Volume 117: 9/1/05

Honey v. Vinegar

My neighbor to the right shakes his head and informs me, "You catch more flies with honey than vinegar." He does this whenever the topic of my neighbor to the left comes up in conversation - it's his way of saying my neighbor to the left is lacking in the interpersonal communication skills department.

Far too many Web sites serve up a nasty dose of vinegar when it comes to the language of error messages: "Somewhere in the transition from offline to on-, we lost the mantra, "The customer is always right." Online, most Web sites treat users as if they're always wrong. At least, the error messages on Web sites make it seem so."1

Frankly, I don't think customers are always right. But, as they are in control here, I absolutely believe you have to give them the appearance of being right. Call me a short, green Machiavelli if you will, but in this case, I'm one hundred percent behind the idea that the ends justify the means. Swallowing the blame for those times when things go wrong - at the very least, finding a way to deflect blame from your visitor - eases the persuasive process and goes a long way to improving your image.

It goes like this ...

A Sampling of Error Messages

Time Out. Your browser was left idle and lost connection. (Can you see the finger wagging at you?)

Can you please modify your search? (Can you hear an extra emphasis on "please" as a foot taps in frustration?)

Invalid code! (Can you hear the sirens going off?)

Unknown login. (More sirens!)

Errors have occurred during the process of your form. Please make the following corrections. (Can you hear the exasperated sigh?)

Error: Phone number cannot contain dashes. (Now you tell me?)

One or more required fields were not filled out correctly. Red = invalid. (Is that a ruler poised over your knuckles?)

I like to imagine error messages are more often than not composed by techie types - folks who are perfectly happy to call a spade a spade and scratch their heads when you suggest the label might not sit well with all spades. Techies code stuff that will process provided you follow the process correctly. A form doesn't submit because you overlooked entering your zip? Well, that's down to you. The error message tells you so. End of story. Wanna go get a cup of coffee?

The thing is, any error message that implies your visitor is an idiot, careless or poorly informed about how you have set up your system is a bad error message. It's a bitter pill to ask your visitors to swallow. Throwing in a "please" doesn't really make the pill taste much better - "Please correct your error" plays little better than "Correct your error." Nor does adopting the "passive-aggressive" angle, as in "We're sorry, but you didn't do this correctly."

It doesn't matter whether your visitor was actually wrong; the last thing you want to encourage them to feel, at any time in their online experience, is testy. After all, they may be about to do you the favor of helping you meet your goals.

In Search of Honey

One of the primary goals in interpersonal communication is to minimize the offensive edicts that breed defensive reactions. The second I tell you "You made a freaking mess of this", you're immediate reaction is, "I did not!", after which you proceed to justify what happened. Defensively.

I'm not suggesting you resort to these sorts of classic therapy-speak phrases: "When you do [fill in the blank], I feel [fill in the blank]." I am suggesting you find a way to change the perception of where the blame for an online errors lie. And then offer a solution. In other words, serve up a dose of honey.

Friendster turned a hideous error message ("Error: You've been logged off due to inactivity") into a thoughtful favor: "Sorry! We thought you left, so we closed your Friendster session for your security/privacy. Please log in again!"2  Not only does Friendster accept the blame for the situation, they turn it into a virtue, so instead of huffing away, you're much happier to log in again.

The Basics

  • In announcing the problem, use "I" or "We" statements: "We can't find that login" or "We can't find an exact match"

  • Instead of heavy-handed "Correct this" commands, opt for the "Please try again" variations

  • Write your messages in the active voice. "One or more required fields were not filled out correctly" is passive. It doesn't directly blame your incompetence, but you most certainly are the implicit idiot. Why not say, "Oops! We don't understand the zip code. Please try again."

  • Head off error messages at the pass. If your system can't cope with something, say so up front: "Please leave out the dashes in your phone number" or "Please use the two-letter state abbreviation."  Better still, extract the format you need from the information they provide.

  • Offer options. If you are speaking with a registered customer, don't berate her with "You have entered an incorrect address." Ask her, as does Amazon, "Could one of these be the correct address?" When logins go awry, provide a help solution.

Your Mission

... should you choose to accept it. Head out into the wilds of your Web site and start making every mistake you can think of to generate one of your error messages. Muck up a form. Select an invalid quantity. Forget to fill in an entry field. If it can be done, someone will find a way to do it. (You can, of course, just get a print-out of all your error messages, but this isn't nearly as much fun.)

Now, read the language that confronts your good-willed, unsuspecting visitors, keeping in mind it doesn't have to go to "you stupid moron" lengths to offend. Never reprimand your visitors or leave them hanging on the wrong end of an online problem. Swallow the blame (it is, after all, a system you created), phrase the situation so it doesn't generate a defensive reaction and offer considerate solutions.

My buddy Bryan Eisenberg revealed this interesting observation in one of his articles for ClickZ,

Bear in mind this strange but true irony: Customers who do experience problems that are handled well by a company often rank their experience with that company higher than customers who don't encounter any problems at all!3

When things go wrong online, the quality of your response is a clear indicator (think bull horn on max volume) of your customer focus. Heed my neighbor to the right: when it comes to cultivating relationships, honey works much better than vinegar!

1 "Error? It Wasn't My Fault!" Jack Aaronson. ROI Marketing, ClickZ. August 26, 2005. http://www.clickz.com/experts/crm/traffic/article.php/3529926

2 ibid

3 "Do Your Web Forms Show Good Form?, Part 3."  Bryan Eisenberg.  ROI Marketing, ClickZ. May 28, 3004.  http://www.clickz.com/experts/crm/traffic/article.php/3359741

Train to Increase Your Conversion Rates

New Services

We've been busy bees over the holidays.  Check out our services page - new services have already been launched and more are coming!

New Publications

You'll also want to keep an eye on our publications page.  We've been working on creating resources that will help you put many of our principles into practice more easily and more efficiently.  We're just about to release several of our newest products, including Which Sells Best?: A Quick Start Guide to Testing for Retailers and The Conversion Experts Handbook.

The Long and Short of Persuasion

Enter the Bear family home. They are discussing optimal copy length.

Papa Bear thinks copy shouldn't be "too long." Mama Bear counters it shouldn't be "too short." Long, short, long, short. You're head can whip back and forth faster than the balls fly at Wimbledon. But if you're caught up in their debate, you're actually missing the point. Fortunately, there's Baby Bear, sensible cub that he's always been, who says copy should always be "just right."

To kindle the mood for our forthcoming Persuasive Online Copywriting Workshop, I've solicited a favor from Wizard of Ads partner Chuck McKay. Chuck wrote the best article (posted it to his blog) I've ever read addressing the subject of long versus short copy.

So here's to Chuck, gracious man that he is, and our joint desire to make Baby Bears of you all!

The Long and Short of Persuasion

by Chuck McKay

Researchers must be careful to neutrally phrase a question, so as not to influence the response.

My partner, Roy Williams, offers a perfect example. When a penitent asked if it was proper to smoke during prayer, he was told it was not. But when the question was rephrased as: "Is it acceptable to pray while smoking?" he was assured that prayer was always appropriate.

Sometimes it's not the phrasing that controls the outcome. Sometimes people ask the wrong question. The wrong question, in this case, is "Which sells better? Long copy or short copy?"

I'm a long copy proponent. That is, I'm opposed to the "nobody will read more than 300 words" school of advertising.

Short copy has inherent risks. Because it has limited amounts of compelling information, response rates are frequently low. There's also the risk of high numbers of cancellations and refund requests because the product or service wasn't what the customer imagined.

The short copy crowd assumes that everyone is like them. "I wouldn't read this," they argue, "therefore no one else would either." However, these people are not interested in what you have for sale. Without any interest, no matter how short the copy is, they will not read it. Will they read 300 words? They won't read 100.

Fans of short copy are almost never successful copywriters.

When the copywriter ignores people who won't buy, and concentrates on those who may, copy invariably grows longer. Be careful, though. Long copy in the hands of an unskilled writer becomes an excuse for sloppy, non-focused, undisciplined writing.

Long copy proponents have research on their side. Split-testing research shows that long copy consistently outperforms short copy. Additional research indicates that although readership does fall off dramatically at 300 words (when the non-interested browsers lose interest) it does not show further erosion until 3,000 words.

This argument over long copy vs. short copy has raged for years. Unfortunately, it's a tangential issue.

It asks the wrong question.

To get to the right question we need to assess the customer's perceived risk, and the emotional commitment necessary to persuade her to buy.

The biggest risk any purchaser makes is the possibility of wasting her money in a bad purchase - one that doesn't suit her needs. The lower the price, the less risk. The less the risk, the lesser amount of emotional commitment. A lessened amount of persuasion becomes necessary.

We've all been in a check out line at a convenience store or a grocery. We've noticed the magazines, the candy bars, the breath mints. In retail, these are known as "impulse items." No emotional involvement required. No financial risk. Impulse items are low priced items.

Long copy may well bore the potential purchaser of low-risk items.

Note that you won't be able to pick up and admire the portable DVD players, or the jewelry, or anything with a stiff price tag as you wait in line. These things don't usually sell on impulse.

The higher the price, the less likely Miss Prospect is to purchase it on a whim. As price goes up, so does the risk that she's making the wrong purchase. As risk goes up, so does the requirement for emotional commitment on the part of the buyer.

When our prospect is considering a major purchase, short copy may leave her wanting to know what she gets for her money.

So, in order to decide how long to make your copy, you'll need to determine the amount of reassurance Miss Prospect requires. If you're selling candy bars, she won't worry about the rent check bouncing. If you're selling college enrollment, and asking for a commitment of $25,000 over the next eighteen months, she will require more assurance.

This leads directly to the right question: How much persuasion does the prospective customer require to be comfortable making the purchase?

Her comfort level will be directly proportional to the number of dollars in the "ask."

The length of your copy should also be proportional to the size of the ask. When asking for a small amount a simple easily remembered message is appropriate. When asking for a large amount your copy must anticipate every objection, every question, every doubt that your prospect has in you, or in the product or service you're selling.

Of course, it must also be well written, persuasive, and compelling.

The message must be salient.

Salience is the relevance of the message to your prospect. It's the most overlooked quality in advertising. It's the reason for the long copy / short copy debate. It's also the reason the debate is bogus.

Remember, your purpose is persuasion.

You're trying to get a total stranger to open her purse and give you money. Write something that speaks directly to her. Give your message salience.

Write what needs to be said to convince Miss Prospect that owning your product or service will affect her life. Get her emotionally involved. Tell a story. Share testimonials. Use statistics. Boost your credibility by whatever means is available to you to remove as much risk as possible. Guarantees are golden. Add as much information as necessary to make the sale, and not a bit more.

Then start cutting any excess from your copy. Remove any word that can be removed without changing the meaning of the sentence.

Do you now have strong, persuasive, motivational copy? Long enough to make your points? Short enough to get right to them?

Assuming that you truly understand your prospect, and have written to her concerns, your writing will automatically be the appropriate length, whatever that length may be.

Chuck tells me he'll be at the Workshop. I'll be there too. And you?

Volume 117: 9/1/05


Contact Us


Bi-Monthly Online Conversion Rate Marketing Newsletter:

We Value Your Privacy!

  • Cardscan
  • ProFlowers
  • GE
  • Leo Schachter Diamonds
  • Overstock.com
  • SAS
  • Disney
  • Gatelys.com
  • NBC Universal
  • PriceWaterhouseCoopers
  • Volvo International
  • Southern Company
  • BuyTelco
  • Allegis Group's Thingamajob
  • WebEx
  • CafePress
  • Elance