Currently Google Adwords (Google Search only) and Microsoft's Live.com. Those two are the only options I've found worth bothering with for B2B software at this time.
We tried Yahoo: It's a complete waste of time and money. Ads were always edited by Yahoo staff (requiring frequent e-mails to clarify the service we were paying for), bids may be changed at random, the interface is a nightmare kludge and there is no way to opt out of Yahoo's 'partner network' (which seems to consist only of spam domain landing pages), so bidding a keyword to the top of the Yahoo search results guarantees a load of expensive visitors from click-farms. I would love to advertise on Yahoo's search result pages and will do so the moment Yahoo Search Marketing is taken over by a competently run organization (Go Microsoft!).
We tried Ask.com: You'll receive a tremendous number of visitors on your first day, all from spam domain landing pages. There is no way to specify "place my ads on Ask.com search results only" (which I'd genuinely like to do), so money given directly to them is wasted. Crazily enough, you can cancel your Ask.com advertising and opt into using Google's "Search Network" instead. That will put your ads on Ask's search results without having to deal with Ask.
Mechanical Turk is good for getting initial users to use your site and "test it". Offer them $.10 to test your site and give you some feedback on the site. It's a good way to get initial beta testers.
Are you able to set a top-limit cap on the amount you spend or the number of users?
I'd be interested in hearing real numbers, in regards to how much you had to invest in Mechanical Turk to get a decent response to further your application.
I have tried AdWords and Facebook. They worked in the sense that people visited my site, but they did not work in the sense that no one bought anything from me; however that may be due to my site and product, not the advertising.
I think internet advertising for web apps is pretty much useless. Sure it might drive you some traffic, but you most likely won't get a lot of sales. So unless you have a $100 profit margin on each user, you won't recover the money that you'll spend on ads.
(Posting from a junk account just to keep things a bit more anonymous.)
The company I work for spends over $50,000 daily on advertising, for a purely virtual (downloadable software) product. This is almost, but not quite, exclusively search advertising. Dollar-wise, over the past few months, the split is:
Google: 87.2%
Yahoo: 8.8%
MSN: 2.4%
Ask: 1.0%
With LookSmart, CJ, NeverBlue and Right Media taking up the sub-one-percent slack.
This is (like other people have noted) largely influenced by what our product is. Search marketing (as opposed to display) is by far best for us, and Google is by far best of search. There is a chance there are some campaigns I've missed.
Depends on your product some. You might also look into MySpace advertising-- it's wicked easy to try out in their beta ad manager, but I find that the CPC is pretty high for toying around with at first.
StumbleUpon seems to be a good one for interesting/clever sites in terms of conversions, and the effective CPC is around $0.05.
With Adwords, there's some fun things digging through and finding keywords your competitors aren't really looking at. There's also a lot of people overlooking keywords in unconventional ordering of the words.
Adwords has worked for my business. You really need to invest time and money into it. It was only after my first $3k and 3 months that I started seeing it work.
With the proviso that its going to either take a LOT of time or using Conversion Optimizer, I suggest taking a look at Google's (AdWords) Content Network. It is far and away the most effective advertising I am able to do, consistently clocking even AdWords on google.com
We tried Yahoo: It's a complete waste of time and money. Ads were always edited by Yahoo staff (requiring frequent e-mails to clarify the service we were paying for), bids may be changed at random, the interface is a nightmare kludge and there is no way to opt out of Yahoo's 'partner network' (which seems to consist only of spam domain landing pages), so bidding a keyword to the top of the Yahoo search results guarantees a load of expensive visitors from click-farms. I would love to advertise on Yahoo's search result pages and will do so the moment Yahoo Search Marketing is taken over by a competently run organization (Go Microsoft!).
We tried Ask.com: You'll receive a tremendous number of visitors on your first day, all from spam domain landing pages. There is no way to specify "place my ads on Ask.com search results only" (which I'd genuinely like to do), so money given directly to them is wasted. Crazily enough, you can cancel your Ask.com advertising and opt into using Google's "Search Network" instead. That will put your ads on Ask's search results without having to deal with Ask.