Re­cently, Prac­ti­cal Ty­pog­ra­phy com­pleted its third year on­line. Last year, in my sec­ond an­nual re­port, I de­scribed hav­ing to face the co­nun­drum shared by many on­line pub­lish­ers: how to con­vert a mas­sive flow of un­paid read­er­ship into dollars.

To ad­dress this prob­lem, I tried an idea that I was pretty sure wouldn’t work. But to my sur­prise, it did. More about that below.

First, how­ever, I want to thank all the read­ers who have paid for this book over the last year—ei­ther with a di­rect pay­ment, or in­di­rectly, by li­cens­ing my fonts. You’ve made it pos­si­ble for me to keep im­prov­ing this book, along with its un­der­ly­ing pub­lish­ing soft­ware, Pollen.

When I launched this book, I said it was “partly an ex­per­i­ment in tak­ing the web se­ri­ously as a book-pub­lish­ing medium”. In turn, I called upon read­ers to join me in prov­ing that this ex­per­i­ment—re­ly­ing on vol­un­tary sup­port by read­ers, rather than DRM and cor­po­rate in­ter­me­di­aries—can work. I’m sat­is­fied that it has. In fact, I’m suf­fi­ciently well per­suaded that I’m work­ing on a new book that will be pub­lished on­line in a sim­i­lar way.

Read­ers who pay for books are the best read­ers in the world. With­out you, there won’t be any more books. Thank you for do­ing your part and sup­port­ing the writ­ing you find valuable.

By stim­u­lat­ing font sales and di­rect pay­ments, Prac­ti­cal Ty­pog­ra­phy gen­er­ated rea­son­able rev­enue in its sec­ond year—about $11,155. But that was a de­cline from its first year, when it gen­er­ated about $14,926.

I had ex­pected rev­enue to go down. What I hadn’t ex­pected is that in the sec­ond year, traf­fic went up. That meant that a smaller per­cent­age of read­ers were sup­port­ing the book with any kind of pay­ment. Why? And was there a way to re­verse the tide?

In a non-sci­en­tific way, I the­o­rized that the prob­lem was the flow of “ef­flu­ents”—high-vol­ume, low-qual­ity traf­fic—pumped in my di­rec­tion by web­sites like Cre­ative Bloq and Buz­zFeed. Since the be­gin­ning of the web, it’s been an ar­ti­cle of faith that more traf­fic is bet­ter. Clearly, that hasn’t been true for a while. We’re deep into the click­bait econ­omy. The con­tin­ued dom­i­nance of ad­ver­tis­ing, and the neg­a­tive-se­lec­tion ef­fects of ad block­ers, have meant that the web in­creas­ingly caters to those with the short­est at­ten­tion spans.

As I saw it, the chal­lenge was how to con­vert more of those short-at­ten­tion-span read­ers to pay­ing cus­tomers, with­out turn­ing Prac­ti­cal Ty­pog­ra­phy into a non­stop pa­rade of wheedling, beg­ging, and nagging.

So what hap­pened in the third year?

In year one, I got about 650,000 read­ers. This in­creased to about 700,000 read­ers in year two. In year three, how­ever, traf­fic went down to about 615,000 read­ers—roughly a 13% drop. Since I track these num­bers with Google An­a­lyt­ics, it’s pos­si­ble the drop is over­stated be­cause of the ac­cel­er­at­ing rise of ad block­ers (in­clud­ing the con­tent block­ers sup­ported by iOS 9, start­ing last Sep­tem­ber), which of­ten block Google An­a­lyt­ics as well.

But con­sis­tent with my com­ments last year, I think it’s wise for web au­thors to de­tach them­selves from their traf­fic fig­ures (or Twit­ter fol­low­ers, or Face­book likes, or Stack Over­flow rep, or GitHub stars). Ex­po­sure doesn’t have any pre­dictable re­la­tion­ship to money. Never has.

In­deed, if you be­lieve that it costs money to serve web pages—and yes, it does—then maybe the best out­come is to col­lect more money from fewer read­ers, as it sug­gests you’re get­ting bet­ter at find­ing the peo­ple who will pay for your work.

As in the pre­vi­ous two years, font sales were the foun­da­tion of the book’s rev­enue. In year one, I es­ti­mated there were 450 font trans­ac­tions at­trib­ut­able to Prac­ti­cal Ty­pog­ra­phy, and im­puted a $25 “re­fer­ral fee” to each, for a to­tal of $11,250. In year two, 325 trans­ac­tions, with im­puted rev­enue of $8125 (an 18% de­cline). In year three, I es­ti­mate the book stim­u­lated 310 trans­ac­tions, for im­puted rev­enue of $7750—less than a 5% de­cline. So that’s pretty good.

The big change was in di­rect pay­ments—where I ask read­ers to vol­un­tar­ily send be­tween $5 and $20, as be­fits their ap­pre­ci­a­tion of the book. In year one, I col­lected 321 pay­ments to­tal­ing $3676. In year two, 244 pay­ments to­tal­ing $3030. But in year three, I col­lected 489 pay­ments to­tal­ing $5472—a 100% in­crease in pay­ments, and an 48% in­crease in revenue.

Not bad! That means that to­tal rev­enue for Prac­ti­cal Ty­pog­ra­phy in­creased from $11,155 in year two to $13,222 this past year—an 18% rev­enue in­crease de­spite a 13% traf­fic drop. A to­tal of 799 read­ers opened their wal­lets, which is bet­ter than the first or sec­ond year.

Of course, this is not life-chang­ing, quit-your-day-job money for any­one. But as I said in my first re­port, it’s un­rea­son­able to ex­pect book pub­lish­ing on the web to ex­ceed the eco­nom­ics of print pub­lish­ing. As I said then, sup­pose the typ­i­cal non­fic­tion book sells about 1000-3000 copies over its life­time, with the au­thor get­ting about $2-3 per book. Even lean­ing op­ti­mistic, that’s to­tal rev­enue of only about $5000-8000. So if the ques­tion is “would I have been bet­ter off pub­lish­ing Prac­ti­cal Ty­pog­ra­phy as a pa­per­back?”—the an­swer re­mains, em­phat­i­cally, no.

At the end of year two, I thought about ways I could en­cour­age more read­ers to pay. As I said above, non­stop wheedling and beg­ging was out—that would’ve made the book per­ma­nently ugly & an­noy­ing. Not really what I’m go­ing for.

A few self-styled web-pub­lish­ing pros sent me well-mean­ing ad­vice—like in­te­grat­ing such-and-such app into the web­site, or start­ing an email newslet­ter, so I could ha­rass read­ers more of­ten, etc. These are nice ideas. If they work for oth­ers, great. But my time is lim­ited. I can’t spend dol­lars to chase dimes.

What I re­al­ized af­ter last year’s re­port is that these mas­sive flows of un­paid traf­fic em­anated from only about 15–20 web­sites. Also, re­call­ing the Na­tional Lam­poon cover, I won­dered if maybe a life-or-death ap­peal might move the needle.

So I set­tled on a strat­egy of gen­tle, tar­geted sham­ing. I added ex­actly one page to this web­site, and then up­dated my .htac­cess file so that vis­i­tors from only those 15–20 sites would see that page when they arrived:

HELP KEEP THIS BOOK ALIVE
(click to see the whole thing)

A lit­tle dra­matic? Maybe. So what? It’s all true: I am sick of bot­tom-feed­ing click­baiters and so­cial-net­work­ing bil­lion­aires prof­it­ing from my ma­te­r­ial, while con­tribut­ing noth­ing in re­turn. (Re­mem­ber: ex­po­sure isn’t money.)

“Hey man, that’s just the way the web works. Haven’t you heard? In­for­ma­tion wants to be—” Yeah, I’ve heard. How’s that work­ing out for you? Do you like the web you’re get­ting for free?

Don’t worry—as a web au­thor, I know bet­ter than to yell at clouds. The click­bait econ­omy will al­ways be part of the web, just like ad-sup­ported net­work tele­vi­sion. I ac­cept it. (Though to be fair to net­work tele­vi­sion, they do pay for what they broadcast.)

But I con­tinue to be­lieve that there’s room for more than just that. Un­for­tu­nately, when we as read­ers click from link to link, and noth­ing is ever asked of us, some­thing gets lost in trans­la­tion: that what makes the web good at all are the in­di­vid­ual hu­mans who share their writ­ing, their pho­tos, their mu­sic, their videos.

“Help keep this book alive” is just a way of re­mind­ing read­ers that the in­for­ma­tion here wasn’t gen­er­ated by a piece of soft­ware. There’s a per­son be­hind it. And if the click­baiters are com­fort­able sell­ing ads against my ma­te­r­ial, why shouldn’t I be equally com­fort­able ask­ing those read­ers to pay some­thing? You could call it a quixotic pur­suit, ex­cept for one thing—

It worked.

—Matthew But­t­er­ick
5 Aug 2016