I’ll be honest with you: I hate e-books, with a burning passion in my soul.  I don’t like reading them — they are inferior, in almost every respect, to a well-designed web page.  They typically contain worse information than could be found in a Google search, with a poorer layout and taking less advantage of the ability of your computer to organize information in a non-linear fashion.  The promotion of e-books takes heavy advantage of the fact that the cost of reproducing them is zero, and as a result there are untold scads of them which are essentially multi-level marketing schemes, with chains of affiliates seven or eight layers deep.  I have had forums which I like succumb to the disease, and its like sitting down at your favorite coffee shop only to be set upon by the local Mary Kay lady.

But while I really hate e-books, I really like Bob Walsh.  He is one of the leading lights of the uISV community and, while I sometimes disagree with him on individual details, he has previously given enough good advice on his blog and in his forum posts to make me largely trust what he has to say.  So when he asked me a few months ago “Hey Patrick, I’m going to be writing an e-book about USPs and want to feature you in it.  Will you help me out here?”, I of course agreed.  (I’ll explain what a USP is in a minute.  Hold your horses.)

That book project turned into Micro-ISVs Sites that Sell! , which Bob released to general acclaim last week.  I’ve been busy showing a friend around town, so I haven’t had time to give it a proper review yet, but now that I have 30 minutes free I want to give you my take on it.

So what’s that USP Thing Again?

A USP is a Unique Selling Proposition, which is a word I had never heard of before hearing Bob beat it like a drum continuously for the last 2 years.  The basic idea is that any product needs to have one core differentiating factor that separates it from all the other junk on the Internet.  This idea has been around for a while, as a quick Google search will tell you.  Bob’s genius is to treat the USP and assorted methods of flaunting it on your website as an engineering problem.

He refers to the USP as a design pattern.  (Sure, you could read about them on Wikipedia too — but I refuse to help them link higher than the original authors for the canonical work on the subject.)  If you majored in CS this term is familiar to you already, but if not, basically design patterns take tasks which we routinely have to do in the programming world and give them names.  After you’ve named something, you can study it, and how it interacts with other named things.  After you’ve studied it, you can systemize it.  And after you’ve systemized it, you can replicate the success of your implementation of it.  That last idea was fairly revolutionary in computer science — more than anything else, it is responsible for bringing many forms of programming from out of the stratosphere of brilliant, socially maladapted geniuses and into the realm of just-another-knowledge-worker task, like accounting.  (Imagine if the field of accounting was invented from scratch for every company’s internal audit systems.  Not just the processes, the terminology and the rationale itself.  The crazy field you’re imagining is not far from where CS used to be!)

So Why Do I Care That Marketing Is A Design Pattern?

Well, let’s get down to brass tacks: it helps you make money.  Essentially every uISV starts out with a five page site which functions as a product brochure and can’t sell their product worth garbage.  Trust me, I know, I did it myself.  You’ve seen this: top menu bar reads home, features, screenshots, download, purchase, support.  Main page has one prominent screenshot of the full application, which is seldomly legible, and about 50-200 words describing what the program does.  The features page has 20 bullet points listing things like formats the program exports to and that it is faster than the competitors because it uses unique, proprietary algorithms.  The screenshots page has six screen shots, the first three being empty and the last three being extraordinarily visually busy, possibly with additional annotation saying “This screenshot is the main screen” and “The 32nd combo box from your left is where you select the widget”. 

I don’t mean to be harsh — my first draft of my website was quite similar to that.  Heck, it is still quite similar to that in some respects.  But in the intervening 18 months I spent an awful lot of time building a mental model of how my customers act, using that model to inform decisions about copywriting and site design, and then testing the bejeesus out of the changes I made.  This blog has detail of a lot of the decisions made in real time.  Bob’s book is, frankly, a much better organized resource: he clearly explains the rationale at the outset, so you don’t have to wait half a year learning it through osmosis.  He also goes into a few case studies of successful uISVs, showing how different parts of their USP work together to better connect with customers.

Let me quote a little bit of Bob’s take on me, so that you understand the quality of the advice the book has:

Creating Relevance: [Bingo Card Creator] is a software tool for educators and parents — it says so in the first line of copy in the site.  If you are not an educator or parent, that’s your cue to leave.  [Patrick: and they do, believe me.  You should see my bounce rate on generic bingo queries, and I’m happy about that!]  And if you are a teacher, that’s the cue that this isn’t one of those awful business enterprise sites selling software way too complex to use.

Below the fold of Bingo Card Creator’s main page is a short, understandable list of features.  [Patrick: I personally would describe them as benefits, not features.  The distinction is one that Bob expounds on at length elsewhere, and it is an important one.  Features are important to you, benefits are important to your customers, optimize accordingly.  And whoops, it just occurs to me that the page calls them features.  Hah, a marketing bug that has persisted for 2 years — time to fix!] 

Here’s the first one:

Save time! You can print bingo cards for an entire classroom on your normal computer and printer in mere minutes. Writing cards by hand takes hours — hours that you could better spend doing what you do best: teaching. Leave the busywork to the computer.

Take a moment to re-read that bullet point.  Did you notice the tone, word selection and the Us teachers against the Them computers feel of it?  [Patrick: Did you notice that I am playing to teacher’s desire to be seen as important and do important things, like teaching, and also denigrating my primary competition, which is manual labor?] Educators do not particularly like, trust, desire, or respect technology the same way programmers do.  Consider this rewrite in corporate business-speak:

*snip the rewrite.  It’s funny, but long, and I can’t copy-paste.  You’ll have to buy the book.*

Same approximate meaning, but an entirely different worldview.  One of the best ways of connecting with a given market is to speak as they speak, to share their worldview.  [Patrick: I work *very* hard to generally infuse the page with a “You can trust me because I’m just like you” vibe.  Bob gives plenty of other examples, too.  I note, as an aside, that if you write how your customers speak, they will find you much more easily on the search engines.]

Little Things That  I Liked

• Hyperlinks!  An amazing concept — you can direct your readers to related or supplementary material on any subject in an efficient manner, and concentrate only on the subjects you are personally best at!  Bob makes excellent use of them throughout the book and its sidebars.  It is a bold step for the world of ebooks, bringing them solidly into the bright new year of 1996.  ;)  (Have I mentioned that I don’t like e-books?)
• Visual design.  Just because it is chock-full of information, doesn’t mean it has to look like Usenet.  It also isn’t physically painful to read on my laptop — the book is horizontally oriented, so I don’t have any annoying letterboxing effect like reading a traditional paper-> PDF conversion.  That made me read a lot more than I skimmed.
• Voice.  Voice is an English-teacher term.  It means that your writing, be it expository (“I want to teach you stuff”) or whatever, still manages to carry across your personality and uniqueness.  It is probably easier to think of it by trying to imagine a news report or corporate memo, which too-often try to eradicate any trace of personality, any flair with writing, and anything other than the facts and just the facts.  You care about voice as a writer because having it will greatly increase the perceived quality of your writing and increase the amount of attention people pay to it.  You care about it as a reader because it makes reading fun, as opposed to a chore, and frequently increases your retention of core concepts if they are well-integrated with the voice.  Bob’s voice is, among other things, unapologetically that of an engineer speaking to engineers, and if you’re an engineer you are much, much more likely to retain the core concepts from this book than if you had read the same stuff in a Sales Copy For Direct Marketers blog post. 

What I Wasn’t Quite So Thrilled With

I think that Bob has some weak points in his book.  It jumps between themes sometimes — while I agree, for example, that AdSense has no place on a uISV website, I don’t know why that information comes earlier in the book than his core themes about USPs.  Half of the thou-shalt-nots he presents in chapter 1 are, in fact, core issues with a lack of USP or borked execution thereof.  They deserve solid prominence in the book.  AdSense is just a tangent.

I also thing that Bob doesn’t stress the last part of the design pattern cycle enough: after you’ve analyzed, planned, and implemented, you should test the heck out of everything.  If your intuition about your market says that they will respond positively to your choice of language, and your market does not respond positively to your choice of language, then your intution is wrong.  If your Iterator pattern causes your website to take 15 seconds to render, time to go back to the drawing board.  If your USP gets conversion to the trial of 5%, ditto.  The topic of iterative refinement could easily fill a book, though.

My other weak point is exterior to the text: the book is priced at $19.  I don’t think that price communicates that this is a serious resource for an owner of a software company, which is designed to make you thousands of dollars once you implement the very actionable suggestions.  Bob knows his market better than I do, obviously, but that suggests to me that he is trying to sell the book to a crowd of people who are probably fence-sitting on whether they’re actually going to open a uISV or not.  If you are actually selling software — customers come to your website and buy things from you — $19 is a fraction of one sale.  Aaron Wall has produced a lot of reasoning recently on why pricing for the folks who least benefit from what you have to offer is not sensible as a long term strategy, and I tend to agree with it.  (As of yesterday, I am also paying Aaron Wall $100 a month for training on a subject I already consider myself intermediate-advanced on, with the expectation that it will probably make me a multiple of that eventually.  More on that later.) 

But that is neither here nor there — the upshot is, as a reader, you have an opportunity to grab this for a fraction of its value to you.  I’d highly suggest it.  (I hear it is also going to be featured on Bits Du Jour soon, probably for a steep discount.  Have I mentioned that I think the price is far too low as it is?)  Bob also asked me for a testimonial prior to publication, which I’ll reproduce here.

For people who enjoy the challenges of being lost in the wilderness, I highly recommend learning to market software by putting up a website and tweaking it incessantly until you find some combination of elements that works. For folks who prefer knowing they will be able to make the rent check, I suppose you could read MicroISV Sites that Sell! instead. This is Marketing 101 written by engineers, for engineers — copious examples of what works, a focus on concrete actions to take over voodoo psychology, and actionable suggestions for the marketing novice.

As usual, Bob hasn’t paid me anything for either that testimonial or this review.  (Well, I suppose being asked to participate in the book counts as being paid a tremendous compliment, but no money has or will change hands.)  I only mention it because I sincerely think it is of value to many of my readers.

… make it the Search Value Pyramid graphic from Aaron Wall of SEOBook.  It’s so true it hurts, as I have to keep reminding myself every time I slip another place on [bingo cards] (oh no, my traffic numbers!) and gain another place on [need bingo cards for reading class] (oh yes, my wallet!)

One of my customers has had issues on five (!) XP boxes installing Bingo Card Creator.  That suggests to me either a problem customer or a borked executable, which has happened to me on GoDaddy in the past.  Sales have been slow for the last 3-4 days, too, which means it is just possible it is happening, but it installs and runs fine on my Vista machine.  Could anybody who runs XP quick install the free trial of Bingo Card Creator and tell me if you can successfully execute the program?  If you see the main screen and the popup suggesting you check for updates to the program, you’re done.

A friend of mine from high school is dropping by from tomorrow through next Saturday, and given that friends from high school do not routinely drop by Japanese rice fields, I will be showing her a good time rather than knocking a few articles off my to-write list.  See you in a week!

My year of GoDaddy hosting, which cost me about $93 if I recall correctly, expires on February 23rd.  I will be transitioning Bingo Card Creator from GoDaddy hosting to one of my Slicehost VPSes slightly before that happens.

This is not a rejection of GoDaddy so much as it is just a natural growth in where the site is going.  I am integrating the Bingo Card Creator and Daily Bingo Card sites, and am planning on doing some serious Rails-based work on the main site itself, starting with some backend improvements (conversion tracking the way Google should be doing it, for starters — “first click wins” not “last click wins”) and a few automated marketing experiments.  As a result, I have no further need for shared hosting, and want my $7 a month back.  ;)

One thing that gave me pause about doing this was an email disruption.  Happily, I was able to transfer email providing from GoDaddy to Google Apps for Domains without much hassle.  It took about 20 minutes to set up, all told.  Now I get to use the Gmail interface I know and (mostly) love with my @bingocardcreator.com addresses.  Hopefully Google will also end improve the deliverability of the emails, although since setting up an SPF record for my site and e-junkie I have had much, much less of a problem with that. 

Another thing which I’ll have to think about is whether my $20 a month Slicehost slice can take an increase in traffic by a factor of about 10 (immediately) and then some fairly rapid growth after Google gives Bingo Card Creator’s link juice and domain trust to the Daily Bingo Card cards.  I’m pretty sure it can, but will be serving key pages as static HTML until I know that traffic spikes won’t bring the whole business to its knees.

Sidenote: for those starting uISVs looking for hosting, I think GoDaddy is an excellent choice.  I think in 18 months there I’ve had one serious problem (unexplained corruption of the trial executable EXE, on and off, for two weeks) and one additional minor outage, which happened in the middle of the night and lasted for a few minutes.

… but not quite released yet.  I finally figured out what was giving Java problems with reading the 200 new bingo card files (stupid Windows Notepad inserting a byte order mark into UTF8 text, stupid Java not reading BOMs in UTF8 despite them being in the standard, stupid me for letting this delay a release by almost a month…)

I’ll be releasing it into the wild as soon as the site redesign is complete — ideally, sometime this weekend.  In the meanwhile, I’ve already put it on Daily Bingo Cards as sort of an early-warning-system for issues.  For once, I rather doubt there will be any.  If there were, my new and improved automated test suite would have croaked when doing the output run for all the DBC cards.

New in 2.0

• ~200 new bingo card Wizards
• better support for foreign character sets
• much, much better printing for many large words and phrases
• assorted bugsquashing
• eased some trial restrictions that were only causing problems as opposed to sales

In the crazy busyness of January I forgot about pre-announcing the price increase, so I’m going to put that off for another month.