Some people hate advertising in general and despise Web ads in particular. I can certainly sympathize; my inbox overfloweth with spam. I fondly remember the good ol' days of Netscape 1.0 when the Web was a cozy, ad-free place mostly populated with hobbyists and college students. And I'm well aware that Corporate America has done damage around the world by promoting mindless consumerism, harmful goods and pernicious social and sexual stereotypes through advertising; kids are particularly vulnerable. For instance, researchers have found young girls often develop eating disorders the more they're exposed to advertising (references). Google ads are filled to the brim with scams that prey on the naive (I personally see red every time I see a vanity publishing scam).
But your book isn't a scam, is it? You worked as hard as you can on it, poured your life into it. It's not wrong to tell people about it.
Will your book get negative backlash if you take out a web ad? If it's an attractive, honest ad, and you don't, say, advertise your erotic horror novel on a children's cartoon site, probably not.
But there's still a risk. Most people who hate web ads with a passion do the logical thing and install ad-blocking software, regardless of whether this hurts the sites they enjoy or not. A few of them will condemn any product -- good or bad -- that is promoted through advertising. They condemn any site -- good or bad -- that hosts ads.
So if you think the true target audience for your book is mostly composed of pedantic, judgmental lit snobs, then yeah, you might want to avoid ads altogether. And if that's the case, "Hey, guys, my new book came out!"-type message board promos aren't going to pass muster with that crowd, either. You're probably stuck waiting for word-of-mouth to materialize.
Fortunately for me, those folks probably wouldn't ever pick up my humor collection, so the publisher and I decided to give web ads a try through the Project Wonderful system. PW mainly runs on web comics sites but also delivers ads to speculative fiction magazine sites like SFReader and Greatest Uncommon Denominator.
It's been a learning process for sure -- one big thing I learned is that web comics readers actually do buy "real" books. Another thing is that actually paying for ad space often yields better results than participating in free banner exchanges. Most people who host an exchange banner stick it down at the bottom of their pages where hardly anyone will see it. Conversely, people who are participating in Project Wonderful often put the ad slots in highly visible locations so that they'll be worth more. The trick is to hit a good balance between cost and exposure.
Anyhow, since the publisher's refined his tactics, monthly sales for my book have quadrupled. There are a lot of elements you have to orchestrate to have a successful web ad:
1. You need an attractive ad.Every ad system allows you to run text ads, but text ads in my experience are a usually a waste of time and money. People just tune them out. Graphical ads do much better. Colorful ads do better than monochrome ads, and ads that move do better than static ads. This is all Psychology 101 stuff -- we're wired to pay attention to movement. The key is to avoid obnoxious colors/movement. Having an ad move too slowly to catch the eye is bad, but setting its frames to flash by too quickly to read is worse. We've all seen those horrible mortgage ads featuring panting sows and dancing people from the uncanny valley. Obnoxious. Avoid.
Some sites are so put off by moving ads that they'll only take static ads. Furthermore, different sites have different sized slots (see PW's templates page for more info). So, you'll need multiple ad sizes, and this is where graphic design skills and proficiency with programs like Photoshop and Adobe Image Ready come in handy (GIMP is free and great for static images, but I've found making animated ads is far easier with Image Ready). If you don't have these kinds of skills, and if your publisher can't provide ready-made ads, you'll have to hire a graphic designer. If your budget is limited, at the very least get a banner ad and a leaderboard ad. A square ad wouldn't hurt, either.
Aside from fitting into different ad spaces, having a variety of ads at your disposal lets you see what works and what doesn't. A specific ad might get a listless response at one site, but work very well on another. So, get multiple ads made if you can.
What should you put in an ad, or tell your graphic designer to put in the ad? The title, cover art, and author names are mandatory -- you want these things prominently displayed so that potential readers who don't click through might recognize the book and pick it up if they see it in a store later. Short, lively review excerpts are good -- no more than a dozen words in a single frame. Less if possible.
It helps if your ad has a "hook". Why would people would want to read your book? Is it funny? Exciting? Scary? Sexy? Informative? Try to convey that in the ad. The goal here, other than to make people aware that the book exists, is to entice them to actually click on the ad to learn more about it.
2. You need an ad host.Some popular book sites, like Smart Bitches, Trashy Books or Ralan.com, provide their own ad hosting and offer flat rates for buying ad that will appear on their site for a set period of time, usually a month at minimum. While these $50-for-a-month type arrangements can be a good deal, you're locked in -- you can't usually swap an ad out for a different one if it performs poorly, and you surely can't cancel it and demand a refund unless the site goes down or there's another problem on the host's end.
Furthermore, you'll experience click-through attrition on ads that stay up a while on sites that have a regular daily readership. You might get 100 clicks the first day, 75 the next, then 50 the next day, and so on until you're only getting a few clicks each day toward the end of your ad month.
So, there's a lot to be said for sites that take ads through auctions like the system Project Wonderful offers. A banner ad slot on a popular site might be going for $50 a day ... but my publisher doesn't have to buy the space for an entire day. He can just bid up enough to secure the ad during lunchtime, spend a few dollars to get a couple thousand exposures and a couple hundred click-throughs, and then cancel the ad and let someone else have their turn in the slot.
Once you're buying space through an auction system, you'll start considering taking out ads on sites that are new to you. Use basic common sense in evaluating them. Does the site get a lot of traffic? Are the site's visitors likely to be interested in the genre you're writing in? Do many of them appear to be readers? Are the ad slots featured in visible places, and are they limited so that the page isn't crawling with competing ads?
3. The ad needs to go someplace usefulWhat do you want to achieve with the ad? Do you want to promote a particular book, or do you want to promote your whole catalog? Whichever you want to do, make sure that the page the ad sends people to is attractive, informative, easy to navigate, and loads quickly. Good review blurbs are a must. Free samples of the book are extremely useful - few people are willing to take a chance on an author they've never heard of before if they don't have the chance to read some of it first.
In my experience, you get the best results if you send people to a place where they can read excerpts and then actually buy the book, such as its page on a major, consumer-trusted seller such as Amazon or Chapters. This works best if the book page has at least a couple of positive reviews featured on that page. So, when you or your publisher are sending out books for review, don't ignore people who mainly post on Amazon or other bookstore sites. They do have value.
Barnes and Noble and Amazon pages also have additional value in that they provide sales rankings. While these rankings involve a lot of secret voodoo and are hard to translate into real numbers, they do give you an idea if a book has actually sold copies or not that day. And being able to track sales greatly protects you from click fraud and other shenanigans from dishonest ad hosts.
4. You need to monitor your ad's performanceOkay, you don't have to monitor an ad's performance. If you're awash in cash but not much time, you can just throw money into ads and hope for the best. But if your budget is limited, it helps to pay attention to what works and what doesn't.
You need to know some jargon going in. A CPM is not a type of missile, and CPC is not an ozone-destroying chemical. CPM is the cost for 1000 loads of the ad on the site's pages. A raw CPM refers to overall page loads; a unique CPM refers to loads presented to individual visitors as determined by their internet address. CPC stands for "cost per click". Paying for ads based on pure click-throughs is a bad idea (see the next paragraph) but watching the number of click-throughs in conjunction with the number of impressions is helpful for determining if the ad is reaching an interested audience or not.
A lot of potential advertisers are concerned about click fraud -- that is, a single person spoofing different IP addresses to make it seem like real site visitors are actually clicking through when they aren't. A slightly lower-tech version of this is a webmaster who's enlisted minions to click on new ads as they appear on the site. This is a legitimate worry. But if the site is high-traffic and it seems legitimate (ie, not a link farm) you probably don't have to worry too much, particularly if you give the ad a test run and see that it's generating results in terms of your book's sales ranking. We use Titlez to keep track of Amazon rankings, and it's been very handy for the purpose.
Remember Richard Dansky's observations about potential reader attrition? That applies just as much to web ads, but at least there are fewer steps for them to go through. Of the people who see an ILDB web ad on a high-traffic science fiction comic site, maybe 2% or 3% might click on it. Of the people who click through, another 0.5%-3% might actually buy the book then and there (the statistics I have indicate that more people apparently come back later and buy the book from another vendor, or put it on their wishlist, or bookmark the site, etc., but for direct sales it's between 0.5%-3% depending on how the wind's blowing and if Mars is ascendant and Venus is in the House of Pancakes).
By my calculations, if 5,000 people on the aforementioned comic site load up a web ad, 2-5 of them will end up clicking through and buying the book in one fell swoop. So if an ad goes out to 2,000 web surfers and the Amazon numbers don't improve a little that day, we know the ad isn't working.
Ultimately, you need to run some test ads, then crunch numbers. How much is a single sale worth to you? You need to know this before you can go further. Figure out how many clicks and exposures you probably need to get to earn that single sale (and bear in mind that some days you'll do all the right things and still not get a bite). If the ad on the site is going to cost more than you'd get for that sale -- don't buy it.
I've seen Facebook ad hosts bragging about a fifty-cent CPM - that is, a buck for 2000 ad impressions. In my experience, that's actually pretty bad, unless the site's visitors are mainly composed of the sort of people who buy the kind of books you're selling. My publisher generally ditches an ad that gets worse than a $0.30 CPM, unless it's highly-targeted. It's entirely possible to find CPMs of 10 cents or even less in Project Wonderful.
Likewise, he looks for low costs-per-click. More than 10 cents is bad, 3 cents is good, one cent is awesome and if you find an ad at that rate, you should ride it for a while to see what happens. At three cents per click, my publisher spends $3.00 to have 100 people hit the book's page. If just one person buys the book, he's broken even; if a statistically-probable 3 people buy it, I've personally earned enough in royalties to go get a burrito.
And that's pretty cool.
I hope you've found this introduction to web advertising helpful. There's no one-size-fits-all approach to this, so it's all pretty daunting at first, and it's definitely work, but it can generate good results. And you don't have to stand up in front of a crowded room of strangers or drive five hours to do it.