1) Advertise real products or services. Seems like 3/4 of the ads I see on facebook are scams or vapor. Carefully vet your advertisers to make sure they're legit.

2) Make ads relevant. I'm a happily married stay-at-home dad. I'm not interested in dating ads, but ads for better bottles could get my attention. Keyword-driven or site-demographic-driven ads are a start. As dpcan suggested, browsers could allow us to tag ourselves and our interests, which would be even better.

3) Get rid of common annoyances: flashing ads, ads that eat CPU and grind my browser to a crawl, ads on sites I might visit from work or with my kids that aren't safe for work/kids, popups/popunders, ads that block my view of the content I'm looking for, and misleading ads. (Again, if you're a host, vet your ads!)

4) Do not compromise user privacy. Don't track my browsing; base ads off of the site I'm on or information I voluntarily tagged myself with.

Many of the webcomics I read use Project Wonderful, and they seem to have a pretty good model. I regularly see interesting ads, and almost never see offensive ads, scams, or CPU killers.

Although I agree with all of these points, I think relevancy is key. If it's determined by some means that I like "Antique Butter Churns", I'm hardly going to field like the ad to the "Antique Butter Churn Store!" website is actually an ad. I'd consider it a helpful recommendation.

If an ad platform could approach a point of such relevancy that users wouldn't see the ads as advertisements but friendly suggestions, that company would change the way the web works!

Item 4 makes item 2 harder to achieve - less relevance per unit of effort.

How? Don't serve ads.

Frequently on HN we get a lot of people who are great at building stuff but fall when it comes to finding and growing a userbase.

We then have people that are great at developing a community, have a popular niche blog or forum, and are itching to monetize their community.

Serving ads to solve this problem is like strip-mining for gold - ridiculous sediment to gold ratio.

So what if we concentrated on the affiliate model. Rather than serving ads to your community, you vet actual products and services that your community would actually use and genuinely endorse the product openly. You could run a drop-in type of "marketplace page" for your endorsements. It would be clear to your community that you receive (large) commissions on these sales, but I reason happy users would want to help your site. I'm talking 50%+ commissions so you can easily profit share in terms of offering a better price than they'd pay going directly anyway.

For product and service providers, I can't see a downside. They pay only from actual sales. The product is seen in a positive light (endorsed) rather than overlooked as an ad. And product providers could/should work closely to community moderators to iterate on their product.

As of now this seems like wishful thinking I know, but this could definitely be worked out, and provided it was dead-easy to get started from both ends, it's worth testing.

Make dropping in an affiliate marketplace easy for content providers. And make revenue sharing, management, and tracking easy for product providers. Develop community around connecting the two in a vetted environment. Profit off transaction fees.

Doing targeted advertising that is utility is the best way to make ads not suck.

The main reason advertising is annoying is because we are looking for something else, we don't need it right now, or its the wrong audience.

If you had a huge spare tire and wanted rock hard abs, and the advertisement could prove the safety of your purchase, you would say "HELL YEAH!"

When I was 16 the advertisement for Proactiv was very engaging. It worked and I was amazed. I'm no longer a zitty teenager so now I find their advertisements annoying.

I recently joined the Spiceworks team, and it is free software. It is advertisement driven, but most in the case where you need it. It is IT management software. As an example, if you see your SQL server getting slammed, you could see an ad that offers you RAM for that specific machine. Does this advertisement suck? Probably not.

One reason advertising could suck is due to audience. If you go to Digg (which has a wide audience) you are going to get wide or undirected ads. This is why targeted advertisement is all the rage these days - because its less annoying, and more to the point effective and win-win in more cases than not.

I'd say to make advertising not suck, don't involve yourself in communities without a common interest or goal. If you are an IT manager and you are in an IT community, you will more than likely get IT advertisements. If you are watching MTV... well...

I'm personally fine with most ads, even the irrelevant ones. What I don't like are the animated banners, and especially Flash monstrosities that pop up over the page you are trying to read and start making terrible sounds. They are the minority, but they are the reason I run AdBlock.

I put sites I like (and that have tasteful ads) in the AdBlock whitelist. I really wish there was a way to only block the annoying ads. Sounds like an idea for an AdBlock subscription list actually.

Yup, that's a feeling I have too. One thought is: how can you involve your community more to decide which ads are useful and get feedback?

http://reddit.com/r/ads is great for this

If these are lead generating ads, you could determine if a click lead to a sale. If you had this data (which was pointed out by the OP about tracking every click - although he insinuated its bad), you could determine what types of ads work and which ones don't.

So this is two fold, you remove the ads that don't work (so you don't annoy your audience), which in turn creates a more potent advertising experience for vendors which makes that real estate more attractive.

The trick is not letting every dog through the door and keeping some class.

Imitate The Deck (http://decknetwork.net/). They produce beautiful, low impact, relevant ads which I frequently consider carefully and often click-through on. They do an amazing job and I've deliberately reduced the power of my adblock blacklists to see those ads.

An example ad: http://hivelogic.com/

They're clearly not even attempting to scale, but they have the ideas that -- if brought to scale -- could make advertising less sinful.

I look forward to using them for GridSpy. Thanks for the recommendation.

We should be able to tag ourselves within our browsers.

When I install FF, I should be asked what applies to me (tech, baseball, family) and then all advertising can look to this public info first and display what's relevant only.

THIS should replace ad-blockers. Then we don't have to destroy our favorite sites' revenues and I don't have to see any prescription medicine ads.

And get rid of ads I have to "skip", ads that "pop-under", and ads that cover what I'm supposed to be reading.

> Google does well because the ads shown on search bring forth some form of utility.

As far as my personal experience goes, I can't remember a time I encountered a Google ad that was useful for me, unless I was searching for "Amazon.com" and Amazon.com shows up in the sponsored links.

But to answer your question, ads can not suck by simply being perfectly targeted. Even though perfect is impossible, that should be the goal of all advertisers nonetheless. Get as much info about people, and send them only ads you know will be useful. Advertiser wins because they got the word out about their great thing, consumer wins because they got a great thing. The only thing this ideal wouldn't help is a shitty product. Shitty products aren't great, so if they get undeserved attention, they just screw over more people.

A friend and I developed a proof-of-concept system in 2006 to address this opportunity. The central concept was that publishers would receive offers via our advertising marketplace, and then use their own selection algorithm to select ads to display (publishers could choose to use our hosted selection tools, use our open-source selection algorithms, or build their own). This would allow the publisher to use any combination of factors, such as ad format and perceived relevance -- not just eCPM. It is a flexible system that give the publisher the final decision to display or veto an ad.

If anyone wants to read the whitepaper, feel free to send me an email. The truly adventurous can check out the patent application (#11948995).

Try a classifieds model where content providers don't place any advertising on their pages at all. They then put a "Classifieds" link on desired pages, that takes you to another page filled with targeted advertising relevant to the article or site as a whole. As a bonus, let the content providers have an opportunity to select or vote on ads to reward quality and improve relevancy. The classifieds page doesn't even need to come from the referring site, as it is completely separate from the content. If advertising is so wonderful, this model can't fail. It's less intrusive and encourages the advertising industry to provide something of value.

Some things to consider for me:

- Don't put ads where content should be, I hate when I'm scrolling and there is an ad where I should be reading (unless it is displayed at the end of an article)

- Put ads in a designated area, and stick to that single area

- Only display relevant ads if your site is about web design I don't want to see ads about belly fat or ebay

- No flash or animated ads, this is so annoying and probably makes me leave sites the most because it is just distracting from both your content and my train of thought

- Don't display too many ads, this is also a major problem and just makes me assume that there is no real content and you just want to make money (news sites seem to do this a lot)

1. Get rid of scam/cheating ads (rebills etc) and also pop unders

2. Don't autoplay audio/video on the ads. Heck, autoplay is bad for content, its super annoying for ads.

3. Increase relevance

4. There are sites where I need to search the content inbetween ads. Would we tolerate if just 30% of our TV screen was video, and the rest were ads? (text, scrolling text etc). Maintain decent ratio between content and ads.

5. Most websites don't even bother to experiment with other ways to make money, like subscription etc. quizlet turns off ads for a donation of 10$ (donations get some more features). Seems like a good experiment. May be others can try that? this could also be an option.

It's impossible to make ads not suck. Fifty years ago, perhaps ads didn't always suck, before we had Metacritic and Amazon reviews and a million professional critics and reviewers writing online and social media telling you your friends' opinions about everything. All of those things are way more useful than an ad, no matter how well you target it. And soon, with augmented-reality stuff, they will be easier and easier to see whenever you are thinking about buying something.

With all that, what the hell is the point of ads? They're just wasting the time of the people producing and consuming them. They do nobody any good at all.

> How can we make ads not suck?

If there was some kind of nonprofit "respectful advertising alliance" that enforced consumer-friendly practices and provided a whitelist to import into adblock plus, I might use it. It would be hard to define what is consumer friendly and enforce the rules, though.

The only site I've disable adblock on is Reddit. This is because their ads use inside jokes and are sometimes humorous and always non-intrusive.

I don't see what is so nice about Google ads. They look crappy on most sites and are rarely useful in my experience. Also not a fan of Google's tracking with adwords.

What about voting/rating ads? I've seen it in a few places, and it seems so simple -- someone has to have information on how it worked out in practice.

I would expect that having an easy, obvious rating meter on each ad would work in two ways. First it could improve targeting by having the user reject ads on a case-by-case basis. Second, ads that get rejected by a wide range of user types can be determined to be bad in general (ugly, offensive, annoying) and thus removed from circulation, automatically.

Does anyone know why this isn't in place more? (Or perhaps I just haven't noticed it?)

What other verticals can you apply "thedeck" model to?

I think it comes down to specialized, targeted advertising outlets. A perfect example is what FusionAds is to the developer community. I enjoy clicking those ads (although I think they should allow more websites to display the ads).

I like the 125x125 image ad format. It's just big enough to get my attention, and small enough so the publisher can stack in a few of them without taking up too much space. I think it's been gaining a lot of popularity lately for those who sell adspace directly to businesses.

A few examples that come to mind

http://betterexplained.com/

http://theoatmeal.com/comics/coffee (bottom of the page)

http://37signals.com/svn/

My concern with asking "How can ads not suck?" is the limitation of the answer. If I take one step back from the question, my view of the problem shifts and the potential solutions grow. Your concern isn't about ads and ad-blockers, it's about generating revenue from a site. You think that ad-blockers are affecting your revenue. They aren't. How you're advertising is affecting your revenue.

I'd start with the question "How can I generate revenue without using ads?" Others may have better questions.

I've taken numerous conscious and subconscious decisions to eliminate ads from my life. I don't watch broadcast television and I don't listen to broadcast radio. I don't regularly visit sites that whore out their pixel space (and even the content pixels with floating ads). It's easy for me because I know where to get good content without these media channels. I like to believe the net result is less fatigue at the end of the day.

Bring down the editorial/advertising wall. Seems to happen a lot online but I think it's going to explode even in the established media once the efficacy of advertising tumbles. More payola, more "product placement", more links within the actual content to advertisers..

Industries tends to find a way to win (or at least fight hard). Our content is going to get a lot worse rather than the advertising better, IMHO.

A lot of the ads I see that "suck" are for scam. Most, if not all of those ads are driven by rebills. If we can get rid of rebills, we can get rid of a lot of the crap ads.

The reason for that is that advertising is market driven, those that make the most money off a user are the ones that will pay the most for a spot.

Rebills and upsells are a way to make a lot of money from a single signup so that's why those advertisers pay top dollar.

And as long as the credit card companies allow scammers to operate this will continue.

I think one secret is community-based ads. I'm happy to receive ads about products directly related to sites I frequent. For example, things produced by YC... I would be happy to peruse an "ads" tab (next to "jobs"/"submit") by my own volition. 2600 magazine does the same thing... ads by subscribers only... and they're always interesting. That's the key. Find ads that people WANT to see, so much that they're almost willing to go out of their way to see.

Then again, if you really have high enough volume, why can't you just try EVERYTHING and then only show ads that get high CTR? Isn't 5k impressions a sufficient sample size?

What about using graphs of campaign clickthru rates for individual people (a la directededge)?

For every youtube video, i should see ads on when the artist is next coming to town by me. Find out where i live by simple ip address geolocation. Get commissions on ticket sales. If the artist isnt coming soon use the panoply of recommendation systems to just find out when similar artists are comin instead

Ditch automated ads, ads that try to steal our focus away from the content, and ads that hog up our bandwidth, screen space, CPU time, etc. They're annoying, usually irrelevant, and serve no good.

When you write, include in your writing very brief endorsements for products related to your subject. Don't just randomly interject with "oh by the way, go buy Product X!", but if Product X is a brand of hard drive and you're writing about hard drives, take a sentence or two to talk about the advantages of that particular brand. In plain text, just like the rest of your article.

There are TWO important parts to this equation. Many people only seem to realize the first. 1) Make your ads not so obvious that machines can tell they're ads. 2) Make your ads not so annoying that people will update their ad blocking rules to remove them.

In other words, working around the ad blocker to present annoying ads is just going to annoy even further. Your ads have to be seen AND they have to not be immediately blocked again and/or just ignored, and that means they have to be manually inserted so that they are intertwined with the content from both a technical and a writing standpoint.

You might even get away with static banners, that don't stand out too much, at the top and bottom of a page (where we can scroll it off the screen), and text links next to the content area. Just not IN the articles, unless it's part OF the article.

Why did I install an ad blocker in the first place? Because ages ago on gameshark.com, an ad popped up, demanding I join the US Army. I've never left Canada and I was maybe 13 at the time. This ad, for 30 seconds, played an animation with loud sound effects, hogging up all my CPU time and bandwidth (now I knew why the page had loaded so slow), and centred itself under the mouse cursor at all times so I couldn't click anything. It was a Flash layer of some sort, rather than a window, so no keyboard shortcut would be able to do anything either. When this ad finally disappeared, I IMMEDIATELY looked up and installed an ad blocker, and never again went to gameshark.com. If I'd had any interest whatsoever in joining the US Army, that also would have killed it permanently. This is the perfect example of how to do EVERYTHING WRONG with online advertising.

PS: one of the worst comment systems I've ever seen. Requiring registration? Fail. Not stating this until AFTER the comment is submitted? Massive fail. Not stating the requirements on the form until they're violated? Cherry on top.