM
It's a $97 product (why do all of these people insist on prices ending in 7 even though tons of research has shown this to be utter BS?).
Be interested to read any research on this. I don't know much about it but as a customer I'd be more drawn to $27 than $29.
Cheers. Can't see anything in there that suggests $97 would be more effective than $99 as a sales price, but then nothing that suggests it wouldn't be either.
$97 is three units away from 100, so when I'm glancing at it I don't automatically raise the cost to $100, but when I see $99 it just rounds up to $100 in my mind.
I think $97 is probably better than say $95 too, in that $95 is the last step before $100 when one counts by fives, you're less likely to do the jump in your brain with 97 as it isn't divisible by 3, 4 or 5.
In 207 fast-moving pages, Whitman teaches you dozens of well-guarded secrets that he learned during his 25+ years in the ad business, including:
*60% of people read only your headline and what to do about it
*Captions under photos get 200% greater readership than non-headline copy
*Ads with sale prices draw 20% more attention
* To double your ad's attention-getting value, you must enlarge it 400%
*Four-color ads are up to 45% more effective than black and white
*Prices ending in "95" are less effective than those ending in "99"
http://www.amazon.com/CA-HVERTISING-Ad-Agency-Psychology-Anything/dp/1601630328
I'll leave it up to you guys to actually read the book and determine the veracity of the author's claims.
Cheers Jason.
If Wal-Mart end their prices in .97 then it's clearly not a bad tactic.
There are people who claim they have split tested this, and that ending in .97 works.
I have seen enough people try to split test and get it wrong. to trust their claims, but I guess they could be right. Even so, I would expect something like that to 'trend' as it gets over used, and also to be affected by the market you addressing.
Damn! Friend sent me an article on this about a year ago and I can't find it. But they tested pricing strategies and did not find support for this.
First, let's be clear - testing might show that 97 outsells 96 in a particular situation. I'm not saying it never will. I'm saying it would be something you would need to test, rather than accepting an old saw with potentially little application to your product. Consider - what are the chances that someone buying a gambling book will be more attracted to "lucky" numbers than someone buying a math textbook?
My understanding is the whole seven thing comes from Ted Nicholas, who tested it pretty extensively, but he tested everything extensively and would probably
This is like the Cliff Notes version and he presents no real data, but FWIW, here it is:
QuoteI have no idea what scientific formula Ted used to arrive at his conclusions, but with all due respect to Ted Nicholas, his theory's a bunch of bull!
How do I know? Because I thoroughly tested his theory myself and discovered no discernible difference in response or profits.
I also base my conclusion on perhaps the greatest authority in marketing: Televison infomercials.... And almost without exception, every infomercial I've watched over the years, regardless of the product, have prices that end with a five or a nine. For example, $29.95, $39.99, $249.95.
http://www.streetdirectory.com/travel_guide/5067/marketing/the_truth_about_the_fallacy_of_7.html
QuoteMyth #6: Your Offer Price Must End in 7
Did you know there was magic in the number 7? Twenty or so years ago, the genius marketer Ted Nicholas ran some direct mail tests. He tested prices ending in 7 verses prices ending in 9. The 7′s outperformed the 9′s. Since then, this strategy has spread through most professions. Sure, it worked for Nicholas back then, but does it work for you? The only way to find out is to test. I know plenty of markets who have tested $7 offers against $10 offers, and the $10 almost always wins. So are we going to say all offers must end in 0′s now? You should try $17 or $27, but since I'm not superstitious I suggest keeping it simple by using multiples of 5 ($15, $25, $30, etc.)
http://dcpracticetools.com/the-7-myths-of-discounted-offers/
QuoteIf Wal-Mart end their prices in .97 then it's clearly not a bad tactic.
Many stores use the last digits as a sort of code. At REI, a price ending in .47 or .97 means the item has been marked down, so this settles disputes when someone says "it should be 20% off." If it hasn't been marked down, it doesn't end in .97 or .47. If it does, then the discount has already been applied. So it simplifies accounting, saves cashiers time in looking up pricing details, and saves as a redundant check for balancing books.
I'm not sure about how Wal-Mart does it, but I would not assume it's because they think that price maximizes profit. Clearly they think it does not *hurt* profit, otherwise they wouldn't do it. But it is even possible that a price might negatively impact revenue and positively impact profit if it saves REI money on accounting, accounting errors and so forth.
There's a lot of difference to money in the bank in the bulk Walmart sell at - 2 cents difference between $9.97 and $9.99 on x million products is worth having, so we can probably assume that they believe its fact it increases sales.
Unless they used to price at $9.95?, in which case all we can assume is that they believed people would still buy stuff if they added 2 cents.
I'm sure it would change over time though - I bet a lot more people are literally counting pennies now than were 2 years ago.
QuoteBecause I thoroughly tested his theory myself ....
And there is my problem .... I don't trust anyone to know how to test something properly except me.. and I don't trust myself a lot of the time.
I spent weeks getting my head round the maths to do statistical relevance testing when the sample size is potentially infinite and the test variables not delivered equally (long story as to why, but far more complicated than it sounds). In doing so, I read just about every 'split test guru' site out there and most of what I read was complete and total bollocks.
Using a chi-square test to get a p-value and then using a conservative significance threshold is good, but it doesn't help if the experiment is not set up correctly.
So the chi-square might tell you that the results are non-random with a 99.95% confidence, but if there are multiple explanations of why the results are non-random, you haven't gotten anywhere.
So I think Col was pointing out the difficulty of crunching more complex, multi-dimensional problems that don't lend themselves to simple chi-square testing.
Most of the conventional wisdom on split testing, IMO, falls down by using the same measures of statistical significance as a medical research test.
We do not need that level of confidence as:
* lives are not at stake
* is far too rigid for our industry
* our sample size is potentially infinite - and it is sample size and 'success' actions that determine significance (mostly)
As long as we can quickly eliminate the failures that would result in the test losing money, we can, and should, test indefinitely (although 'seasonality' and time in general would then become a factor to take into account)
This flexible sample size changes the maths completely.
Also, there are a number of strategies we can use that are not open to medical researchers - for example, we could use a fairly loose level of statistical significance to identify and eliminate the variants that perform worse than the control, whilst using a far higher level of significance (ie by running longer so we have a larger sample size) to decide which variants we end up running with.
Our concern is different from a medical researcher - we just need to get the test into profit as soon as possible - and if that means we accidentally go with the variant that long term testing would show to be the second best, who cares? we are in profit already, so carrying on testing is now cost free.
We have so much more flexibility, and so many different out of control variables in play, that it makes testing more challenging, and the interpretation of the results far more contentious than whether a new drug can be recommended for people with the lurgies.
Which is why I don't trust anyone else's tests - and only trust my own a little more than that :)
Quotecan be delivered
Woah!! I never said it couldn't - it can - and often very easily - its what I do for clients now.
(of course, it all depends on how sh## a position you are starting from :))
My point is that most people claiming to know what they are talking about in 'split testing' are not fully understanding the issues - their tests are often badly set up, and their conclusions suspect as a result.
[added]
I know this may seem obvious - but it still needs mentioning ..... in an open ended test, deciding when to stop, or at least change, a test, is crucial - and human nature being what it is, there are psychological issues with calling it complete when it matches expectations, or leaving it running longer when the results don't match expectations. Very common when working with clients, and even present to a small extent when working on your own stuff.
QuoteWe Agree
... oh right - way to kill a good thread Jason - I had me popcorn out and everything ;)
Its not bad - but I think I improved it