Increasing Ad Revenue

When a website is popular and has a lot of traffic, then it’s easy to get good ad revenue. But if you are a small fish in a big pond, then it’s not so great. There is a good chance you could get booted from Google’s AdSense program because some idiot sitting on the other side of the globe is illogically clicking the links or whatever the reason may be, or worse yet, you may not even be selected into the program.

Ofcourse, there are alternatives such as AdBrite. However, don’t stick with just a single Ad Network. Based on popularity of AdBrite, I tried doing this and my CTR is very low and further, AdBrite seem to be serving ads with very low rates (CPM or CPC) for low volume sites. It’s understandable that they want to maximize their advertisers value of advertising through their network. However, low volume doesn’t mean quality leads.

The main problem with all the context-less advertising networks is, they heavily rely on category of the website (how targeted can you get with just a handful of categories?) and geo location of the visitor. So, with current techniques used by these networks, there is no easy way for them to measure the worthiness of a site or the right way to match the advertisers to the publishers besides based on metrics such as alexa ranking.

So, the best approach is to go with multiple ad networks. Some of the ones I have tried are AdEngage and Bidvertiser. I noticed a different CTR with each network. This could be based on how and where I placed the ads from these different networks, but there also might certainly be some factor related to what ads are getting served on your site through these non-context ad networks. With just a few days of experimenting, I am quite pleased at present with Bidvertiser. I am still running AdBrite but may gradually get rid of it if I don’t get good CTRs or if they keep showing only “Advertise On This Site” and not really putting ads (what a waste of space, and I guess this has to do with supply and demand. They perhaps have more ad real estate from publishers than the advertisers). The key is to go with networks that have a more balanced supply and demand so that it’s possible to get higher CPC/CPM ads as well and not the bottom most of the pile.

“A Campaign to Block Firefox Users?”

Yes, this is what was discussed on slashdot. Then, I just happened to see this facebook advertising warning which mentioned about smart students using ad blockers.

The reason why many services on the internet are free today is because of the hope of many publishers hoping to make revenue out of advertising. However, if this is prevented by special software, then that is definitely going to hurt these publishers who need to either fold themselves or need to start making their content premium.

In general, any ad that doesn’t cause irritation to the eye (such as those ads that keep flashing, or animating too fast) are not really hurting the readers. So, why bother to block? Infact, think of such ads as a way to learn about new products and services.

Anyway, my intention here is not to advice end users about what they should do or not do but wanted to discuss a specific feature of Firefox that is helping blocking of ads and an idea on how to prevent that.

Firefox has many extensions, one of the most popular being the GreaseMonkey extension which allows one to write simple JavaScripts that get executed upon loading a page. These scripts have the ability to modify the page structure and do such things as changing every email link or a pdf link as a normal text followed by a special link (so that you know up front that these are not simple navigation links) to complex things such as removing sections of pages, adding additional links.

Below are a few ways to prevent such scripts removing sections of the page

1. Keep changing the dom structure of these sections. For example, if div is used once, next time use table. If using css, keep changing the class name. Essentially, make it so random, that no generic algorithm can be developed to navigate through the DOM and delete the sections.
2. Instead of generating the html for these sections as part of the overall page generation, make it loaded through a script. This way, as soon as the page is loaded, the DOM structure will not have the ad section and hence the user script can’t figure out and remove the relevant DOM. Once the page is loaded, with a very little random delay, start adding the Ad section to the page.

Interactive Catalogs

What good is all the time and effort spent on advertising, if your website is not good enough to engage your audience so that they come back to you, book mark you, put you on del.icio.us or digit and other web 2.0 services etc? And not to mention, make that last step of closing the deal and open up their wallet?

The article Website Design a Top Concern for Online Retailers has some interesting statistics on how online retailers are working on improving their websites.

One point that I find interesting is

“Only 35.1 percent of merchants use rich media: Among them, 20.7 percent are using videos, 16.6 percent offer zooming in on images; 10.4 percent, interactive catalogs; 9.8 percent, dynamic imaging; and 7.8 percent color swatching.”

These numbers are so low. Only 10.4% using interactive catalogs? One idea for interactive catalogs is Image Cloud based Cloud Store which showcases Amazon affiliate page in the form of a Web 2.0 concept of tag cloud, only the tag is replaced by the product image and just like the size of the tag varies in tag cloud, the online interactive product catalog’s product images are varied in size based on their list price. Placing the cursor on a product opens an inline popup showing additional details of the product. A very cool effect using html image maps, dhtml and javascript.

Here is an example.

Multiple Middle-Men In Online Advertising

May be this phenomenon is new to me, but I observed something interesting on AdBrite. We all know that AdBrite facilitates a way to make advertisers and publishers meet each other through their platform and in the process, make money by taking a commission in the revenue earned by the publishers. However, what I noticed is, some of the companies advertising on AdBrite are not direct end companies, but themselves are marketing companies representing their clients. The two companies that I noticed are doubleclick.net and mediaplex.com.

Here is how it works. AdBrite has 3 types of ads (text ads, banners and interstitials). Of these, banners and interstitials are actually links from the advertiser directly rendered on to the publishers page. Text Ads are not directly served from the advertiser’s server. This is because, text ads are served by AdBrite in multiples and it doesn’t make sense to each of them from multiple servers to assemble. So, it is required to provide the text of the text-ads along with the target link on AdBrite.

So, what’s happening is, mediaplex.com for example, will bid on AdBrite.com for their network advertising and offer as little as $0.002 per interstitial. These numbers are much lower than what a direct advertiser typically pays, which could be $0.005 to $0.01 or more. So, mediaplex.com would itself work with companies to advertise for them through them and perhaps charging much more per interstitial visit than what they are paying through AdBrite. Essentially, not only is AdBrite getting a cut out of your ad revenue, which is very understandable for providing their excellent platform, these other middle men are also making money, we don’t know how much but could be much more than what AdBrite itself might be making!

I think it’s a matter of time for the companies to realize that platforms like AdBrite eliminate the need for working through others and advertise the ads themselves directly. When this happens, then both the publishers and advertisers will truly benefit from each other. Otherwise, at $0.002, the CPM for interstitials is very very low, especially because interstitials are displayed typically once per session visit.

As to the technology on how this middle-man business works, typically AdBrite type networks are context-less networks offering mainly brand awareness type advertisement. So, the only requirement for this type of ads is the sheer number of visits and not necessarily the content itself. So, the middlemen are essentially buying these visits through the interstitials. Their interstitial link itself will keep redirecting/dynamically generating different content each time, rotating the ads of their own customers. Similarly for Banner ads, when the image is served up from their website, they choose to display different ad each time.

While branding ads typically lack content context, there is still some vital context, the demographics which could be as simple as the ip to geocoding to figure out the location of the visitor, to a more sophisticated cookie-tracked visit patterns. Hence, in theory, these middlemen can offer more sophisticated targeting for their (which ofcourse, AdBrite itself should be able to offer) and hence can potentially make even more money than the flat rate they pay on AdBrite!

Essentially, this is similar to the consulting companies in the IT world, where there are multiple middle-men each grabbing a project from a customer for a higher hourly rate and giving it to a guy who can execute/outsource to a lower hourly rate! So, AdBrite is essentially that guy to whom you can outsource for cheap.

Many business models are going to evolve in the future based on this concept.

Now imagine this. What happens if someone becomes an AdSense publisher for high-end click rates and starts serving their page through cheap interstitials? If they could achieve a 0.5% CTR through this technique, by paying a mere $2.00 per 1000 visits, and if each AdSense click fetches them $1.00, they are better of by $3.00. All this, with absolutely no work on their side apart from the initial setup.

CRO – Conversion Rate Optimization

You invested in SEO (Search Engine Optimization) and managed to steadily get visitors to your website. What next?

That’s where the CRO comes into picture? What exactly is CRO? Is it applicable for only websites that sell products or is it relevant for content providers as well? Is it something that can be outsourced? Based on my ongoing effort and understanding I will try to answer some of these questions here.

Conversion Rate is one of the key metrics for any online advertisement system. Say you spend $5 CPM for banner ads and of those impressions, say 2% end up as clicks. That means, you get 10 visits. So, the cost of bringing in a person to your website is, $0.50. Now, of these, say only 10% end up taking an action that you desired. Then, the cost of bringing in people that perform your desired action is $5. So, unless the action generates more than $5, it’s not worth continuing with your current advertising strategy.

So, Conversion Rate Optimization is an art (I don’t think this is or will ever become a science) of optimizing your page content to increase the conversion rate. There is a lot of emphasis on the landing page because, the way it is designed has a lot of impact on your top-line.

CRO is not limited to just people who are selling products/services on the web. Many content providers don’t advertise as much but invest time and effort in creating great content with the hope that they have visitors finding them through regular search. And they make money when the user clicks an ad on the visited page. So, where should the ad be placed such that the chances of clicking the ad is the highest? What kind of ads will generate the best CTR (Click Through Rate)? While some of the influencing factors including the user interface, the performance of the website (remember, a typical surfer has no more attention span than a few seconds since your competitor is just a click away, perhaps below your position in the search results), the others such as the quality of the ads, their relevancy are not entirely in your hands when using a third party system like AdSense or AdBrite (though, I would assume AdSense is much better as it knows the context than AdBrite).

Finally, is CRO something that can be bought from a 3rd party? I think CRO should not be considered same as SEO for a couple of reasons. One is that the success of SEO can be easily measured by anyone as all it takes is to search for the keywords the pages are SEOed and see the position of the pages in the search results. However, conversion rate information is more private and only you as the owner of the website know about it when you sign into the ad publishing account that you use. Obviously, you probably don’t want a third party to know about your financial details.

Another reason why you and only you should spend more time on CRO is that you understand your customers better, you understand your content/products better. Why do you need someone else to tell you what works best? All it takes is just to experiment a bit. Unlike SEO, where it’s the search engines that decide where your page should be in the search results, CRO is completely in your own hands.

If you are a complete DIY person, if you thought SEO itself takes a lot of time and effort, you have no idea how hard CRO is going to be. Start allocating some of the time and effort that you are putting for SEO into CRO. Good luck!