How many postcards is a GOOD test?

[jem]

Happy Superbowl Sunday folks,

So we're trying out a few different direct mail ideas, have a bunch of different designs & copy we want to try.

I'm a huge fan of testing. Normally when split testing a landing page about 200 visitors is representative and gives a reasonable indication of how the pages are going to perform.

I was wondering if this same number holds true for postcard marketing. I suspect it gets more difficult when you also factor into account multiple touches (i.e. you need to touch somebody 7 times before they bite)...

Any thoughts/advice are appreciated

Thanks,
Simon

[greenoak]

we do postcards....but the testing idea wouldnt work very well for us...since our cards are totally related to a big event in the store....and we couldnt produce the event just for test purposes....
i guess i feel like they come because of the quality of the event more than the quality of the postcard.... and that would mean a great postcard couldnt make a great event...
i like your idea of 7 touches....
ann....

[orion_joel]

In essence you can take a test to however many people you want to, but the question is will that be truly representative of the market. Additionally can you test an idea, to the large enough number to get that representative number, where you are testing on people with similar decision making patterns.

I would suspect that you may be able to get some reasonable results if you had groups of 100 that were targeted prospects, eg in the market for your type of product. Where as if you were to even just take 1000 random people from the phone book you may not get as good results. So really it depends on your product, who the test is going to be on and even a few other factors.

The thing that i kind of disagree with is the 7 touches. This theory i have seen a lot, the number sometimes changes up or down. But most commonly i have seen it in the advertising industry for magazines and newspaper's pushing the fact that people need to see your ad X number of times before it registers with. My believe is that this is a marketing tactic, which maybe has some truth in it, but more so to tell longer term advertising spots in their publications.

If you are going direct mail, eg postcards, you will find that the number of times you need to get in front of the client is less. If you have the right offer, and people want the product you only need one touch. I think this is something that Paul would definetly be able to give a solid answer to so hopefully he drops by the thread at some point.

[Steve B]

Good question Simon.

I had the same question about 18 months ago and had trouble finding an answer. There are tons of advertising people repeating general theories, but I could find very little statistical information that seemed to be credible. I finally stumbled on ONE article that was very helpful that was more academic in nature and focused on the science of the statistics (of course, I can't put my hands on it now). The answer is likely going to be over a thousand. But, it depends on the success percentage that you need to get in order for the mailing to be considered a "success". I forget the exact formula, but I can tell you for my product it was close to 3,000. I think my success target was a little over 1%. If I needed a 3% success rate for the mailing to be successful, the required sample size would have been lower.

B.T.W. - Joel, I agree with you about your thoughts on the 7 touches being a marketing tactic. I can tell you, based on my experience as an advertisor that I usually did better the FIRST time I am in a publication. We also publish a coupon magazine and several of our advertisors reported this also. Of course, with our magazine that we publish we push for longer contracts and I often found myself telling prospects the same thing that all the advertisors were telling me that you need to repeat. It just didn't prove to be true for my dog fence business (and several others).

[jem]

Thanks for the replies.

The multiple touch theory (not sure it is seven - have seen btwn 3-9) is something I have seen in Dan Kennedy's work - he proposes a 3 step letter system. He claims to have stats to show that you get xx% of the replies on the 1st mailing yy% on the 2nd etc. You don't get them if you don't do multiple mailings.

Having never done this I cannot speak 1st hand but he comes highly recommended...who knows!!!!!

I have a large consistent list (say 20k small businesses with websites in my state) so want to get my message right. happy to test 10 x 100 or 5 x 200 (hopefully around 1000 but thats where I need guidance) then start doing bigger blasts (1000+ at a time)

Appreciate the replies folks - thanks

Simon

[orion_joel]

Steve - i think that the X number touches does have some credibility, however it depends on the target market and the product. For example someone selling cowboy boots, may advertise, they may do well on the first run, for two reasons one the ad is different to what they see every other time they read, more so if the ad is prominent. The second reason which will apply no matter what is that the people looking for this type of product, will in many chances see the ad, not because it is there but because sub consciously they are looking for it.

Simon - I would suggest that just getting in and doing a test on what you believe is a reasonable sample, if you have 10 different layouts or sets of copy to try do 10 x 100, if not do 5 x 200. On a list your size, i think should should be able to get a reasonable idea on which one to try on a large sample from 100 or 200. You could make it a learning experience which you can create a product or service from later on. Eg How to best market using postcards, a theory behind testing.

[greenoak]

i believe in 7 touches.... and spend hundreds on the idea every year.....
..running ads for weeks, not just once...
i dont quite see how a single ad would work for me... the customer might not need me till a few months down the road and my then wouldnt they have forgotten all about most single ads...
ann

[Steve B]

Simon you are wise to determine the proper size of the test versus just picking whatever number you think might be enough. If you are testing a number that is not statistically significant - then you run the risk of making decisions on chance occurrences. This defeats the whole purpose of testing since your decisions will be no better than if you just guessed at which one would work the best. Actually, it can be worse than just guessing because you could be led down the wrong path and be thinking that you have "proof" that you have chosen the best one.

For instance, if the people who first thought of selling bottled water would have asked me and 10 or 20 other people (that just happen to think like me) about the idea, they would have concluded that it was the dumbest idea in the world. They would have dropped the idea and done something else. But, if they tested it on 10,000 people they would have gotten a much more accurate sample of the entire population and would have realized there are enough people that would pay for bottled water that they can make millions! They might have gotten lucky and made the right decision if they only asked 10 people who happen to be the type that would pay for bottled water, but the odds of them getting the true picture increases as their sample size goes up.

Direct mail marketing is a science as much as it is an art.

[KristineS]

Regarding the "seven touches" idea you might want to keep in mind that advertising and direct mail are two different animals. Running an advertisement multiple times is a good idea, it helps set your message in the minds of viewers that see it every month or every time they read the magazine.

Direct mail is a different idea, in this case you're sending your message out directly to someone who may or may not be interested. If they're interested you'll probably hook them the first time through. If they're not interested, all the subsequent postcards will do is annoy them. Since many people don't pay a lot of attention to DM pieces, it may not annoy them enough to make them remember your business, but it might, and if they do, it will be in a negative way.

Standard return on a Direct Mail piece is 1 or 2 percent. If you get 5 percent you've hit the lottery. I've not seen any research that shows that multiple mailings increase that rate. I'll do some searching and see if I can find any information.

[jem]

At this point I'm leaning towards 5 batches of 200. I just 'feel' 100 ain't enough to get an idea (of course this is just intuition!)...

I'm going to give them a shot over the next few weeks - I'll let everybody know how it goes....

Simon

[Steve B]

Standard return on a Direct Mail piece is 1 or 2 percent.

Kristine, is the above based on your experience? I've heard that figure so often, but it seems to be that everyone is re-quoting information that is already out there. I'm hoping you may have specific experience with a direct mail campaign.

[KristineS]

We have done direct mail, and I'd say that number is pretty accurate. Having a targeted list of qualified prospects can amp that number up a bit, but if you get a 1% return you're doing well.

Direct mail is a tough nut to crack. We've tried postcards and letters and they seem to generate about the same rate of return. One thing we haven't done is purchase a mailing list, all our lists have either been generated from trade shows or our own research. It is possible a purchased mailing list might have some impact on the rate of return, but I don't think it would have much of an effect.

[Steve B]

The best I've ever done was 1/4 of 1%. I know it depends quite a bit on what you're selling. My product/service costs a min. of $850 - so my success would be lower than someone selling $25 steak knives. I did purchase a very targeted list too.

If I could ever find a formula that generated 1% success - I wouldn't have to do any other advertising.

[orion_joel]

While, i am disagreeing concept to the 7 touches theory, what i am really disagreeing with is the idea that people need to see something 7 times to act on it or to register it.

What i agree entirely with is that you do need to consistently advertise. However i do not believe that the actual 7 touches comes into play a huge amount. You can advertise once a week every week for a year, i would expect that a majority of the people that respond to that ad will be seeing it for the first time, of course there will be some people that take their time in responding and may see it 5, 7, 10 ,or even 50 times but this is not necessarily based on the theory of seeing the ad 7 times. I can see an ad 5 times and do nothing until the 6th time i see it because i don't need it till that last time. Which again is another huge factor in the 7 touches theory, if people do not need what is being sold they will not act on the ad.

[Steve B]

I found the explanation I was referring to earlier. If you feel like reading a little bit of an academic statistical explanation, you can follow the link below. It explains what I was trying to say very well. If you don't feel like reading it - the bottom line is that you will have to think in terms of thousands, not hundreds to get a valid test. The exact number will change based on some variables specific to your marketing effort.

Scientific testing - sample size calculations for marketing tests (http://www.lucidview.com/sample_size.htm)

[jem]

Steve,

Many thanks for that....

Put's a damper on my 5 batches of 200!!!

Even though it's probably not statistically valid I think we will still give it a shot. Better to be trying something than waiting for the perfect scenario (big believer in ACTION)

Wish me luck!

Simon

[cocoy]

The most recent one I did was 200. Got a 2% reply and from that 1% actual customers. This was about 1 month ago. These were also targeted.

[Steve B]

Nothing wrong with trying a smaller number, just don't draw any firm conclusions from your results (either good or bad).

[Just H]

It's been several years since my marketing class but I distinctly remember learning that a 1% return on a direct mail piece was about as good as could be expected, generally anyway. As far as local newspaper ads, where the same group of individuals reads through consistently, I've heard it's about 10 viewings before they even recognize the ad as something they've seen before. Seems pretty discouraging but definitely makes comparing the costs vital. For instance, cost of the direct mail piece (from design to printing to addressing and mailing costs) vs. the cost of an ad for, say, every Saturday ten weeks straight. And I've heard the consistency in the size and placement of ad are more important to the recognition than the size of the ad by itself.

Doing one of each of these would be a great experiment if you have the marketing money to research what actually works well. Most clients want to include some identifiable piece of info on each (like a promotional code or certain offer) to best measure from which the client made the contact.

My client that does mailings most consistently is in real estate and generally sends out 200 pieces from a list she has compiled, which includes buyers that have previously contacted her, sellers that she has represented and all the residents in the current subdivision she's targeting. This lets the previous sellers and buyers keep her in mind for their next move and to have her foremost in their minds when a friend or neighbor brings up selling/buying. It also lets everyone in the area know that she sold a certain property on their street/block/etc and what she sold it for, so if they might have been thinking of putting their place on the market or have friends who want to move into the neighborhood. Again, to be the one that comes to mind with proven experience in that area.

[orion_joel]

Jem, i would imagine that on a text of 200 you may only need a 1% conversion to actual customer to recover your cost. However guess this depends on the costs of the postcards, though.

Even if the result's are not statistically accurate, you should be able to gauge i would imagine your customer responses to the marketing. For example you may get 2 or 3 customers from one trial call up really ready to get in and buy right now, where another may get 5 or 6 who call up are hesitant but kind of do want to buy. Whatever reaction i would expect you will be able to judge which one to try again on a larger, sample.

[billbenson]

Steve B, I would think your market is a bit different. You might have people that think it's cruelty. You will also have people in deed restricted areas that have a problem with dogs getting out and can't put up a fence.

My postal carrier is also terrified of dogs. I have a boxer and a yorkie. Funny thing is the yorkie is the one that would bite her. Boxer is friendly as can be. In any case, maybe talking to your mail carrier or doing a seminar for the local postal service would get you some good leads? Dunno?

If I were to want an electric pet fence, I'd just type it into google and see who is out there. It's almost a presold item? Particularly because you are regional.

[Vivid Color Zack]

Keep in mind that with a larger campaign your postage is going to drop enough to offset the cost of doing more cards. The postage saved in bulk rates is pretty close to half.

Edit: That is just for 100 vs 200, don't expect postage to drop by half every time you double quantity. Sorry if that was implied.

In CA you can't even get a discount on postage for less than 200 pieces. I'm pretty sure that's a federal regulation actually. Bulk rates will drop even more if you increase your quantity. Do you happen to be a non-profit? Rates go down A LOT for non-profit bulk mail.

I think you should try to go bigger for your first try. I would rather get exposure to 1000 people than to 200 people 5 times. Like Kristine said, if they don't want your services the first time, they're most likely not going to need it the second third fourth and fifth times.

This of course has exceptions based on industry (what is yours again?) like carpet cleaning or auto repair where you rarely think about it unless you need it right then.

Are you running a promotion of any kind? How is your postcard going to grab the potential new customer?

[huggytree]

i tried the post card idea...i sent out a 10% off card to everyone in the area who had just moved in....300 mailed out and not 1 call....that was 7 months ago, so not even 1%

i send fliers to every contractor 3x a year....over & over & over & over for 2 years now.. that works great....everyone new always says - you send them over and over and i decided to give you a try....repetition is important because they may not need you today, but next time they might....you have to hit them exactly when they need you.

im a small 1 man plumbing shop with over 100 other competitors....everyone knows my company name...repetition is definately the key...a 1 time post card wont work well in my opinion.

[Business Attorney]

This of course has exceptions based on industry (what is yours again?) like carpet cleaning or auto repair where you rarely think about it unless you need it right then.

The thing is, the "exception" covers a great many businesses. Lawyers, doctors, dentists, plumbers, roofers, furnace repair, the list goes on. Sure, some people may take action before a critical need arises, but they will only respond if a need exists at some level. Since most people promptly throw out unsolicited mail, they are not likely to have the post card someone mailed to them 6 months ago, nor are they likely to remember a single post card among the tons of catalogs and credit card solicitations.

I'm not a marketing professional myself, but I am much more likely to buy a product or service from someone who has solicited my business over a period of time rather than pick the one that happens to show up serendipitously at the time a particular need arises. That doesn't mean I will always reject the newcomer, but that business will have to beat the other on some basis.

I don't know whether the "7 touches" holds water, but for me, it makes sense.

[Just H]

It seems you also have to take into account the market and your mailing target for whether to use the bulk postage or not. My client who does the most postcard mailings was using the bulk rate through her company but wasn't getting any response at all. Since her market is somewhat narrowed by area and previous clients, she now uses a regular postcard stamp - she says it portrays her as more "someone they know and can trust" than any local office who "just wants the business".

I really don't believe this applies across occupations and offers though. I'm trying to recall back to all the "ad mail" or I actually thought of it first as "junk mail" (which is telling in itself), and I do often glance through to see if there's something I'm in need of. The postage itself is not necessarily a factor for me though the pieces in envelopes that I can't see right off is determined as "throw-out" mostly by the postage and how it's addressed. I've read somewhere before that up to 75% of the bulk postage stuff is immediately thrown out as junk mail. Let me see if I can find that article, looks like it's here (http://www.freshpromotions.com.au/direct-mail-promotion.html).

Appears it'ss to review the research, be aware of your target market and try a couple things to see what brings the best response.

[Paul Elliott]

Touches?

The "7 touches" has been misinterpreted and often misapplied. It comes from sales contact studies. The average sale is made after 7-9 contacts by the sales person (also quoted as low as 5 and as high as 11). 80% of sales people never make the second contact.

A contact is face-to-face, phone, mail, email, etc., and is best done with a combination of all the above. These figures are from a presentation I heard about 10 years ago.

One of the most effective contacts that is rarely done, is the handwritten thank you note after ever face-to-face visit.

Live postage almost always produces better responses.

For a TV ad campaign to get its message through for the viewer to take action, it takes 27 times--from a marketing seminar about 3 years ago.

Thanks for the formula. Did you run your numbers through it?

In years past when I was doing a lot of direct mail, we found that 1,000 to 3,000 pieces was needed to get reliable statistics.

Is 1-3% good for direct mail? Likely, nyet! Though not necessarily.

I have seen several mass mailings that pulled over 15%.

What is the theoretical maximum pull? 100%? Again, nyet! :confused:

Now, this will put everything in doubt!

I did a mailing of 250 to a hungry, well-qualified list ... Pull? 120%!

I was stunned! My baby blues snapped w-i-d-e open! Our problem immediately turned into a logistical nightmare of fulfillment. Nice problem to have!

Oh, yeah ... how did it happen? The recipients called, faxed, and emailed their friends who immediately begged to be included. It was a massive financial success--each person who took up the offer was worth $8,000 to $12,000 and the conversion was 60%!

NEVER be impressed with a small or large percentage pull ... analyze more deeply. It all depends on the product being sold and the cost of the mailing.

In the above example, the mailing cost about $2.50 each. However, who argue with spending $250 on such a mailing. :D :D

With some mailings, 1/4% might be a massive success. Certainly, a 1/2% in the above example would have been.

How many mailings should one do?

Drew A. Kaplan found that he could mail the identical DAK catalog of electronic equipment three times to get the maximum results. He did find that a different cover on the catalog helped a little, but his yield peaked between the 2nd and 3rd mailings.

Jay Abraham found the same results and usually recommends mailing the same package to the same list.

As with many other things, a more careful and probing approach is necessary to make sense out of what one is planning or the reported results.

Paul

[jem]

Paul - great post.

I am an 'early' student of direct mail and from what I can glean there are MANY rules of thumb, urban myths and what not. It seems the real answer is to test ideas that HAVE been proven (and maybe some that have not!) and see what works for your product/list.

TEST, TRACK, TEST, TRACK etc

I have not been following up (enough) in my opinion - so need to try that.

Does anybody have a view how long between mailings is a good place to start? 5 days, 7 days, 10 days?

Many thanks - great info here!

Simon

[Paul Elliott]

TEST, TRACK, TEST, TRACK etc

That's the only way to go. Test EVERYTHING!

It needs to be done carefully AND persistently.


Does anybody have a view how long between mailings is a good place to start? 5 days, 7 days, 10 days?

It depends on your program and the urgency you have/intend to create. On most of our programs, we mailed every 3-4 weeks. 80% of your response will be complete within 2 weeks of when the piece "hits"--lands in the recipient's mailbox. Therefore, it may not pay to "drop"--when your piece goes in the mail--inside of 3 weeks.

When you do mail the same item to the same list multiple times, purge your list of all who have responded before mailing again, otherwise, you will look foolish to those who have already responded.

If you're mailing an envelope, put a teaser headline on the outside of the envelope. I encourage you to put your logo along with your return address--something that is recognizable--on the outside of the envelope.

If you're mailing a postcard, put a headline on front and another one on the back.

Again, use your logo. You want to brand yourself.

Paul

[jem]

Paul,

Many thanks - that is very helpful!

Simon

[Steve B]

I agree with everything Paul said except one thing. Putting a teaser on the envelope seems to be so overdone and it just SCREAMS that the envelope contains "junk mail". If you want them to open the envelope, I wouldn't do that.

I don't have any data to back this up - so Paul may be right, but I know how I react to those teasers on the envelopes. I see it as a challenge to toss the envelope without finding out more.

[Vivid Color Zack]

^ I base a lot of my design ideas on how I personally react. Not very scientific, in fact it's probably restricting my views a lot. If I toss every mailer I see that has an explosion symbol on the front with "DEAL OF A LIFETIME" or "LEAVE YOUR DEBTS BEHIND" I assume I'm not the only one doing this.

[Paul Elliott]

I agree with everything Paul said except one thing. Putting a teaser on the envelope seems to be so overdone and it just SCREAMS that the envelope contains "junk mail". If you want them to open the envelope, I wouldn't do that.

I don't have any data to back this up - so Paul may be right, but I know how I react to those teasers on the envelopes. I see it as a challenge to toss the envelope without finding out more.

Steve, I didn't test my last one--18 months ago--but got a 4% response rate. In years past when I did split test, a well-crafted envelope tease vs. none made a difference from 2% vs. 0.9% up to 12% vs. 1.2%.

I've never had one that was = or lower on the tease.

I believe the quality of the headline/tease is very important. Keep in mind that the headline/tease has only 1 goal--to get the reader to read further. Do it well. Avoid the things that look from the start like a scam.

You are usually better with a question that can't be answered "yes" or "no."

One classic that sold millions of units was, "Do You Make These Mistakes In English?" Of course, you have to read more to give an answer. Clever, huh?

Paul

[Paul Elliott]

^ I base a lot of my design ideas on how I personally react. Not very scientific, in fact it's probably restricting my views a lot. If I toss every mailer I see that has an explosion symbol on the front with "DEAL OF A LIFETIME" or "LEAVE YOUR DEBTS BEHIND" I assume I'm not the only one doing this.

Zack, you're right it isn't scientific at all. While you are certainly NOT the only one who pitches them, you nor I have ANY idea how many do. This is why testing is SO important.

While your response--gut feeling--is a place to start, NEVER continue to do a piece--more than once--without testing SOMEthing. Otherwise, you must assume you are simply wasting money.

Paul

[Steve B]

Thanks Paul. That goes to show you I'm not normal.

I also would have laughed at anyone that wanted me to invest in bottled water based on what I think of the idea.

[Paul Elliott]

Thanks Paul. That goes to show you I'm not normal.

I don't accept that, Steve. You may be supernormal. :D

Paul