free special report  free weekly tips  "Ways to win" 

 

Testing covers on the Newsstand
Although some say they use intuition, go with their gut, or rely on voodoo, there is also a science to designing covers that well on the newsstand.

At the start of any discussion on newsstand testing, we first need to talk about who the newsstand reader is. We’ve learned through numerous surveys that have been done throughout the magazine publishing industry over the past two decades that the newsstand reader is someone completely different from the subscriber.
There are many people out there who will never subscribe to a magazine, but will buy copies on the newsstand. They are what we call a ‘cherry picker’: they don’t want to commit to a 12-times-a-year magazine, they want to select the issues that are interesting to them. Even if they buy three or four issues of a monthly title a year, and they spend almost the same amount of money as a subscription, that’s how they want to do it.
Because newsstand buyers are choosing an individual issue, they tend to read it more thoroughly and look at the ads more closely. They are a valuable and to a large extent, an irreplaceably resource. They are inspired obviously by the cover, and by the editorial content of a particular issue.

1. The economics of cover design
Sales can differ from one issue to another on a percentage basis. Although the sales can be affected by seasonality, we know that distribution does not change that much from one issue to the next (for better or worse). So we know covers make a difference.
How much of a difference does 1 percent in sales make in your magazine? How much of a financial difference? This is a number that every publisher, every art director, every editor should know: what difference does it make to your bottom line from one issue to the next if you can change your newsstand sales by 1 percent?
The difference will vary from publisher to publisher, based on the factors each takes into account and how each factor is weighted. In judging the true financial impact of 1 percent in sale, however, remember to include the revenue that comes from the subscriptions generated by the newsstand sales and the contribution to rate base that one twelve-time reader would have (in other words, for every one reader added to every issue, how much is that contribution to rate base worth). The resulting annualized revenue contribution creates a justification for using that money for improving the package and testing.

2. What makes readers buy?
What makes the newsstand buyer buy? Studies have shown they choose a magazine at newsstand for the following reasons:
Table of contents: 10 percent;
Name or brand: 15 percent;
Cover itself: 75 percent
Put this together, and it means that if the average newsstand buyer, the cherry picker, buys three or four issues a year of a monthly, and if this newsstand buyer is 75 percent influenced by the cover, then you should be able to change this browser from a 3-time buyer to a 4-time buyer. What difference does that make to your bottom line? That money can be put into your cover, it can be put into your package, your newsstand testing.
3. What should you test?
There’s a huge range of things to test, starting with elements of the cover. Here are a few possibilities:
Test the cover type faces –
Try testing serif vs. sans serif type. In a magazine in a computer or fitness category, what difference will it make? The answer is likely going to depend on what category you’re in and who your competition is. Art directors and consumers see serif type as more traditional, reassuring, having a more historical value. San serif is seen as younger and more modern. And if you want to make a difference in how you fit in with your competition, if you want to be seen as a younger, hipper publication, how about testing slanted lines vs. straight? We’ve seen that slanted lines increase the feeling of excitement on the cover. Most slants go in an upward direction.
Test the cover images –
You can test interior vs. exterior photos of a home, car, whatever. You can test male vs. female or people vs. equipment. Our experience is that whatever your category is, your best bet is to put that product on the cover. If it’s a music magazine, is the product the entertainer or the instruments? It really depends on what that magazine is.
If it’s a computer magazine, are you better off showing a person with a computer or without? In our experience, you are better off not showing the person at all. For a baby magazine, our experience is that the baby is the product and if you have a really cute baby on the cover, you are doing well.
Test the logo –
The logo is an important thing to test. I recommend that anyone considering a logo change should test it first. Preferably, test it repeatedly. There might be some resistance to a new logo and ‘look’ for the first few issues, and then you might find that it becomes more successful. Or it could be the opposite: you might have a wave of interest from people who are intrigued because the look is really different and that might wane later on.
Test the package –
You can test the package itself, for example, the premiums, CDs or other value-added products. When you do these kinds of tests, you have to look at the cost of creative and bagging it on, and the difference in sales.
Test the price –
Cover price testing is something we have done extensively; and there’s a lot of research available in this area. There are a number of things you need to know about cover price. One is to use both U.S. and Canadian pricing and make sure you have a test for both.

4. Isolate your test elements
When you are testing, isolate the test element. If you are testing a price or a specific premium, that should be the only element that is different from one part of the print order to the other. If you are testing a cover, make sure that only one element – the size of the cover lines, the color of the background, the specific background image – is different from one part of the print order to another.
If an entire cover concept is being tested, and too many elements are different (i.e. the cover lines, the cover image, the overall cover treatment) then you will not be able to determine which element causes the difference in sale that results.
You have to take one element. For example, if you drop the price and you put a sticker on the cover that says ‘New Low Price,’ you aren’t testing the price so much as you are testing the promotion. There’s nothing wrong with testing that – it’s great to do – just understand that it’s not a true price test. You are testing the impact of the sticker on your cover. You need to pick one element to test and test only that.

5. What should you test against?
Over the years, we have done as much as we can to refine and focus the testing and make it as stable as possible. We know that by testing this year vs. last year, you’re testing this year’s market vs. last year’s. This is true even if you’re using the same month. It’s really better to take one issue and split it.

6. Newsstand pricing trends
For years we’ve talked about a $1 up, 10 percent down rule of cover pricing, and to some degree it seems to hold true on the cover price tests we’ve been doing recently. It’s a good rule of thumb when doing your testing.
$4.99 is what we call an avalanche point – the point at which your sales drop so rapidly that your profitability goes with them. Price does have a dramatic impact on your profitability, whether you up it or down it.
The chart on this page shows an example of newsstand pricing test results where you can see that up until $4.95, your sales are dropping, but your profitability is going up. At $4.95, your sales drop so quickly that your revenues and profitability take a nosedive as well. That’s why we call it the ‘avalanche point.’
Different magazines in different categories can have different avalanche points. Overseas, different markets have different avalanche points based on their local cover prices. What we’ve also found is that if there is a premium added at the avalanche point, it changes the graph. An added premium will cause the sales to pop up instead of dropping off.

7. Value more important than price
This shows that value is more important to customers than price. And through years of testing right up until now, when things are changing so rapidly, this rule still holds true.
The average consumer would rather pay more for more value than get a price break on a product that doesn’t deliver the same value to them. So if it’s a choice of dropping your price or upping your price and adding some extra signatures, a special section or a CD ROM, it’s better to go with the value-add rather than cutting the price.

8. New price sensitivity issues
For years, we noticed it was possible to really push newsstand prices. Over time consumers would lose their price sensitivity, so they might go away when you raised the price, but they came back again and reestablished prices at the new level. Publishers began to go beyond those traditional avalanche points – to $6.99, $7.99, $8.99, $9.99.
But just within the past year, it appears there is more price resistance than I’ve seen in years of price testing on the newsstands. A year or two ago if someone had asked me about down-pricing their magazine, I would have said don’t bother. For years, nothing appeared to be gained by lowering your price. If anything, it seemed to cheapen the product.
That’s changing too. A number of magazines have tried down-pricing and it has increased their sales. We’ve gotten to a point where consumer price sensitivity is getting more apparent that it has been in a very long time. It needs continual testing, it really does.

9. Setting up your tests
The online program Decision Analysis provides a statistical model that can help in setting up price groups. Using it, you can see that for a desired confidence level of 99 percent and maximum error of only 2 percent, at an average sale of 5,000 copies, you will want to test about half your draw; at an average sale of 10,000 copies, you can reduce your test group to about 33 percent; and at an average sale of about 15,000 copies, you can test about 20 percent of the draw.
In other words, there is an inverse relationship between the size of your newsstand sale and the percentage that you must use for a test group in order to get statistically significant results.
If you up your scale to 100,000 or 200,000 you can test as little as 16 percent and still feel confident. And that’s good, because then you can get to multiple splits-test maybe three different price points.

10. Test early and often.
We always sit down with our publishers every August and say ‘What should we be testing in the next year? We don’t need to know anything right now, but think about what you or your publisher or your board of directors is going to want to know.’
You want to test, and because it takes a fairly long time, you need to test before you have to make the changes. Listen to what is happening among your ad sales people, your board, etc. If people feel you aren’t charging enough or you’re charging too much; you’re using colors that are not bright enough – test now before you have to make the change without testing.
If you are going to test something on the cover, you don’t have to tell anyone other than those responsibly for setting up your splits; and your printer, who is responsible for making sure your issues go where they are supposed to go. If it’s a price test, retailers need 60 days to get the new pricing into their systems, or they won’t be able to read your UPC codes.
Test several issues. The newsstand is volatile and people’s responses are going to change from one issue to the next. If you are testing a price, price sensitivity may go away or newsstand buyers may discover the price change the second time around. Allow yourself time to analyze.

11. Wait for the final results.
Some people say that they can have test results within five weeks of putting the issue out. The problem is the split returns might be slow. You don’t want to risk it. If you analyze based on returns flow, you may find in the end that the rate of return is different in the test group than in the control group. This could result in your jumping to a conclusion that later turns out to be incorrect.
The danger in this is that frequently the first results presented are the only ones remembered. It is very difficult to correct that initial impression that was created, even if it is counteracted by the final test analysis.
For that reason, it makes sense to take the extra few weeks to get the final test results rather than trying to draw conclusions from the early returns flow. Don’t analyze too early, and if you do track it early, don’t broadcast the results until the issues are final and you’ve really looked at them.

 

 

Contact Us

(603) 721-9668