“So what did you think of the Season Finale of Heroes?””Dude no spoilers I haven’t seen it yet.”

“Oh, alright, highlight the following to see what I thought: *start highlight* The fight scene between Hiro and Sylar, which was supposed to be the highlight of the entire series, was over in less than two seconds.  He stabs him, the end.  What a let down! *end highlight*”

“Hi forum people!  Buy cheap Lost DVDs at spammerplanet.com for only $49.95 per season!  Free shipping!”

“Spam!”

“Spam!  Reported!”

“Where are the “#”#%ing mods?  I tell you, this forum has been going downhill since Tommy stopped running it.”

Recently, in the Business of Software forum, somebody made the observation that I have never posted to Usenet about Bingo Card Creator, and they reasoned from this that posting to Usenet about consumer software is unpopular.  (For those who may not know me: Bingo Card Creator is software which produces bingo cards for teachers.  I wrote and sell it as a small business.)  I’m both flattered and frightened that I’m supposedly representative of good marketing practices.  However, I think the conclusion should be a little broader than they one they drew: it is ineffective to directly market to a community which you are not a part of, be that a forum community, a mailing list, or a Usenet group, regardless of what you are selling.

What Community-Oriented Marketing Is

The key thing about Usenet or a PHPBB forum or your local school’s teacher mailing list is not the technology that is used to drive them.  The key thing is that they have a community, quite possibly a very tight knit one which has built up over years.  That community has its little social rituals, in-jokes, standards of acceptable behavior, shared history, friendships, rivalries, dramas, a whole tapestry of meaning for its members that you, the outsider, know nothing about.

If you attempt to sell something directly to the members of a community you are not a part of, you risk a great chance of falling afoul of community norms and an almost certain chance of wasting your time.  Many communities are quite opposed to the commercialization of what they perceive, correctly, as their shared social space.  Some have debates rage for years about whether its appropriate or not to put AdSense ads on a forum sidebar (sidenote to board admins: please don’t.  Regardless of whether its appropriate or not nobody will click them.  They’re coming for the community, not to be marketed to.  The only way to use AdSense on a forum is to put interstitial ads between posts and harvest misclicks.)  Some of them are filled with Slashdot-esque folks who are fundamentally opposed to people other than themselves making money for any reason.  Some are filled with folks who either do not have money or should not be spending it if they do have it.  If you’re not a member, you will not know the lay of the land, and you might step directly into one of the minefields.

Legitimate Ways To Market To A Community You Are Not A Member Of

Of course, there are a variety of approved ways to drop an advertisement in a community.  You could, for example, buy an AdSense ad there — see above, though, its tremendously unlikely to be effective.  Some communities have established Advertising boards — this should be a pretty big hint to you that they really would not appreciate an ad in their main forum.  Of course, traffic to their advertising board is a bare fraction of what it is to the main forum (if everyone wanted to see ads they’d be accepting of ads there!).  I spent about 2 hours when I started up Bingo Card Creator running around the Internet, finding ESL forums, looking to see if they had an advertising board, and dropping in a tasteful and honest ad for Bingo Card Creator if they did.

The Relative (In-)effectiveness Of Direct Marketing To A Community You Are Not a Member Of

The best performing ad out of those sends me a total of perhaps 10 visits a month, resulting in 2 trial downloads.  Think of that: one hour per trial download per month (TDPM — many marketing expenses are evergreen on the Internet, so its handy to watch how a one-time investment continues sending you traffic as opposed to watching the one-time spike of traffic immediately after posting).  By comparison, there are many, many better ways to deploy one hour of my scarce time.  One way would be to work at McDonalds, because a trial download is only worth about fifty cents to me.  But thats no fun.

Writing my Squidoo page took me 2 hours.  That was worth 5 TDPM.  Already thats doubly effective.

Writing a list of Dolch sight words for my website was “slightly” more effective.  It generates about 200 TDPM, partially from organic search and partially from folks who pass the link to their friends.  That page took, yep, about an hour to write and we’ll call it another hour work of linking from my blog over the last year.

Writing a single blog post about Free Bingo Cards took me about 15 minutes.  That generates about 150 TDPM.  (Note: I had quite a bit of help on that post thanks to an impromptu brigade of friends who decided to link to it.)

Oh, and in my portfolio of marketing efforts, there is one community link that sends me 20 TDPM.  It cost me 0 minutes to write, because I didn’t write it.  If I had written it, it wouldn’t have sent me a single hit.  The reason the link is so effective is because someone who was trusted in their community put their reputation on the line and said “Hey, check out this site, it will help you educate your children”.  It wasn’t an anonymous fly-by-night posting from some Internet entrepeneur (sadly, half of the world now thinks that is a euphemism for “spammer”), it was a recommendation given to the community by their childrens’ teacher. 

Trust Is Key

That is what community-oriented marketing comes down to: trust.  If you don’t have it, then building it up will take quite a bit of time, and you have much better options for marketing in terms of time spent per marginal exposure gained.  (See the above list for some ideas.)  If you are trusted somewhere, you might be able to effectively market there, based solely on your existing trust.

I personally haven’t used that method.  I am trusted, for example, in a community of ESL teachers close to where I live.  I know I could send out a email to the list and get 50 downloads of Bingo Card Creator in a day.  However, I’m trusted precisely because I have not been a self-promoter for the last three years, and I see no reason to throw that trust away now for a piddling amount of money.  On the other hand, I was a volunteer translator for a major Japanese ESL textbook, and they were kind enough to throw me a link from my biography (on the acknowledgements page) to Bingo Card Creator, which can’t possibly hurt.  That is marketing, but it is marketing which enhances my trust in the community rather than detracting from it. 

I strongly suggest that you do the same — don’t aggressively push your product at anybody who knows you and could possibly use it.  That makes you into the Internet equivalent of the Mary Kay lady, somebody who aggressively tries to promote her business to all of her soon-to-be-former friends to the exclusion of anything approaching a real, honest relationship with them.  Instead, continue going about interacting with your communities just like you do right now, and the marketing will more or less fall into your lap.

Finally, continue providing an excellent product and service to customers.  Customers are the first, last, and best marketing team you will ever have.  They are already trusted in more communities than you can even conceive of, and when they plug your product for you their words will be trusted and their consciences will be unburdened, because they are doing it to help their friends rather than to help themselves.  Its a win-win situation for everyone involved.

And, yes, I was severely disappointed with the last episode of Heroes.  Grr.  They’d better improve for next season. 

Cutting a customer off: There is a certain school of thought that says you should have a maximum level of tolerance for any particular customer using support resources, and after that point you say “No more”. I actually think this is (potentially, depending on execution) a decent idea, which might suprise people in light of my recent paen to excellent customer service posts. You might also be suprised to learn that I’ve described someone doing it in the last 24 hours.

Here’s the cruel math of telephone customer service: the average cost of servicing a phone call is $12. The average profit of a small order is less. You cannot afford to absorb a support call for every small order. Class poll: who caught the fact that when the the representative offered free shipping and cookies to a minor no-profit-in-this-transaction customer it terminated a (potentially hostile) support incident in under a minute, totally obviated the need for a second call or an escalation to the supervisor, and still got the sale? And that that customer was so ecstatic to be brushed off he came back bearing hundreds of thousands of dollars? Thats the difference between support and customer service. As a support incident, that call was a waste of time/money. As an opportunity for demonstrating you’ve got an unparalleled dedication to customer service, that call is as good an opportunity as every other customer contact you make.

Saving Money/Time on Support: Your first line of defense against “wasting time” (never, ever, ever think of talking to a customer as a waste of time: see below) in support is producing a quality product. I sell to a rather non-technical market. I could be spending the rest of my life fielding calls on how to use the product — and I rather don’t want to, so I coded the project to be immediately usable by anyone who is capable of finding the Internet. The handholding starts at downloading/installation (clicking enter until you can’t anymore works and will dump you at my program’s main screen) and continues to my main screen (which doesn’t just explain what you need to do, it reads my app’s typical use case out to you, step by step). Improving your application is probably one of the best-scaling support investments you can make: if you consistently find yourself copy-pasting a canned “That feature is in the Tools Menu, 3rd from the bottom” response, you should probably go about making it more obvious. If you have a work-around for that annoying printing bug, fix the annoying printing bug.

How to Think of Support: I would generally advise against thinking that you’re wasting time doing support. Most people are rather poor actors — heck, most actors are rather poor actors. If you’re annoyed by the incident thats likely to come across to your customers. Think of it as an investment if it helps you — you’re investing in your reputation as a customer-service powerhouse. Against a reputation like that, firing off a few emails every day is cheap. (Think of it: suppose that 5% is an accurate accounting of the number of “needy” customers you have. Suppose you’re rolling in the dough at 1,000 orders a month. 5% of 1000 is 50, multiply by say 4 inquiries each is 200, averages to 7 per day. 7 emails a day is nothing — you can take care of that while brewing coffee.)

Politeness and a smile are free: You can’t always say yes to a customer request (although I’d strongly suggest defaulting to yes and requiring good reasons to say no, rather than the other way around). You can, however, have 100% of your customer-facing communications be polite and positive. For instance, compare the following two emails.

Bob,

You emailed me about this support issue 4 times this week. I’ve done what I can for you. Its obvious things aren’t working out. I’m refunding your purchase price.

Sincerely, Peevish uISV

Thats the wrong way to do things.

Dear Bob,

I have done some research regarding your support request. As it turns out, our product is regrettably not the best on the market for your needs. In our professional opinion, MicroFoos’ Foozle 2006 is a closer fit for your requirements. While it pains us to have not been able to help you, here at Pleasant ISV we are totally dedicated to customer satisfaction. Accordingly, we could not in good conscience accept your money with these issues outstanding, and have instructed our credit card processor to refund you.

Thank you for choosing Pleasant ISV and we look forward to the opportunity of serving you in the future.

Sincerely, Pleasant ISV

Content-wise, these emails describe the exact same set of circumstances. In terms of customer perception, these two emails are worlds apart from each other. Peevish ISV’s mail is, well, brusque and strongly leaves the impression that there was something wrong with the customer. Pleasant ISV’s mail doesn’t blame anyone (no, really, read it again — it doesn’t say or imply a single negative thing about Pleasant). It screams “we’re competent, we’re professionals, and we don’t accept anything less than the best”, and it leaves the door open rather than slamming it on your customer’s fingers.

An inspirational quote: “ We are what we repeatedly do. Excellence, then, is not an act, but a habit.” — I have seen this one variously attributed to Aristotle and Adlai E. Stevenson. In any event, if you make excellence the routine in your customer service, people will know the difference. Promising good support/promising good customer service scales very well for your business. Think of it like the guarantee: you can’t afford for everyone to ask for their money back, but you know in advance that the overwhelming majority of people will not ask for their money back. And you know that the number of customers reassured by having the guarantee outweighs what you’ll spend on it. Support is exactly the same. Most people won’t need it, but lots of people are reassured by the fact that it is available if they do need it. Your customers have been trained for years to distrust software, despite the fact that most customers will not have any problems: software is impersonal, software gets in the way of them doing what they need to do, software breaks, and when software breaks they spend 10 hours in tech-support heck talking to people who hate them and don’t have any answers.

What if you could reassure customers? What if you had a deserved reputation for *not* blowing up in people’s faces, and for being a joy to work with? So beat the drum and beat it loudly:

If you have any problems, or just want to ask a question, talk to us. We have an actual human here. Even better — not just an actual human, you’ll get all your support emails answered directly from the head engineer/company president! (Try that with your other software vendors some time!) We care about your concerns and will work to make them right. Take a look at what Mary Sue of Normal, IL and Bob Smith of San Fransisco had to say about us: “Wow, I had expected to get a canned reply but they got back to me within 15 minutes and kept working until my problem was solved.” and “Pleasant uISV is the best in the industry, bar none. Once I bought one of their products and it wasn’t working out for me. They gave me my money back without me even asking and referred me to a competitor! It was more important that I be happy than that they make a buck. I’ll never stop using Pleasant and I recommend them to everyone I do business with.”

Or, you could say something like

Support incidents: Every customer has a maximum of four support incidents, after which they must pay a non-negotiable charge of $24.95 per incident. We do not answer requests about generic computer configuration problems, setting up web pages, etc etc.

Which of these two companies would you rather do business with? Which sounds like a risky investment? Which sounds warm and inviting?

But my favorite part of the wall was letters. Quill had, to my knowledge as a lowly worker bee, two policies regarding letters: first, they were all collected and any workers individually singled out in them had the action noted on a publicly-visible board and in their file (and Quill would FIND you: there was at least one letter addressed to “that lovely Spanish woman with the Texan accent” and, by God, she got the credit for it). Second, the letters and/or copies (some people couldn’t bear to part with them) were posted on the bulletin board next to the numbers.

Anyhow, my favorite letter wasn’t a letter at all. It was a prayer card from a convent in, as I recall, New York. For anonymity’s sake we’ll call them the Sisters of Perpetual Gladness. Oh, they were a happy bunch, and it was contagious. Do you know why we had a prayer card from the Sisters of Perpetual Gladness? I don’t, because the people named on the card (whom the Sisters had dedicated a novena to — thats nine days of prayer, for you non-Catholics in the room) wouldn’t say exactly what they had done. Thats my personal metric for customer excellence: Nuns Praying For You (NPFY). Paypal, yeah, there’s a NPFY-0 business if I ever saw one. Quill probably has the highest NPFY of any any business I’ve ever been associated with. Keep in mind we were earning NPFY60+ for, well, prompt delivery of paper and pens.
But the Sisters were loyal customers and in Order Entry we lived for their call. I got one once — it made my year. And the Sister blessed me, too. I blessed her back, which was probably technically against company policy (although if you’ve been reading closely you know that when stacked against customer satisfaction policies were more like guidelines and guidelines were more like advice and advice was more like suggestions and suggestions were more like comments). Anyhow, for the Sisters, if they had asked Order Entry “We’d like an 18-wheeler filled full of printer paper and the most we’re willing to pay is, hmm, nothing. Oh, overnight delivery, too.” the only thing we would have said is “But Sister, where will we put your cookies?”

There was only one problem with the Sisters. We had some heavy-duty CRM (customer relations management) software at Quill. It supported locking a customer to one particular desk or extension — for example, if you’re the Big Important Government Account with a 6,000 page bid request your order is probably too important to be handled by $10-an-hour me, so you got locked to a specialist. The Sisters did not require any special handling like that. As a matter of fact, they required a specially-written rule to ban handling like that for their account. The reason was that every lowly worker bee could put a special handling restriction on an account, and everyone could take off a restriction from someone at equal-or-less access level, which made the Sisters’ account history look like a Wiki revert war (“They’re MINE!” “NO, they’re MINE!” “My family is from New York! I should get to call them*” “But you’re Jewish!” “So are you!” “Yeah, but I dated a Catholic once!” etc, etc)

* The Sisters’ order always had a problem requiring calling them, immediately. Typically, it was that they had missed an opportunity to get free cookies. At any given time, you see, Quill had hundreds or thousands of promotions running. Dozens of these ended in the punch-line “Free Mrs. Fields’ Cookies”. We got a very stern notice every month “Here are a list of the new promotion codes, in case a customer forgets. Remember, Mrs. Fields’ Cookies are NOT appropriate as a we’re-sorry and the customer must specifically mention the promotion to receive the premium.” An exceptionally morally upright employee would phrase the phone call something like this:

“God bless you. This is Sisters of Perpetual Gladness, Sister Mary speaking. How can I help you?”

“Hello Sister Mary, this is Bob for Quill Corporation Office Supplies. I was wondering if I could speak to you about your order of *checks watch* two minutes ago?”

“Oh, certainly.”

“Well, Sister Mary, we reviewed your order and there seems to be a problem. The last 47 times you’ve ordered with Quill, you ordered one free tub of Mrs. Fields’ Cookies. We see there are no cookies on this order, and wondered if there might be an omission.”

“Oh, thats certainly kind of you, but we didn’t see the free offer…”

“Didn’t see? Why, no problem, we can look it up for you. Do you know what catalog you were using?”

“Oh, I don’t know, I don’t have it in front of me.”

“Hmm, I’ll bet it was the *check premium list* California july legal small offices circular.”

“Err, are you sure?”

“Well, did your catalog have the word Quill and some brightly colored office supplies on the front cover?”

“Yes, yes, it did.”

“Well, mystery solved! So did the circular. The promo was on page 437b. Oh, don’t worry, I’ll go right ahead and get this order and your cookies you to you. You don’t have to do a thing.”

“Why, thank you very much young man. God bless you.”

“God bless you too, Sister Mary.”

Minimally, you should thank them for choosing your software then inform them you have refunded their purchase at their request, and sign off gracefully. If you feel the need to ask why, you can then ask why, but phrase it in such a way as it sounds like a favor to you rather than a business request. People will generally respond pretty favorably to requests for a 10 second favor after you’ve just given them money.

I learned this lesson through my first real employment, which was as an order entry operator at Quill Corporation Office Supplies. Quill is outwardly fanatical about customer service: without violating any confidences regarding their internal guidelines, suffice it to say they would rather have some customers get away with murder than inconvinience another customer to the tune of 5 bucks. I have seen their customer satisfaction metrics and the comparison to other companies in the industry, and the fanaticism works.

You might be inclined to ask why the person is returning the software, and then process the return pending a reasonable explanation. This is a mistake. There is no question that your customer can get money from you: the question is whether they go through you or they chargeback. I can scarcely find words to describe how much better for you it is that they go through you. Well, OK, here’s an attempt: if they go through you, it costs you either “nothing” or “very little” depending on your payment processor, if they chargeback you will frequently get hit with a large fee ($25+) on top of paying your payment processor fees, and frequent chargebacks will get your account yanked.

You might be inclined to write off a customer who has requested a refund. From a support perspective, this is correct. From a politeness perspective, wrong move. You’re not Quill with 6 figures worth of customers, but you are in a niche market, and your niche talks to each other. The next time your ex-customer is chatting with his friend and his friend expresses the same need he had, he will always, invariably, talk about his experience with you. Since he’s your ex-customer, this will not be a maximally positive conversation, but you can choose whether its “Well, I bought from Bingo Card Creator. It didn’t do exactly what I wanted but when I asked for a refund I got it in 10 minutes and I was impressed by his professionalism” or “I bought from those “#$”#%”$#!s at Bingo Card Creator. What a scam. It crashed my computer three times and when I demanded my money back I had to fight him three days for it. Burn in “#$#”$”, “#$”#%”#%”. So, Bob, how are your kids doing?”

Big deal, you think, how many people are going to buy or not buy on the basis of a quick conversation with a co-worker? Answer: this is the most influential form of advertising you will ever get in your life. I remember getting an all-hands memo when Quill gained a $X00,000 a year account because the purchasing officer had a good experience seeking a technically-out-of-bounds accomodation when ordering a single cartridge of toner for his wife. He wanted free shipping on a $Y order. We didn’t offer free shipping until $45. The representative he spoke to said “No problem sir, we’ll make an exception for you. And have a box of Mrs. Fields’ Cookies as our thanks for calling in”. Do you know what a box of Mrs. Fields cookies and eating a shipping charge costs us? Suffice it to say that its more than the amount of profit the company made on a $Y order. Do you know how much profit a $X00,000 account makes in a year? It pays for a whole lot of cookies.

(P.S. If your uISV ever requires office supplies, I heartily recommend my old colleagues at Quill. You will find no better service in the industry than you will get at Quill. If I’m wrong, the cookies are on me.)

Stupidly simple thing to keep in mind as a uISV: always leave enough money in your account to cover the most damaging single return you’re currently liable for. If you have a 100% money-back guarantee policy for 30 days, and you’ve sold $1000 of product in $25 increments, keep minimally $25 in your account. If you’re using Paypal, this will decrease the amount of time it takes to get your refund authorized. Even if you’re not, if you have to do account-balancing tricks to get your customer their refund (in my case, that could possibly require an international wire from Japan to the US, followed by funding my Paypal account and then processing the refund), it will take a while and peeve off your customer. What if two people request refunds? In this case, you’ve probably done something wrong. :)

Oh, why offer a refund? In a nutshell, because the marginal number of customers the refund offer will get you outweighs by many orders of magnitude the amount of charges you will eat processing refunds. Steve Pavlina covered this better than I ever could (search for “guarantee”, although the rest of his advice is decent, too).

Some things which I’ve found useful, and which can be done fairly easily without compromising your ethics or your user experience:

1) A different, meaningful title on every page. “Meaningful” means “makes sense to humans AND includes keywords”. If your title is something like CompanyName Website, you’re losing valuable SEO currency unless people routinely search for CompanyName. So, for example, my homepage is “Make Your Own Custom Bingo Cards With Bingo Card Creator”. The first five words of that is a decently common, high value (near 80% conversion) search string, which I rank rather highly on on Yahoo (5thish) and MSN (1st). So do I use that sitewide? No! My purchasing page, for example, says “Make Printable Bingo Cards with Bingo Card Creator”. That simple expedient puts that page at #4 on MSN for both “make printable bingo cards” and “printable bingo cards”, both of which are (as far as my keywords go) competitive. Incidentally: all 3 big search engines treat keywords to the left of the title as the most important, so you don’t want a title like “Company Name: Delivering Cheap and Cost Effective Widgets” if you have people searching for “Widgets”. How many interior pages do you have on your site which get read once in a blue moon? A simple change of their title to include some keywords and, bam, free traffic and some utilization of all that time you spent writing, e.g., your privacy policy.

2) Use URLs to your advantage. MSN is a sucker for this, and it helps elsewhere too. If you have landing pages for, say, “Dolch sight words bingo”, make the URL blahblah/dolch-sight-words-bingo.htm (*cough* 2nd on MSN *cough*) rather than “bingo.htm” or, God forbid, “content.php?article=23″. The same goes for your site’s domain name. Obviously, you don’t have quite as much flexibility as that, but you should have total control over your URLs (caveat: a lot of users hover over links and read the URLs. More, I’m guessing, than who actually read page titles. Don’t name your purchasing page involved-widget-search-string.htm). Oh, that reminds me — use dashes, not underscores, to represent a space. All three big search engines think foo_bar is a single word and foo-bar matches the query “foo bar”.

3) Pluralization matters. I had assumed all the search engines did stemming (natural language processing term, means collapsing verb endings and plural forms before continuing processing on input). It appears that they care more about fidelity to the exact words used than about stemming. The difference between SERP (search engine result pages) for “cheap red widget” and “cheap red widgets” is night-and-day. For example, drop the s from “words” in “Dolch sight words bingo” (oh Lordie, I’m so going to rank on that for this post… dang it all), and I slide 5 places on every search engine. I picked the variant that was getting more impressions in my PPC campaign. Recently, Google released a tool which will let you discover which of the alternatives is more popular. Make sure you select “Global search volume trends” from the dropdown box. Unforunately, unlike Yahoo Search Marketing (YSM) they won’t give you exact counts of searches.

4) Variations matter. “first grade” != “1st grade” != “grade one”. Note, however, that if you’re writing naturally you might use all three of these in a single paragraph. Hey, does that give you any ideas? It should :)

5) Consistent navigation helps you. If your users and your desire to SEO every come into conflict, you have done something terribly wrong. Borking your navigation because you’re afraid having non-keywords in prominent places is a big “something wrong”: the search engines are really good about stripping out templates iff that template is mostly consistent across your site. You can check if you accomplished proper templating fairly easily: if when you search for a keyword you get “My Keyword Is In My Title … Keyword leads the heading, too!” with the template code ellipsed out on the SERP (search engine result page), then congratulations, the search engines are properly ignoring the template. If on the other hand you see something like “My Keyword Is In My Title Products Services Free Trial Contact About Us Keyword Leads The Heading”, you could probably stand to take a look at whether your template is internally consistent across several pages. Oh, incidentally, another failure mode is not having several pages actually indexed.

6) Thou shalt not duplicate content. Duplicate content means using substantially the same words on two pages. Your navigation will generally not count unless it vastly outsizes the actual text on a page (this is a big no-no, and you can guesttimate whether it does by pointing Google’s AdWords keyword suggest tool at the page and seeing whether it can extract any keywords or not). If you do duplicate content, perhaps because you wanted to have two substantially identical pages rank for different variations on search (say, to avoid that pluralization niggle I pointed out above), you run the risk of “going supplemental”.

Going supplemental means your page gets quarantined out of the main search index for Google, and is put in basically a low-quality jail. When Google gets a search query, they display quality results and then, if there are no quality results, they give “supplemental results”. Obviously, most strings being searched for have quality results on them somewhere on the web, and people who click through all the way to supplemental results are a very rare breed. Accordingly, “going supplemental” is the kiss of death, so don’t duplicate content. Want a safe but time-consuming way to accomplish the same thing? Write two versions of the same page which are substantially identical *when read by a human* but substantially dissimilar *when read by a computer*.

Want an example? My academic background is AI and natural languages. I’ve seen some awesome things done with text processing, and it can give the impression that the algorithm has genuine content understanding (e.g. did you know that a 30 line gawk script that can guess whether you’re a man or woman on basis of a single college term paper with 80% accuracy?) But there is really no ghost in the machine, and some tasks require *actual* understanding rather than even the coolest applications of statistics. Here’s an example: tell me whats special about the next paragraph. Now try to figure out what a computer would have to do to come to the same conclusion.

The definition of duplicate content is two pages which use mostly the same verbiage. Analysis for it does not count the user interface unless the interface is much larger than the parts of the page which draw the attention of the reader. This should be avoided. Software which analyzes a page for keywords, such as the one provided by The Big G’s CPC service, can be used to check whether you’re in danger or not: if it reports that it has insufficient information to provide quality results, you should write more. The penalty for having two mostly identical documents is severe: you can vanish off the most common search engine results pages.

Yep, you’re right, this is indeed a thought-for-thought recapitulation of a paragraph you just read. But the word order and vocabulary used are almost totally orthogonal, and even algorithmicly derived lists of synonyms will fail for most of the words (“vanish = disappear” is easy for a computer to understand, “vanish = ‘going supplemental'” is all but impossible).

On a lighter note, if search engines were girls: Google is the bombshell who everyone drools over. She knows this and lords it over everyone. You’ve really got to work to get Google’s trust. She also has a weird fascination with blogging. Yahoo is wild and erratic. MSN is the world’s cheapest date, who will practically faint if you show any interest in her at all.

Over the past month I’ve been doing a lot of tweaking of various things on my website, and generally after leaving a tweak up for two to five weekdays (my weekend traffic is highly skewed by the work habits of my customers) I evaluate whether it was effective or not. Its not A/B testing but its, shall we say, empirically driven iterative improvements. (As an aside: I’m sure my Japanese economics professor would say that I’ve just started doing things The Japanese Way (TM), as iterative improvement is routinely credited for everything from the quality of Japanese cars to… the quality of Japanese cars.) Anyhow, there are surely other, better ways to increase your conversions (such as optimizing your landing pages), but you can accomplish all of the following while brewing a pot of coffee.

1) Make your “download now” link an image. This delivered roughly a 100% boost in my conversion rate. Yeah, you read that right. I kept things very simple: it looks like a button (so its obvious you can click it), has unambiguous text (“Download (newline) Free Trial”), and I made it blue, which is figure is a nice inviting color which stands out on a site that tends to orange/yellow/white.

2) Within your main content area, put your goal for that page as the first textual link. This was roughly a 20% increase to the conversion rate. I don’t check what percentage of people get to my download page from the right-hand button versus from clicking on an exhortation in the text (although, come to think of it, I really should and could do that by just appending ?clickedOn=wherever to the link and then browsing Analytics’ “Dynamic Content” data), but systematic use of this convention has helped me alot. Here is my theory: some people will read your entire page, but a lot are just sort of browsing, and they’ll randomly click on whatever looks interesting — and you might as well have that be your goal (purchase or download, as appropriate).

3) There should be a download/purchase link on EVERY page. Want some fun stats? 71% of my visitors who download go straight for the cheese: they arrive at either my home page or an optimized landing page for a particular ad, and their next action is the free trial. What do the other 29% do? They dig around like veritable gophers. I have a non-trivial number of users who hit *every* accessible page in the site before downloading, including a rather suprising number who go through every link in the nav bar in order. And at some point the gophers decide they saw something they liked and convert. This has happened on every page I have written, from my free teacher resources to my about me blurb to… deciding to download after reading my privacy policy.

4) Your main page should have a screen shot above the fold. My program is not much to look at, but the printed output is pretty nice if I do say so myself (not exactly fine art, but it looks well done for elementary school bingo cards). I’ve tested having no screenshots/scans on the front page, having a screenshot only, having a screenshot and scan, having a scan only, etc. And I’ve tested where to put the screen shot. The conclusion: having a screenshot matters, a lot. Suprisingly (or perhaps not, since many, many visitors will leave in literally 10s or less if they don’t see something they like), it really matters that it is above the fold. To the tune of 25%. I put it south of my sales pitch on the main page for 3 days and changed nothing about the site in that time, and conversions nosedived. (Incidentally, I put the scan at the bottom of the page, after discovering having it was better than not having it but that if I put it above the spiel few people would scroll down to read it.)

5) If you use thumbnails, pop-up the full-size version in a new-window. This is probably a function of having non-technical customers, but you would be absolutely shocked at the percentage of users who in 2006 have not heard of the “back” button. I did some analysis on what people did right before they left: the largest group arrived at the site and left immediately. The second largest group clicked on a thumbnail and couldn’t find their way out of it.
Anyhow, you can quickly implement these, wait 48 hours, and revert them just as quickly if they didn’t work. You’ll notice that this data-driven improvement can’t happen without, well, data. Do you have analytics software installed yet? If not, what are you waiting for? (The only acceptable answer to that question: “My Google Analytics invite.”)

The most important thing for succeeding with AdWords is that you need to install analytics software. Let me repeat that: you will have no clue whether your (potentially very expensive) AdWords campaign is actually making money unless you install analytics software. I really love Google Analytics because of its tight integration with AdWords, but if you want to use somebody else, hey, whatever floats your boat. And you need to enable conversion tracking for at least your trial downloads and ideally for your completed purchases.

The easiest way to do this is scatter your site with download links that all go to a central page (mine is thanks_for_downloading.htm) which has the tracking Javascript on it and a meta-refresh to the executable, plus a “if you don’t see your download starting within 5 seconds click here”. One thing I do is make sure all my links use text like “You should download our free trial to …” so that everyone who clicks on the link knows that they are committing to a download. The reason behind this is that if they click on a link saying “Free Trial” or something, see the download begin, and immediately hit cancel you’ll never know and you’ll think that person was a successful conversion.

OK, got your Analytics set up? Alright, here we go:

1) Eyes on the prize. The prize is conversions, to your trial and eventually to being satisfied customers who have paid you money. You’ll be buried in numbers — click-through rate (CTR), cost per click (CPC), number of clicks, number of impressions, conversion rate (CR), cost per conversion (aka cost per action, CPA), blah blah yackety smackety, and you can slice this data a million ways. There are only two numbers you care about: CR and CPA. Everything else is noise — potentially meaningful noise when you’re optimizing your ad, but otherwise its just noise. The only thing that will get you money is to get people to download your trial and decide to take the plunge. If it costs you $30 per trial download and you sell a $24.95 product, congratulations, you should a) get serious about optimizing your ads or b) cancel AdWords today.

2) Opt out of the Google Content Network. You can find this option in campaign settings. There is one simple reason for this: these sites fail to deliver customers who convert, which hurts CR and ups CPA. Or, in plain English, you pay them money and get nothing in return, so don’t pay them money! If you mistakenly leave them on for a week, like I did, you’ll very quickly figure out why: the vast majority of clicks are from “Made for AdSense” (MFA) pages, which are generally scrapes of content which exists elsewhere on the Internet, and more than a little bit shady. I assume that most of these folks are either site owners, bots, or users who are clueless enough that they land on a MFA page and take it to be valuable information (when they almost never are). None of these folks convert.

3) Segment, segment, segment. You can make multiple ad groups within a single campaign. Make use of this feature. An Ad Group should be thematically coherent: for example, one of my Ad Groups is based around the theme “You’re searching for something to make bingo cards, I provide something to make bingo cards”. Another is “You’re searching for information about Dolch sight words, I provide a resource to teach Dolch sight words”. If you’re smart your software solves one or several pains — pitch your ads on a per-pain basis. Why do this when you could save time by throwing everything in a single ad group? Because if you segment, your CTRs and CRs will be higher, since you’re showing the most relevant ad text to the searcher.

3) Watch that CTR, but not toooo closely. The best guess is that the average CTR is about 2%. If you’re at 1%, you’re still OK. If you’re at significantly above 2%, you’re *probably* OK (but see below). But if you’re below 1%, you’re going to start costing yourself money soon. The reason is you have a Quality Score, which is essentially a witches brew of factors that Google uses to determine whether they display your ad or not at a particular price. If your QS is low, Google will keep bumping up your minimum bid to be displayed. That costs you money, so you want to keep your QS nice and high, and one easy and transparent way to do so is keeping your CTR healthy.

4) Writing ad copy. God darn it Jim I’m an engineer, not a marketer. Here’s everything I know: include a call to action (“Download our free trial today.” works decently for me), make sure you use keywords from the search in the ad copy if possible, and speak directly to the pain. You can try out many ads at once — Google will automagically pick the one with the highest CTR for you. Thats Good For Google, since high CTRs mean they make money, but its not necessarily Good For You. You want ads with a high CR, because those are the ones that make *you* money. This means you should periodically check how your CR is doing and pull ads that aren’t making you money.
5) The importance of landing pages. You’ve got five seconds to overwhelm someone’s inborn defenses against spending time/money on your product. Make the use of them. Don’t be the silly advertiser who just directs everyone to the main page — have an optimized landing page for each ad group (or segment even beyond that — for each keyword, for each ad variation, whatever you can afford on your time budget). This means pages which speak to the pains which you solve. You want an example? Compare www.bingocardcreator.com, which is a generic pitch of my software to my main niche (teachers), to my landing page for sight words. Anyone landing on that page was looking for resources to teach sight words and clicked on an advertisement promising some variation of “I will save you time and money playing sight words bingo”. I greet them in a personal manner, immediately tell them download the free trial (something like 30% of the clickers do so immediately), and then go about pitching the activity (talking about the pain, basically) and providing them lots of reasons to believe that I’m the best possible solution to the pain.

6) You probably don’t want to let Google budget for you. Well, in one sense you do — you’ll establish a maximum you want to pay per month and Google will cap your expenditures at or near that maximum. This is good. What you don’t want is for Google to “spend up” to your maximum, which is what they will do by default if you let them budget for you. Lets pretend I have a budget of $30 for 30 days (I do). See, what happens under that setting is that they will adjust your bid timed to reach exactly $30 in 30 days… But if you only spend $5 in your first 10 days, then they’ll adjust your budget to hit the $25 target in 20 days… and they do this by bidding up your maximum cost per click. Supposing your click volume is not yet high enough, they’ll raise it again and again and again. You’re almost guaranteed to make your monthly limit. Great for Google, but there is a point at which you’re not making money (where your CPA * your conversion for demos to purchases exceeds your net profit per sale). You’re better off manually limiting your expenditure.

7) There is likely more traffic than you can afford to service. For a small advertiser, you are probably not able to absorb a click from everyone who wants to click on an ad in that day. So, reduce your maximum bid. It doesn’t matter if you’re in 1st position, 2nd position, 3rd position, or 17th position if you’re still maxing out your budget every day — I haven’t seen any difference in conversion rates based on where the ad is on the page (there is obviously a difference in CTR but, oh well, CTR only makes Google money).

8) You only want qualified buyers to click your ads. Here’s an issue for my business: I sell a program to make bingo cards which is targetted at teachers. I’ll accept orders from people who are not teachers, but I know if you’re not a teacher or a parent you’re highly unlikely to want to buy my product. So if you’re looking for something to print bingo cards for the game on Tuesday night I’m happy to show you my website for free (organic search) but not happy to pay a nickle to pitch my site to you. Yet I routinely end up paying $.15 to pitch to this person, because one of my campaigns is overly broad. You don’t want overly broad campaigns. There are three ways to target your niche more precisely:

• Exclusion words. I pay for someone searching “make custom bingo cards”, with broad matching (it will hit “make custom reading bingo cards”, for example). However, I can specify exclusion words, which means if they search for foo they don’t get one of my ads regardless of how many of my keywords they hit. Consider carefully whether you really want to pay for anyone searching for “free keyword keyword keyword”. Currently, my conversion from people searching for free stuff is pretty nice (its actually higher than folks who didn’t specify if they were searching for free stuff or not). Similarly, if your keywords are ambiguous, exclude words which would resolve the ambiguity against you. For example, if you’re selling gardening software to people searching for “potter” (I don’t know why you would do this, but play along), you’d want to exclude Harry Hermione Ron magic Hogwarts etc etc. Note that excluding words does not appear to decrease the amount of money you have to pay (I’m not totally positive about this), so you’re probably better off not paying for Potter in any event.

• Speaking your customers language. In general, especially if other software exists in your niche, the two to three word description of what your software does will be expensive. On the other hand, natural variations such as “How do I <solve my pain>” are likely to be very, very cheap. Listing off a couple dozen variations of that natural search query gets you lots of very qualified traffic for very cheap.

• Make your ad text clear as to what they get for clicking. Suppose you could come up with some ad text with an obscenely high CTR by slightly stretching the truth as to what was behind the link. This is NOT a good idea. Remember, CTR is money for Google, not for you. Ideally, you’d want ad text that turned off 100% of people who would not convert while still capturing 100% of people who would. You’ll not likely be able to do that, but you can audition various ad texts to see what gets the lower CPA. Here’s three ads from my “you’re looking for software to print bingo cards” ad group:

Print Custom Bingo Cards

Your own text or use our lessons.
Download our free trial now!
www.BingoCardCreator.com

Bingo Cards for Teaching
Print custom cards on your own PC.
Download our free trial.
www.BingoCardCreator.com

Lessons Ready In Minutes
Make your budget go farther and

save prep time. Try for free!
www.BingoCardCreator.com

• Here’s the results: variation #1 has a high CTR (6%) and a high CR (20%ish), but the CPA is poor compared to targetting teachers specifically (roughly quadruple what I pay elsewhere). The reason is that I pay a lot of money to pitch to folks who weren’t interested in teaching. Variation #2, on the other hand, has a lower CTR (4.5%) but a higher CR (unstable since I’ve only had it up for two days, but I’m estimating it will settle in the 40% region). Doing the math, thats roughly 50% extra downloads for the same amount of money (or, equivalently, 1/3rd off my CPA). Variation #3 just sucks as an ad (sub-1% CTR, no significant conversions) and it will be killed right after I get done with this post. You can see why it sucks, too: its not pitching anything at the pain people are searching for.