Twitter Ads

Everyone seems to be talking about Twitter's new ad model this week. The newly unveiled promoted tweets in search results ad model was shown by Twitter's CFO this week and clearly is a big step forward for the company. While, I have looked at it and I have not formulated a set in stone opinion on whether it will work or not but I do think it is only a small step towards fully monetizing the website.

Why do I say this?

First, the ads are only going to be displayed on a tiny fraction of the site, being seen by users who search via the web interface only. Twitter has not disclosed numbers on how many folks search but I have to figure its a tiny fraction and when you ad that small number by the number of users who generally click on ads you get an even smaller number. Do you see where I am headed with this? 

Second, Google was successful by matching desire with satisfaction. Meaning that people desired to find stuff on the web and Google provided satisfactory results to those queries. 

Folks don't go on Twitter to search. Their main intent is for quick snippets of information without searching or spending alot of time finding it. I would venture to say that the average Twitter user finds his connections through friends and via other means than searching on the UI. 

Let's hope these ads are relevant when landing in the results page and its not at the expense of the experience that these ads take precedence over other more relevant results. In any event, its nice to see Twitter has started to monetize the site and we can only hope that it will not interrupt the experience too much.

3 Common Implementation Errors

Digital marketing has ushered in a new wave of analytical marketing where everything can be measured right down to the click. But unfortunately its is not without its faults and the issues come in all different shapes and sizes. Everything from complex coding to simple human error is common place. And in my role at Microsoft Advertising, I see a wide variety of these issues ranging from the mundane to the utterly complex.

One of the most common errors I see that leads to more client headaches than any is publisher implementation. These varied errors cause more wasted time than any other type of error when investigating differences between publisher and client. The good news is they almost always can be prevented by a little thought and preparation before hand.

Here are a few simple and easy rules to make sure your tag is implemented correctly.

  1. First make sure the tag is placed on the page. Its quite common for publisher to implement the tags on the wrong page or just not at all. So look to see if the ad is running and on the right page. It seems like a no-brainer but you would be surprised how often this is the case.
  2. Second, make sure the entire tag is implemented from beginning to end on the page and not broken. The method for QA-ing in this manner depends on the type of tag but generally you want to spot the opening tag which will look something like this: < iframe > or <script type=”text/javascript”>. Then you will want to follow the tag until you see the closing command. It will look similar to this: </iframe > or </script>
  3. Finally,  make sure cache busters are implemented. Many times its the case that browsers are caching an older version of the tag and not counting all the impressions. To QA this simply look for the cache buster in the string. Different systems have different versions of cache buster so make sure the tag you look at has the appropriate buster on it.

This is certainly not all the possible implementation issues I have come across in my time. However, I would say it accounts for the bulk of them. Not to mention, its always a terrific idea to QA these first and foremost as you can almost always save yourself some time on the back-end when your reporting shows up with gaps. Hopefully this provides a good starting point for publishers and advertisers alike and happy implementing!

 

 

*** This is a repost from stemato.com my original blog.