Cognitive Biases — The Anchoring Effect
How often are we hooked on that first bit of information we get? Do you fall for it too?
Anchoring or focalism is a cognitive bias that influencing our decision-making abilities. This bias uses our reliance on an a certain piece of information. Typically, the first bit of information we receive becomes an anchor and all future evaluations are based on this anchor piece of information.
Kahneman and Tversky experiments
The anchoring effect is a well documented bias and the best researchers in this field are widely regarded as Daniel Kahneman and Amos Tversky, who’ve have carried out many experiments.
One of their books I would recommend is Thinking, Fast and Slow. Here is just one example of their testing from September 27th 1974:
Two groups of high school students estimated the answers to the below sums, within 5 seconds. One group (group A) estimated the result of:
8 x 7 x 6 x 5 x 4 x 3 x 2 x 1
while the other group of students (group B) estimated the product of:
1 x 2 x 3 x 4 x 5 x 6 x 7 x 8
Given the short time the students had, they were only able to do the first couple of multiplications and then had to guess the rest based on their initial sums.
What Kahneman and Tversky found was that the median of the students sums for the the descending sequence (group A) was 2,250 whilst the ascending median (group B) had an estimate median of 512. The actual correct answer to the multiplication sum is 40,320.
When someone sets the anchor, in this case the result of the first few sums, they then evaluate whether the overall answer is a little or a lot higher or lower and then adjust their estimate to take that into account.
Group A had higher value sums to work out first (8 x 7 x 6…), so their initial calculation would be higher than that of Group B, who had the lower value sums (1 x 2 x 3…). Because group A’s initial calculation (the anchor point) was higher than group B’s, their overall answers were generally also higher than group B’s.
3 Tier Advertising
A common example of the anchor bias is the 3 tiered approach. What we can see in the example above is three price points; an expensive £18.02 top tiered package, a £6.59 mid-range package and a ‘cheap’ £2.92 package.
We take the most expensive price point as our anchor point, which then fuels the perception that the £6.59 package is better value than the £18.02 one, because it’s cheaper and the middle option.
In many situations, people make estimates by starting from an initial value that is adjusted to yield the final answer. The initial value, or starting point, may be suggested by the formulation of the problem, or it may be the result of partial computation. In either case, adjustments are typically insufficient…that is, different starting points yield different estimates, which are biased toward the initial values.
– “Judgment Under Uncertainty” by Kahneman, Slovic and Tversky
Further Examples
In terms of selling and marketing, the seller would benefit from establishing the value of a product first. Once a seller sets a price, the negations with potential buyers will revolve around that initial valuation, because the buyer is unconsciously anchored to that price.
We’ve all been exposed to this hundreds of times without even noticing. For example, if you’re told a house is worth £200,000 then you would assume it’s got to be worth more than a £150,000 house, even if they appeared exactly the same at first glance.
But is it really worth an extra £50,000? Have you been tricked into anchoring on to what you’ve been told it’s worth?
The anchoring effect can easily slip in unannounced. Drazen Prelec and Dan Ariely conducted an experiment at MIT in 2006 where they had students bid on items in a bizarre auction.
Ariely explains in his book, Predictably Irrational, that the researchers would hold up a bottle of wine, or a textbook, or a cordless trackball and then describe in detail how great it was.
Then, each student had to write down the last two digits of their social security number as if it was the price of the item. If the last two digits were 11, then the bottle of wine was priced at $11. If the two numbers were 88, the cordless trackball was $88.
After they wrote down the pretend price, they bid on the items. Sure enough, the anchoring effect scrambled the students ability to judge the value of the items. People with high social security numbers paid up to 346% more than those with low social security numbers.
For example, people with numbers from 80 to 99 paid on average $26 for the trackball, while those with 00 to 19 paid around $9.
Social security numbers were the anchor in this experiment only because we requested them. We could have just as well asked for the current temperature or the manufacturer’s suggested retail price. Any question, in fact, would have created the anchor. Does that seem rational? Of course not.
“Predictably Irrational” by Dan Ariely
Despite the above examples, cognitive biases don’t have to be used negatively all the time.
In a 1975 study by Catalan and Lewis et al, researchers asked a group of students to volunteer as camp counsellors for two hours per week for two years. They all said no.
The researchers followed up by asking if they would volunteer to supervise a single two-hour trip. Half said yes.
Another group of students were asked straight out if they would volunteer to supervise a single two-hour trip. This time, only 17 percent agreed.
In the second group, less students agreed to a single two-hour trip because they did not have the anchor of a two-year commitment, so it made a single trip seem more off-putting as they had nothing to compare it to.
Anchoring Bias and Product Pricing
When shopping, you tend to see prices ending in a 9. Thanks to neuropsychological data, marketers know that a vase priced at $39 versus $40 is not merely perceived as $1 cheaper, it is perceived as $10 cheaper.
This tactic has been dubbed ‘charm pricing’. In the book Priceless, by William Poundstone, he dissected 8 studies on the use of charm prices and found that on average they increased sales by 24%.
Does anything perform better than the number 9?
If the number 9 performs so well, is there anything that is more powerful than these ‘charm prices’?
In an experiment by MIT and the University of Chicago, a standard women’s clothing item was tested at three separate price points: $34, $39 and $44. Researchers expected the item to sell best at the cheapest price, however, the item sold best at $39.
Why? At $39, customers assumed that the clothing item was discounted just because it ended in the number 9. This anchoring to the number 9 gave customers the impression that the $39 item was better value because of this perceived discount.
Despite this, researchers have found that although the number 9 does increase demand for and interest in a product, it’s also dependent on the market around it.
For example, if customers knew they could get the same item for $34, rather than $39, they’d probably opt for the cheaper price, despite the latter ending in a 9. This goes to show that context can sometimes trump the anchoring bias of the number 9.
Black Friday
Black Friday is a classic example of where the anchoring effect comes into play. We often see discounted prices or items grouped together to make them seem cheaper or that you’re getting the best discount.
All this discounting is doing is setting an anchored benchmark to create the illusion that you’re getting a great deal. The thing to remember is to only buy something if you need it and consider whether the cost is worth the use you’ll get out of it.
If you want to sell something, make your initial price high. People will anchor from that original price and make their decision off that.
Remember the anchoring bias can be used in many different situations on an almost daily basis such as salary negotiations, buying a car, even day to day shopping. The only real way to conquer the anchoring bias is to compare market prices and not buy frivolously.
Further Reading
The anchoring effect bias is just part of the larger series around cognitive biases and how they can influence decision making. Here are a few other biases to read:
