Put Your Copy To The Test

In my cri­tique con­sult­ing ser­vices, I’ve found the most com­mon mis­take clients make is the igno­rance of, or indif­fer­ence to, split-​​testing. Lit­tle do they know this under­uti­lized mar­ket­ing prac­tice can be one of the most profitable.

Split-​​testing is the sim­ple process of split­ting your audi­ence into read­ing two or more ver­sions of your copy — whether it’s your web­site, saleslet­ter, email, etc — to deter­mine which ver­sion pulls the best.

I’m a big fan of test­ing. And I teach it as much as I can, and if my cri­tiques are any indi­ca­tion, I also believe 99% of mar­keters out there do not test at all. It’s a shame, because they’re leav­ing so much money on the table.

I applaud those peo­ple who do test. The prob­lem, how­ever, is that some of them reveal their test results, share their insights, and make claims, which can be a bad thing.

Sure, test results are cool. Per­haps even insight­ful. But some may be mis­in­ter­preted, and doing so also feeds into this lack­adaisi­cal mind­set of those non-​​testers, think­ing that such test results are uni­ver­sally applic­a­ble and there­fore test­ing is unnecessary.

Tak­ing any test result as gospel, with­out val­i­da­tion, can be an expen­sive proposition…

I always rec­om­mend you test every­thing first. Never take any test result as applic­a­ble to your spe­cific sit­u­a­tion. Even mine, as I do share my test results from time to time.

Why? Because only a hand­ful of these test results, if any (and believe me, they are rare), are sta­tis­ti­cally sig­nif­i­cant enough to be truly portable to other offers or markets.

So, you’re prob­a­bly scratch­ing your head, wondering:

“What good are test results, then?”

Now, I agree that some results, tested thor­oughly under var­i­ous con­di­tions, can be widely accepted. They have slim mar­gins of error, and they’re sta­tis­ti­cally valid enough that the like­li­hood of achiev­ing the same kinds of results in another area is higher.

(Then again, those “sure­fire” test results are equally lim­ited in their appli­ca­tion, as they were gen­er­ated at spe­cific points in time. They may change over time as mar­kets evolve. So even the best test results are never constant.)

It’s true, how­ever, that hav­ing access to some­one else’s test results — like the ones I reveal in my Copy Doc­tor mem­ber­ship web­site — puts you sev­eral steps ahead of try­ing to blindly test any­thing with­out any help or direc­tion, let alone of test­ing noth­ing at all.

Know­ing what to test is an awe­some benefit.

Never dis­count other people’s test results. Look at them as point­ers, which are great ideas for test­ing. But they are not, and should never be taken as, absolutes.

Each “pointer” any mar­keter gives has been achieved under very spe­cific and some­times very unique con­di­tions. Dupli­cat­ing those results is highly con­tin­gent on dupli­cat­ing the same con­di­tions under which those tests were conducted.

For exam­ple, if you’re test­ing col­ors, then the tar­get mar­ket, the type of prod­uct, the price point, and the offer have all played immense roles in the weight of those test results. And even if there’s a small chance that they are the same as yours, the time dur­ing which the tests were con­ducted is def­i­nitely different.

Let me give you some examples.

Last year, another copy­writer posted his test results on a pop­u­lar Inter­net mar­ket­ing forum. I agree with most of them, largely because they seem to par­al­lel mine, but there were a few excep­tions. So I’ve added some of my com­ments, reprinted here…

1. Your head­line should always be in Tahoma, dark red, size 20. It must also have quo­ta­tion marks before and after. Also, head­lines that start with “Who Else” always gets a higher response.

Yes, that’s what I’ve tested, too. But the results do vary. On aver­age, the increase is less than 100%. Even less than 50%. (Mean­ing, they don’t dou­ble or more, as some mar­keters sug­gest.) And in some cases, the response went down rather than up.

In the major­ity of tests, we found the great­est bump in results has less to do with col­ors than it has with the headline’s dis­tinc­tive appeal. In other words, when the head­line is dis­tinctly dif­fer­ent than the rest of the body copy, it draws more attention.

For instance, sans-​​serif fonts for head­lines and head­ers (i.e., sub­heads within the copy), when the body copy uses serif fonts — and vice versa — have the best results. Dark red head­lines, and then dark blue, seem to out­pull other colors.

Sans-​​serif are pro­por­tional width fonts with no curlicues, such as Tahoma, Tre­buchet, Ver­dana, Arial, Hel­vetica, Geneva, etc — this blog uses Arial, for exam­ple. Serif fonts have curlicues, such as Times New Roman, School­book, Geor­gia, Gara­mond, etc.

Serif fonts are best with a pri­mar­ily techno­pho­bic audience.

(Techno­phobes are peo­ple who are not tech­ni­cally inclined, such as new­bies or peo­ple who are averse to tech­nol­ogy. Granted, as more and more peo­ple use tech­nol­ogy on a reg­u­lar basis, techno­phobes are a shrink­ing bunch.)

With technophiles, on the other hand (i.e., peo­ple who love and use tech­nol­ogy a lot), I’ve had bet­ter results with sans-​​serif fonts used within the body copy, and even sans-​​serif fonts for the main head­line and head­ers through­out the copy.

My guess as to why all of this is so? For techno­phobes, serif fonts mimic direct mail and offline com­mu­ni­ca­tions, which non-​​techies are used to. Objec­tively, deliv­er­ing copy in the for­mat your audi­ence is used to as much as pos­si­ble is the key.

The mid­dle ground seems to be fixed-​​width fonts, such as Courier or Courier New. If your audi­ence is com­prised of both ends of the spec­trum of what Jupiter Research calls “techno­graph­ics” — sim­i­lar to demo­graph­ics — Courier is your safest bet.

My the­ory is, Courier mim­ics plain text email for technophiles and type­writer type for techno­phobes, thus appeal­ing to both. So if you’re not sure of your tar­get market’s techno­graph­ics, use Courier or Courier New as your main font to play it safe.

Bot­tom line, it depends on the audience.

And with­out a doubt, doing proper mar­ket research before you do any split-​​testing — or worse, guess­ing from the get-​​go — will tell you a lot about your audi­ence as to what will appeal to them best. (I’ll come back to the whole “who else” argu­ment a lit­tle later.)

2. Your head­line should be sim­ple and focus only on the SINGLE biggest ben­e­fit of your prod­uct (don’t try to get fancy and do heaps of ben­e­fits at once). Make it clear and compelling.

True. But again, that’s a sug­ges­tion. In fact, I’ve found that using less and less ben­e­fits in your head­line increases response, likely because a benefit-​​driven head­line alerts the reader that what fol­lows is a saleslet­ter. (And peo­ple hate being sold.)

I’ve had good results with head­lines that didn’t have any ben­e­fits at all. Such as…

• Newsy, editorial-​​like headlines,
• Head­lines that intro­duce a story,
• Head­lines that ask a question,
• Head­lines that give an incom­plete idea,
• Head­lines that start a conversation,

And so on.

These head­lines pull the reader into the copy, espe­cially because they start an idea, imply a ben­e­fit (rather than state one out­right), pique the reader’s curios­ity, cre­ate doubt or intrigue, or con­tinue the con­ver­sa­tion going on in the mind of the reader.

Here are some exam­ples of clas­sic head­lines that are not benefit-​​driven but have his­tor­i­cally proven to be for some of the most prof­itable ads in history:

• “Do You Make These Mis­takes In English?”
• “The Tale Of Two Young Men”
• “The Insult That Made A Man Out Of Mac”
• “When Doc­tors Feel Rot­ten, This Is What They Do”
• “Often A Brides­maid, Never A Bride”
• “Using A Lawyer May Be Dan­ger­ous To Your Wealth”
• “Have You Ever Seen a Grown Man Cry?”

Again, there are no real or direct ben­e­fits in these head­lines. But there’s either an implied ben­e­fit, or a state­ment that tick­les people’s curios­ity and forces them to start read­ing, which is the whole pur­pose of a head­line in the first place.

Now, some peo­ple have ranted and raved about the “who else” head­line. I tested this head­line like crazy. Per­son­ally, I hate it. But like it or not, it seems to pull more than any other head­line I’ve tried — granted, it is slowly declin­ing in recent times.

In fact, while I’m at it let me share with you five head­line types I’ve tested, which have pro­duced some of the high­est responses — and the rea­sons I believe they work…

1. “Who Else”

The “who else” head­line is based on the orig­i­nal clas­sic, “Who Else Wants a Screen Star Fig­ure?” This type of head­line is the win­ner is most of my tests, and now used by count­less mar­keters. I’m sure you’ve come across at least one of them.

Mind you, the sec­ond win­ner in line is not too far down. So “Who else” may be the win­ner but the mar­gin is slim. Plus, recent tests show that the response for a “who else” head­line is declin­ing, likely due to is overuse.

In other words, just like the benefit-​​driven head­line, too many mar­keters use “who else” nowa­days, and there­fore peo­ple are becom­ing more aware of it. It screams “saleslet­ter!” and thus scares read­ers off when they see it being used.

2. “If/​Then”

This head­line is very close to “who else” in terms of concept.

It was used in one of the most suc­cess­ful ad cam­paigns for self-​​made multi-​​millionaire Charles Givens, writ­ten by one of my favorite copy­writ­ers, Gary Ben­civenga.

Since then, it has been used in many other indus­tries with great suc­cess. For exam­ple, it also worked in an ad for exer­cise guru Richard Sim­mons, which basi­cally said some­thing like, “If you give me 20 min­utes a month, then I guar­an­tee you will lose weight.”

I’ve also used this type of head­line with John Reese’s Traf­fic Secrets, the saleslet­ter that sold over a mil­lion dol­lars worth of prod­uct in under one day. It started as, “If you can copy and paste, then you have what it takes to…”

Sim­ply, it promises a ben­e­fit but only if the reader can meet a very sim­ple, very obvi­ous con­di­tion. In other words, “If you meet this con­di­tion, then I will make this promise.”

The key is that the con­di­tion must be easy to meet. If the con­di­tion is too hard, it defeats the pur­pose of using such a head­line. Also, if your tar­get cus­tomer can eas­ily meet the con­di­tion, this type of head­line can also act as an effec­tive qualifier.

3. “Give Me/​And I’ll”

This is close to the “if/​then” head­line. Essen­tially, rather than ask­ing if the reader meets a con­di­tion, it makes a promise or offers a ben­e­fit when the reader meets it.

In other words, it doesn’t put any con­di­tion on the reader but rather asks a small favor from her, which is a sim­ple request that’s obvi­ously easy to do, and dis­pro­por­tion­ate to the value of the ben­e­fit they receive in return for complying.

It goes some­thing like this: “Give me some­thing (i.e., do some­thing or meet this con­di­tion), and in return I’ll make you this promise (i.e., you will enjoy this benefit).”

This high­lights a clas­sic Cial­dini prin­ci­ple, the Prin­ci­ple of Reci­procity. Dr. Robert Cial­dini, in his famous book “Influ­ence: The Psy­chol­ogy of Per­sua­sion,” states that when you do some­one a favor, peo­ple will feel oblig­ated to return it.

In this case, the head­line is ask­ing for a favor and promises to return it.

Take a look at Alex Mandossian’s Traf­fic Con­ver­sion Secrets course, which was a tele­sem­i­nar series deliv­ered over eight months. The head­line goes: “Give Me 8 Months, And I’ll Spoon­feed You My Tested Meth­ods For Con­vert­ing More Vis­i­tors Into Cash.”

4. “How To”

The “how-​​to” head­line is self-​​explanatory. Any head­line that says “how to [achieve a result]” has been a sure­fire win­ner in many cases. I use it all the time. And the rea­son is, the head­line makes the saleslet­ter look more like an arti­cle or editorial.

Regard­less of whichever head­line method you choose, your saleslet­ter should always pro­vide some infor­ma­tive or edu­ca­tional con­tent. But the goal is to deliver the “what” and to sell them on the “how,” which is only achieved by buy­ing your product.

For exam­ple…

• “How to Trans­form Your Unpro­duc­tive Web­site Into a Rag­ing Cash Machine.”
• “How to Grow Big­ger, Plumper, Juicier Toma­toes in Only Two-​​Thirds of The Time.”
• “How to Get a Near-​​New Car For 60% Less at Gov­ern­ment Seizure Auctions.”

5. Third-​​Person Verbs

Head­lines that start with a verb have always pro­duced really good results.

Verbs that direct the audi­ence to do some­thing, often to receive a ben­e­fit or to achieve a cer­tain result, have always given great test results. Verbs like “Dis­cover,” “Mas­ter,” “Access,” “Get,” “Learn,” “Con­quer,” “Build,” and so on work well.

But here’s the kicker. In recent tests, I’ve learned that, in some instances, putting your verbs in the third per­son increases response. It’s an idea I got from copy­writer Bob Bly.

He tells the story of an ad that had a typo. The ad was for a course on how to play the piano. The intended head­line was, “Put Music Back Into Your Life.” The typo was an extra “s” acci­den­tally placed after the verb “put,” as in “Puts Music Back Into Your Life.”

At first, the error was dis­con­cert­ing to the ad owner.

But to their sur­prise, they real­ized that error mul­ti­plied the ad’s response. My the­ory? The third per­son makes the head­line appear as if the prod­uct or ser­vice does the work for you rather than you doing it your­self. It seems effortless.

In Bly’s exam­ple, rather than ask­ing the reader to “put music back into her life,” which requires effort, the ad implied that the course some­how does it for them.

Again, it goes to the nat­ural human propen­sity of get­ting things done. It’s all about con­ve­nience and lazi­ness. And after I heard about it, I’ve tested it with great results.

My friend Armand Morin tested this, too, with his direc­tory gen­er­a­tor soft­ware. The head­line says, (this soft­ware) “Instantly Cre­ates a New Breed of Web­site…” rather than “Instantly Cre­ate a New Breed of Web­site” (with this software).

3. Add audio directly under­neath your head­line. But not just any audio. You want audio that excites your reader and keeps them read­ing. Don’t waf­fle on. Keep it short and sweet.

I agree. This is a test I made based on a sug­ges­tion from my friend John Reese. John used it with his Auc​tion​Se​crets​.com web­site, and told me his sales increased sig­nif­i­cantly. So I tried it. And after test­ing it got a 44% increase in response.

It is par­tic­u­larly effec­tive when com­bined with a “who else” head­line. The ques­tion in the head­line, using “who else” (or any other ques­tion, for that mat­ter), imme­di­ately engages the reader — at least to lis­ten to the audio, since the audio answers the question.

Used in con­cert with the “If/​Then” and “Give Me/​And I’ll” head­lines, the audio fin­ishes the state­ment, answers the ques­tion, or offers a ben­e­fit when they meet the condition.

After it asks you, “do you meet this con­di­tion?” the head­line directs the reader to lis­ten to the audio for a very spe­cial mes­sage if they do meet it. In other words, the audio says, “I can make this promise (or you can enjoy this ben­e­fit),” or “you should read this copy which is meant specif­i­cally for you (because you meet this con­di­tion),” etc.

4. Pep­per your order links through­out your copy (not just at the end like most peo­ple). Repeat them again and again before the end.

Now, this is a bit contentious.

I found that it depends on the prod­uct. What I dis­cov­ered was that remov­ing all links and leav­ing only the one sin­gle order link increased response dra­mat­i­cally, espe­cially for long copy, one-​​time pur­chases — espe­cially with infor­ma­tion products.

Why? Because more often than not, adding too many links gives peo­ple an oppor­tu­nity to pro­cras­ti­nate and bail out. If you have long copy, and since peo­ple never read long copy at first, they will then scan your copy to jus­tify the need to read it in the first place.

Often, they will want to know the price, not to dis­cover if there’s any value in the offer — besides, how can they know when they haven’t read it yet? — but to dis­cover if the price is appeal­ing enough to jus­tify the time it will take to read the long copy.

Increas­ing the num­ber of links adds “eye grav­ity,” and gives peo­ple a chance to skip the long copy, check out the price before they start read­ing, and then… leave.

On the other hand, if you sell an inex­pen­sive prod­uct, a com­mod­ity prod­uct, or a dig­i­tal prod­uct, if your audi­ence is on a list to which you can follow-​​up, and if the copy is highly tar­geted to an audi­ence that’s already pre-​​sold (i.e., it’s gen­er­ated through affil­i­ate pro­mo­tions that endorse the prod­uct), then many links do increase sales.

Because most of the time, peo­ple are already sold. They just want to buy.

Armand Morin has tested this exten­sively — in fact, some of his saleslet­ters have over 100 order links! At first, when I heard he saw a jump in sales because of adding more order links, my assump­tion was that this was true for all products.

It’s not.

Remem­ber, Armand uses affil­i­ate mar­ket­ing, a “namesqueeze” process (i.e., an opt-​​in page before hit­ting the salespage, so they join a list whether they buy or not), and a care­fully crafted follow-​​up autore­spon­der series that keeps pre-​​selling the recipient.

So when they’ve decided to order, they’re sim­ply led to the sales copy — which is a web­page with many order links. And in this case, doing so boosted sales.

Oth­er­wise, if you sell an expen­sive prod­uct and you really want peo­ple to read your copy first, then don’t put too many links. You want to give the least amount of dis­trac­tions as pos­si­ble as well as the least amount of oppor­tu­ni­ties to bail out as possible.

John Reese said it this way:

“Use only one link, mainly in your call-​​to-​​action sec­tion. And if you want to add more, sim­ply add some after that sec­tion, which are not order links but sim­ply anchored book­marks that jump peo­ple back to the call-​​to-​​action section.”

(The lat­ter part of John’s tip is pre­sum­ably because, when peo­ple scan, they usu­ally pay atten­tion to the top and the bot­tom. Which is why head­lines and P.S.‘s at the end are cru­cial. But in this case, adding “live” order links after the order sec­tion is just another oppor­tu­nity for peo­ple to bail out. That’s my the­ory, anyway.)

5. Pep­per tes­ti­mo­ni­als through­out your copy. Don’t limit your tes­ti­mo­ni­als to blocks some­where in the mid­dle. Use the first one early on and pep­per the oth­ers through­out your copy.

My answer to this one is both “yes” and “no.” Yes, you should pep­per tes­ti­mo­ni­als through­out. But adding a few together, par­tic­u­larly if they’re sim­i­lar and serve a spe­cific pur­pose, also helps to clus­ter them for greater impact.

How­ever, the real key here is to make them con­gru­ent with the thoughts and flow of the copy. Because too many peo­ple nowa­days plunk their tes­ti­mo­ni­als in huge clus­ters in the mid­dle of the copy with no pur­pose or rel­e­vance to that par­tic­u­lar point in the pitch.

I’ve seen bet­ter test results when tes­ti­mo­ni­als are mean­ing­ful, not only in them­selves but also in terms of how they relate to the rest of the copy at that point. So pick and use tes­ti­mo­ni­als that are rel­e­vant to spe­cific sec­tions of your pitch.

For exam­ple, say a sec­tion in your copy is meant to han­dle objec­tions pre-​​emptively. This is a good place to add tes­ti­mo­ni­als from peo­ple who were also skep­tics, had the same objec­tions, and became con­vinced after buy­ing the product.

In this case, the tes­ti­mo­nial is answer­ing the objec­tion (more specif­i­cally, a seem­ingly objec­tive third party is answer­ing the objec­tion), and not you.

Say you sell a soft­ware that promises to increase your search engine rank­ings. So your copy at some point starts to talk to the skep­tic who’s been burnt by other search engine soft­ware, and dis­cusses how it really is dif­fer­ent than the others.

You tell your reader how supe­rior your soft­ware is to oth­ers on the mar­ket — likely soft­ware that failed to deliver — and that it really does deliver on its promises.

But instead of stat­ing it out­right, use tes­ti­mo­ni­als to drive that same point home. The copy might say, after the objec­tion was stated in the copy:

“As you can see, [soft­ware] truly is bet­ter than most alter­na­tives on the mar­ket that only offer [lesser result or ben­e­fit]. Take the case of John Smith, who nearly didn’t buy my soft­ware because he’s not only a diehard skep­tic, but also bought other alter­na­tives that pro­duced less than favor­able results. Like you, he was extremely doubt­ful. But after he reluc­tantly agreed to give the soft­ware a try, here’s what he had to say…”

… Fol­lowed, of course, by his testimonial.

In my tests, I’ve found that adding tes­ti­mo­ni­als too early in the let­ter decreases response. Per­haps for the same rea­sons men­tioned ear­lier: they scream “salesletter!”

Unless you’re sell­ing to an estab­lished client base, to a highly tar­geted audi­ence, or to traf­fic cre­ated by affil­i­ates who pre-​​sell your prod­uct for you (which, in this case, tes­ti­mo­ni­als early in the copy do work well), try to put your tes­ti­mo­ni­als a lit­tle later.

I pre­fer putting them near the intro­duc­tion of the prod­uct, the objec­tions, the offer, the ben­e­fits (near a bul­leted list of ben­e­fits, for exam­ple), and of course, the price.

A final note. Actu­ally, a caveat.

Keep in mind that any­thing I say here may or may not work in your copy.

Just like I said about other people’s test results, mine should be used as guid­ance, not gospel. I know the chances that they will work are high since I’ve tested them thoroughly.

But never, ever take my word for it. Sure, try them. Test them out. Val­i­date them for your­self and your par­tic­u­lar mar­ket, indus­try, and offer. See if my test results match yours. If they do, great. If not, test them again but with other vari­a­tions, or move on.

And if you have sev­eral web­sites or saleslet­ters, don’t stop there. Test it some more. You’ll be glad you did. Or like some of my stu­dents, you’ll be mad — mad because you’ll soon real­ize how much money you left sit­ting on the table all this time by not testing.

About the Author

Last 5 Posts By Michel Fortin

• Superior Value Equals Superior Sales
• On Not Playing The Blame Game
• The Future of The Internet is Cloudy
• Carve Your Niche By Dominating One
• Can Copy And Content Commingle?

Other Related Posts

• How to Improve Your Email Open Rates (24)
• Are Bottlenecks Clogging Your Sales? (22)
• Three Tips To Raise Your Response (9)
• There's More to Tracking than Conversion (13)
• The Truth About Mega-Headlines (16)