Measuring Reach of Paid Advertising on Facebook

As social media programs are getting more and more popular, it is important to understand how to measure the effect of the paid advertising. In this post, I will like to talk about the basic concept of reach, as ad exposure or reach is the first step in the game of placing ads.

Typical, Facebook campaigns are done to increase the fan base of a particular page or brand. In these campaigns, there are three types of ad impressions placed on Facebook:

1. Paid Media – Paid ads with the message of "Become a Fan" of the particular brand
   
2. Earned/Paid Media – Paid ads with the message of "Become a Fan" but with the social element showing a blurb with the name of a few of your friends who are fans of the particular brand. Since, these ads have the social component these are classified within the earned media category as well.
   
3. Earned Media – These are the instances of ads which show the list of friends who are fans of the brand and have a link below "Become a Fan". Since, the advertiser didn't pay for these impressions, these types of ads or impressions are termed as Earned Media.
   

Thus, measuring reach from all these impressions is not an easy task. It is impossible to measure it using third party ad servers like DoubleClick as the Earned Media is not served via any ad servers. It will be interesting to find out if Facebook can provide information on the number "Earned Media" impressions served for a particular brand. The next step will be to provide the number of unique impressions seen by each user; this will be the reach from the earned media impressions.

Another way of estimating this reach is based on the number of users who become fans from the paid media. On an average, each user might have 50 friends (this is not an accurate number, just for calculation purposes) thus, if a campaign generated 200 fans, the additional impressions generated will be 200*50 = 10,000. However, then some of the users who are exposed to these 10,000 impressions will become fans, thus generating more earned media. Thus, it is a cycle which goes on – beauty of social media. A good estimate will be to use 3 iterations in social environment like Facebook.

A recent study by Facebook and Nielsen shows that the Earned media impressions are the most impactful in driving awareness, ad recall and also purchase intent. Thus, it is very critical to drive these impressions. Though less in volume, these are the ones which have the biggest bang for the buck.

Social Media

The key objectives of Social Media are to listen, answer consumer's questions, embrace the thoughts and respond.

Respond to convert the negative sentiment into positive. React fast enough to negate any crisis like the McAfee anti-virus situation, a few days ago. McAfee reacted really fast by apologizing for the issue that helped control a negative impact on their brand. Do not wait like Dell, a few years ago when the battery explosion issue was out in social networks. Dell responded by saying it is a single isolated incident but then in a few days, the entire internet was full of complaints.

Embrace your user's opinions and then present it to the world, like Microsoft. Microsoft is utilizing this strategy in the "Windows 7 was my idea" campaign.

Each of the above mentioned tactic probably needs to use a different platform – Facebook, Twitter, Consumer forums, Review sites etc.

How to measure the impact of social media activities?

Some of the key metrics (monitored over time) that can be used to measure the impact of social media are:

1. Brand mentions vs. competitors
   
2. Brand attribute mentions with the brand
   
3. Sentiment
   
4. Correlate the changes in mentions or brand attributes to media/PR efforts
   

There are a few tools in the industry which can help gather the above metrics like Radian6, Nielsen Buzz. These tools might have additional widgets, reports which can provide additional insights about your brand.

How to prevent ads from appearing at places where you don’t want

In the previous post, I mentioned how verification companies can help monitor where your ads are appearing. But that is a cure to the problem after the fact; your ad already appeared on inappropriate section of the internet.

There are services available which prevent your ad from appearing at pre-determined section of the internet.

Companies have the technology to monitor the section of the internet where the ad is about to be served and stop it if it is not appropriate. These companies have very extensive algorithm which rates almost all the pages on the internet and then use this rating to make the decision of preventing the ad to be served.

Essentially, they create a firewall in between the publisher and the consumer before the ad is served.

I haven't used these technologies but feel that using them will really make the publishers accountable in terms of content and geographical location. It gives more and more control in markets/agencies hands.

Do you know where your Ads are?

As advertisers and media agencies, we sign the Insertion Orders (IOs) and then assume that the publishers are placing our ads at the correct placement and also targeting the right audience.

What goes on in reality? Obviously, a certain percentage of these ads end up in areas of the Internet where it doesn't make any sense (Adult content, areas not relevant to the brand etc) or might get served below the fold or even get served to out of the country audience.

What if you are an e-tailer in UK and your deal of the week ads are being served to people in India or USA or Brazil. It's a waste of media i.e. waste of money. There will hardly be anyone in India, USA or Brazil looking to buy from your online store in UK.

Similarly, if you are a technology company targeting your ads towards gamers but your ads appear within parenting section. The likelihood of conversion in parenting section are really slim.

There are now audit services available that can audit your online media buys and tell you exactly where your ads are being served. Are your ads being served to an international audience, are your ads being served in the areas that are not relevant.

DoubleVerify is a service which can provide you this information. They claim to have relationships with majority of the publishers and seem to be well known. It seems to be pretty simple to use them if you use third party ad-serving like DoubleClick, Atlas.

DoubleVerify provides you screenshots of your ads in inappropriate content, international traffic, below the fold ads, multiple impressions on the same page. All this information can be used by advertisers or media agencies to renegotiate the media buy with the publishers. Net net bringing additional savings to your client or your own business.

Improve your SEO rankings

There are a few key elements which can help in improving your SEO rankings. This post will talk about a few of the key things necessary for improving and also helping your site getting indexed by major search engines like Google, Yahoo!, Bing.

1. URL: Make sure that the URL is short and friendly to read. Including keywords within the URL, it is easy to index these pages. For example: www.xyz.com/desktop/OEM-name
   
2. Web site design: It is known that if frames are used within the pages, it is not a good user experience and also hard for the search engines to index the content.
   
3. Navigation: Avoid flash based navigation categories. Search Engines are not able to go and read the content within flash, causing issues and lower SEO rankings. It is a known best practice to use the navigation lists as HTML links.
   
4. Link similar pages: If you know that users will be reading similar pages, link them to each other.
   
5. Add keywords within the content: Make sure there are a lot of keywords within the content of the website; this enforces that the site is about a particular search engine category.

Using Web Analytics data to understand the user behavior

There is a lot of data available in the web analytics platforms like Omniture, Webtrends etc. One of the key reports is the click-stream report. Having access to raw data can help understand the differences in the site browsing behavior in the first visit vs. the second vs. subsequent visits.

Analyzing the above mentioned data can help understand the patterns where in the process do the users' drop-off the website before converting. Also, how different is the behavior from first to the second to subsequent. This information can be used in multiple ways:

1. Optimize the site content – make the content more easily accessible that the users are reading in the later visits
   
2. Using cookie tracking – optimize how the site is presented to the users in the first visit vs. subsequent visits
   

These small changes can definitely lead to high conversions rates especially for e-commerce sites.

This technique can also be used in the way how ads are served to users who have dropped off the process before converting. Dynamic ad serving companies like Tumri and Teracent can be used to change the color, messaging to first visit vs. a user who has been to the web site.

Promoted Tweets – Twitter Paid Ads

Twitter just announced today it will be starting advertising in form of paid tweets. These tweets will be known as Promoted Tweets.

From a targeting perspective, it is a great opportunity for Twitter to start text mining the current tweets or conversations and then serve the ads. This can become a very well targeted advertising network based on conversations, user profiles etc.

For example: Many companies have twitter handles which promote/provide information about their products. Thus, profiling the users based on their interest and posts will be a great way to target ads.

Instead of relying on any 3^rd or 4^th party database to profile users, Twitter has the ability to segment users based on its in house data – the past and current tweets.

It will be interesting from a measurement perspective, to measure how these ads will work.

I hope to see some webinars or research papers coming up soon on Twitter ads.

Audience Segments

A lot of technology companies like Bizo, Audience Science are helping advertisers target the valuable user segments. These companies have access to huge databases containing information about these users. This information can be used to target B2B users for example – industry segment, functional area, size of companies.

Information to these companies comes from registration information at various places over the internet. Based on the cookie tracking, these users can then later be found & targeted with the ads on other sites throughout the internet like Yahoo!, MSN etc.

It does seem like a simple ad network but with what I will like to stress is that these networks have the information about these users. Of course, this information might not be a 100% accurate but it is still better than using an ad network to target users with no information except their behavior a.k.a. BT (behavior targeting).