BuySellAds.com

My book Barcodes with iOS is nearing completion. Buy it now to get early access!
Our DNA is written in Objective-C
Jump

Experiments in Blog Advertising

For the past few months I invited ads of fellow developers onto the Cocoanetics blog. I was quite dissatisfied with the payout from Google Adsense (like $1 per week) and this prompted me to search for additional methods to offset the cost of my hobby. I previously wrote about how you can express your appreciation and that my server bills for $60 per month. This mark continues to be my break even point.

Let me share the results we got with different kinds of banner advertisements and what we learned from this experiment.

Front Page

Let’s first inspect the results from placing ads on the site frontpage, the page you see if you go directly to cocoanetics.com.

This was for the camThis! app and liked right to the iTunes address. I love the production value of the ad, it has a bit of a geeky Portal game quality. Though if it weren’t for the statement that this app is about “streaming live to everyone” then it might be hard to instinctively know what this app is about just from glancing at the graphics.

I went digging into my Google Analytics outbound links, but I could not find a single click through. That’s the rock bottom of ad failures. To be fair, a major contributing factor might have been the placement on the front page. People coming from Google searches generally end up on an individual article page related to their search query. For somebody to get to the homepage to see this app he would have to go to my site directly. That’s 7.15% of traffic. Of these 150 people how many do you think would be looking for a video streaming app. Very few.

It dawned on me that for an ad to be successful on Cocoanetics.com it would have to have a context with the topics that are of interest to iOS developers. That’s why I had higher hopes for this ad for nKey, a Brazilian iOS development outfit. Again, we see high production value in terms of graphics, but the lack of a clear call to action or message. And psychologically the grid in the background might be seen as a metaphor of a strategy “full of holes”.

This kind of ad is what professionals call an “image advertisement”. The aim of these is not to get a click-through, but instead have people associate a certain message or quality with the company and it’s logo. Image Ads are usually done by large brands like Coca Cola and they are only effective indirectly and if done en masse.

If this indeed was the intention of nKey then they have achieved one thing: not getting click-throughs. We cannot measure however whether more people will now subconsciously think of nKey when looking for “Brazilian iOS Development”. It would have been smarter to drop the info on Brazil as this would probably trigger certain fearful instincts and instead focus on some concrete offerings. Something like “we can build your app before you even know what it is”. Or “free app feasability analyis”, or something else that makes us lust for more information and promises instant gratification.

Article Pages

First let’s have a look at the ads that were placed at the MORE break on app pages of this blog. With daily unique visitors around 1500 generally looking for specific articles this ad would be seen by the most people.

The first and longest running advertisement was for the Cyberspace browser app. In retrospect I was rather successful, probably because of the attractive imagery, and pushing several buttons of us iOS geeks. Cyberspace. Space. Something special.

The campaign was extended once and in total we tracked 68 unique click throughs over a period of 2 months. We terminated it mostly because there was too little tangible effect on app downloads.

After 2 weeks of interim Google ads a new campaign was started, this time for ScoreEuchre, an app to keep track of Euchre Scores (I think). Since it was in the middle of a month I added another month at no extra charge. Though what a disappointment this was.

One month since this started we only have 19 unique clicks. That got me to stop and ponder a bit what the cause for this big difference of effectivity could be. Why would an ad with the exact same placement gather only half as many clicks?

It might have to do with the copy and design of the banner. If you step back and feel for your first impression mine is that it’s somehow blurry, has like compression artifacts around the text and sort of a passive-aggressive color scheme. It does have a clear messages that this is about an app and that the developer is obviously looking for reviews, offering promo codes in exchange.

Here’s the first problem. Last time I tried to write a review on iTunes for an app I had gotten via promo code I wasn’t able to. Either that was a temporary glitch or Apple has finally seen the light and only accepts reviews from people who actually paid for the app. So the first premise “Promo Code for Reviews” is out, if this refers to iTunes reviews.

The second premise is even less attractive. For I while I offered promo codes to all of my apps via a prominently displayed text ad at the top of every page. This was literally seen by thousands of visitors, but only occasionally somebody would actually ask me for a promo code. Developers seem to be quick in making up their mind about the worth of other developer’s apps. And if we like these apps we usually don’t pinch pennies but pay the price. Gratification, even more instant than via promo code, because these codes you have to request, copy/paste, etc. Buying an app you like is just a click and one password away. And another thing, developers are developers, not writers. I am the exception to the rule as I love to put my thoughts into writing. But I know of very few developers who are actively nurturing a blog, let alone put reviews up for other people’s work.

Lastly and most importantly, what is this Euchre? Some sort of card game. Sorry for my ignorance, but I have never played this game. And also I know of no body who does. It’s definitely a niche product for a niche different from my target audience.

Sponsored Tweets?

I find from my bit.ly stats that any kind of article will get between 100 and 200 click throughs. Which is an order of magnitude higher than the number of click throughs that banners will gather over the course of a month. I’ve yet to get any sort of stats what the benefit is for sites that I send traffic to by recommending them to my 2500 followers.

To simplify the process for myself I moved away from manual ads and instead implemented the BuySellAds.com system. Have a look at the index to see the currently available placements. There you now also have the option of buying sponsored tweets.

I have no idea if and how this might work out, but maybe you can be the first to purchase an ad there, so that we see if it works? :-)

Learnings

For an ad to be effective on Cocoanetics.com it needs to chime with the geeky core values of the average iOS developer. Saving time. Getting things done. Social recognition. Niche Knowledge. Getting good at your craft.

People coming to Cocoanetics do so because for the most part they are looking for solutions to their everyday iOS development problems. They don’t appear to be looking for other developer’s apps, because they are making their own. Also they are not looking for promo codes, as there are better suited places to cater to the market of developer’s desperate to get more exposure.

To sum it up in one sentence I’m offering this: “an ad for your product and service only makes sense to put on a site if you have a good indication that this site’s target audience might also be interested in it”. One word: Demographics.

Don’t get me wrong, I’m happy to display your banners any time, but I am interested that you get maximum bang for your buck. You’ve seen that ads that developer’s make themselves range in production value, but generally fail to transport a clear message and “call to action”. It definitely has a big impact if you have a professional graphic artist design your ad banner.


Categories: Advertising

%d bloggers like this: