Build Your Own Ad Network

Last year, Darren wrote a great article on what he wants from an ad network, and I’ve been re-reading that post this morning as I jot down some ideas and fancies of my own. I’d like to revisit the topic today, and ask Performancing bloggers what they want from an ad network?

Its a difficult topic. It’s easy to grumble about adsense, or moan about blogads, but it’s quite another kettle of fish when it comes down to actually working out a better system….

Right now, there are two things that I really want out of a network. Actually there are lots of things, but these are the big ones:

  1. For it to be a good earner
  2. To be treated as a business partner, not another number in the system

Have a read through Darrens post above, it was written right at the start of Performancing and likely you haven’t seen it. Then, help me get this conversation started by telling me what it is YOU want from a network.

22 thoughts on “Build Your Own Ad Network

  1. You’re right, Brad, but the other option, which magazines often offer for smaller businesses, is ad design services. Not sure if Internet ad networks offer it, and I’m not sure if Nick and company want such a burden on top of a running an ad network, but that would be extra revenue for them. No doubt there are tons of designers out there, and several here at Performancing, who’d like the work. It’s typically a one-time cost, or it’s worked into the ad price.

  2. Mini-banners can be effective if your advertisers are mostly companies, they usually have somebody on tap that can make up a banner from stock corporate logs etc. However, if you are targeting individuals as advertisers then the text ad has advantages because there are fewer barriers to purchasing an ad run like knowing how to use Photoshop yourself, asking a friend to make a banner for you or buying one. Text ads have the advantage as an impulse buy for one man shops looking to try advertising on your site.

  3. I like the small (non-animated) banner ads at boing boing. These are 150×60, but could be 120×90. While the site has a lot of them, I mostly ignore them consciously. However, sometimes one will catch my eye and I’ll click on the ad. It’s much more like reading a print magazine, and I don’t have to look at a sea of text ads proclaiming all manner of mad things.

    These mini-banners seem more laid back, and allow brand-recognition – which is a bonus for the advertiser. I’m actually surprised that Google has nothing like this. Adult sites *cough cough* have a lot of these and they seem rather *cough cough* effective.

  4. There are bound to be attempts at fraud in any system, would it be any more vulnerable than adsense with click fraud, cpm systems with page view generators, etc?

  5. it’s cheaper to build a contextual network that is automated rather than a quality-assured, flat fee system.

    If one were to go that way, it wouldn’t be possible to grow beyond a certain point without sacrificing either quality, response time or add to the overheads (because you would need to hire more people).

    That’s why community participation in quality control would help – however, such a system would be easy to scam if the advertisers weren’t careful.

  6. I agree that the pay-per-click model is flawed and tends to corrupt content. If we step back and think about what we’re all trying to accomplish, it’s easy to see that this model isn’t ideal. For one thing, people may click on ads just to see what they’re really about, especially if the publisher goes for the link format rather than the text ad. While this may benefit the publisher in some cases, it doesn’t really help either the visitor, whose time may be wasted, or the advertiser, who may end up paying for nothing.

    What we really want is a great match between audience and products or services advertised as well as visibility for the advertiser. That is the holy grail, and it is worth a lot of money. So the more we can narrowcast, and the more we know about our audience, the more value we as publishers can deliver. Then we’ll get the right advertisers and we won’t have to rely on people clicking through.

    As if it’s so easy.

  7. I dont think there’s any doubt that contextual ppc corrupts content, it’s a pretty well established, and verifiable fact as far as im concerned, though i’d love to find a solution to it…

    I also like the flat fee models, and am taking a (another) look at adbrite and similar programs this week. I like the idea that the branding is paid for aswell as the clicks, i feel rather short changed with ppc models….

  8. I have already said that I feel adsense could be bad for bloggers. One of the things I like about text link ads is you get paid just for having them there, regardless of what your visitors do

    I think you have hit it with that. PPC advertising models favor the advertiser over the publisher because they get all that branding exposure for free on the site even when nobody is clicking. I also think that PPC ad models corrupt the content over time as articles on blogs start being sculpted for Adsense and getting people to leave by clicking on an ad. Getting paid by encouraging people to leave is a flawed business model.

    I’ve been using Adbrite on several sites and I prefer their model of being paid for time length of ad run as opposed to CPM or PPC.

  9. What is your preferred payment model? Pay Per click is the current darling of advertising networks but what do you think about CPM or Affiliate types (lead or sales commission, etc)? I have already said that I feel adsense could be bad for bloggers. One of the things I like about text link ads is you get paid just for having them there, regardless of what your visitors do, and you don’t have to change your template to accommodate them

    Also do you prefer textual or graphical? I dislike animated banners but sometimes its nice for your blog to not be a sea of text. Having CSS control over ads though would be nice.

    Do you like advertisers delivered to you automatically or would you prefer to pick and choose? For most of my sites in the past I would leave them on autopilot, but for others I would be much more hands on which would extend to being picky about advertisers. Come to think of it, the option to approach advertisers would be nice also, particularly if I got rewarded over and above the advertising fee..

  10. As I commented to the Tada lists post, ability to go back is important for any new product. To me it would raise the ease-of-use by an order of magnitude if I could try the new network in 2 minutes and go back to AdSense if I don’t like it.

    How to do it? Hmm,
    – What about having the blocks of absolutely same size as the AdSense has (in addition to own formats), so that I wouldn’t have to reformat my templates yet another time?
    – Or what if I give you my AdSense code and you give me yours to replace?
    – Or could you make it as simple as replacing with in the AdSense blocks?
    – Or maybe you can even provide a tool to statistically compare revenues from your network against the AdSense? E.g. I give you my AdSense code, you give a replacement that sometimes shows the original AdSense, sometimes your own ads and does a split testing for me

  11. Wonderful comments, thanks everyone.

    I particularly like what Ahmed had to say about community involvement. As you all know the very loose biz model at performancing has always been heading toward some kind of advertising deal, and although that may well not take the shape of a network in the end, it is certainly an option, and at present im having a lot of fun trying to write a spec that encompasses some of these core features and concepts whilst “keeping it real”.

    I’ll let you know how i get on

  12. I was just reading Darren’s fine piece on the perfect ad network, and I remembered something else. Every time I change a certain one of my Web sites, the clicks stop. The hits go up, but the clicks stop. Then gradually they start up again.

    Now this may not be a case of cause and effect, but it happens every time I change anything there. I’d like 1) to know why this happens, and 2) for it to stop. I added a mere two links to a long directory on Saturday, and this week I’m getting no clicks again. Just as many page views, but no hits. It’s crazy-making.

  13. My biggest gripe with AdSense so far, Nick, is the crappy (i.e., unattractive and/or irrelevant to my visitors) ads that show up on my sites. I’ve blocked a few, but that helps only minimally. I’m not savvy enough about the whole process to suggest a way to fix that–not yet anyway. But to me, this is the number one problem to date.

  14. In no particular order, here are a few of things I’d love to see in an ad network (which doesn’t exclude the possibility that they’ve been implemented in existing ad networks):

    1. Pageview counts have all bots filtered out, guaranteed?
    2. Network offers text, graphic, and affiliate ads. Not for every advertiser, but as appropriate.
    3. An option for ads in a digital magazine. For example, I’m still planning to resurrect my fringe monthly review print magazine in both online (HTML) and PDF form. It’d be nice to have PDF-based graphic ads for which I could earn CPM revenue, just like a print magazine would. This would be separate from website CPM ads, as the reader would likely be more demographically targeted.
    4. An API so that I could write advanced statistical reporting plugins (such as MMAs – Multiple Moving Averages).
  15. The great thing with i.e. good affiliate marketing networks is that they have a straightforward way to take care of all that commercial aspects and that they offer many different ways for the ad clients to spend their budget.

    A good ad network must pull ad clients! Sounds easy but that’s a lot of work to get it started and keep fresh ad content flowing. It’s definitely worth a commission of 15% to 25% if the ad network offers many advertisers willing to spend money for leads to participating network members.

    So my two main points to look after would be a trusty and efficient back office and the ad content.

    In no way I would ever grant exclusivity to an ad network! Well, I am corrupt though … 🙂

  16. One point that we have to look at is that the more features one puts in an ad network, the more complicated it gets to manage and use. With that in mind, you have to be really careful about which features to let in and which ones to reject and wait for the community to develop plugins for.

    So, what should an ad network contain?

    Quality Control (a good example of this is The Deck, although they might have taken it too far) – after a certain point it becomes very difficult to check ads manually, so perhaps a community review system might come into play here.

    Open API – don’t hide anything from the users and allow them to build all sorts of plugins and tools. A couple of things that Darren mentioned (stats and change logs) should be developed by the community, not necessarily by the ad network themeselves. As long as they let us use the system according to our needs, it’s great.

    Open channels – once again, difficult to achieve after a certain point in growth, but allowing publishers to communicate with advertisers and being open to questions and advice is quite important.

    Making ads lucrative for the publishers means one of two things:

    – the advertisers pay top dollar and/or
    – the ad network takes as small a cut as possible

    You can convince the advertisers to pay more IF the publisher sites are governed by strict quality control measures (Deck reference again here).

    hope this helps.

  17. As a designer and web standards advocate, the thing that deterred me and many other designers I’m sure, is the terrible code that is served up with ads like those from AdSense.

    We strive to ensure our sites use valid, error free code, so the appeal of earning a few AdSense dollars at the expense of introducing a stack of errors into our site is totally non-existent.

    If I could find an affiliate ad network, which provided valid, error free code snippets I’d sign up straight away. If I could find one that offered text ads with CSS layout in mind, utilising a few id’s and classes to make my life easier, I’d be over the moon. And if their graphic ads were diverse in design, message and size I’d probably explode 😉

  18. Unfortunately, there just isnt any good way to do that. Far too much room for error…

  19. I use some similar extension also (don’t remember the exact name), but it is not a simple solution – it works only for Firefox users, that have to install the extension (and get an idea that there is such an extension), then you cannot easily grab some colors because they change while you hover the mouse. All together it means that the solution is good for techies, but of little use for “just bloggers”.

  20. Good points artem, thanks.

    To help with your color picking, if you use firefox, try the colorzilla extension. It will allow you to find the right colors with the mouse from any web page.

    Very handy…

  21. I might look a bit too naive, but AdSense isn’t as easy to use as I’d like it to be. Basing on AdSense experience here is what I’d really like to have in the ease of use directory:

    1. How do I insert the ad into the blog?
    Yeah, I know about the HTML snipplets. The problem is that a starting blogger (like I am) has to somehow insert this skyscraper into the Blogger, WordPress or whatever he uses template. There are a lot of good 2 column templates in Blogger (e.g. their white Minima template), but there are neither instructions on how to add the third skyscraper column to the template, no how to widen the main column and use the float:left skyscraper. It might sound as not a big issue, but the need to dig the template code actually does prevent me from using the AdSense as much as I’d like to

    2. How do I tune the ad for my blog?
    Blending the color scheme is a great idea. The problem is that I don’t know the colors of my blog if I hadn’t designed it. And there are so many different colors in the CSS. Could the add network suggest the correct colors to me?

    Hmm.. I guess my both requests boil down to “Here’s my URL, propose me the integration steps”

Comments are closed.