Over the last several years, social media has exploded. If you’re thinking about promoting a product or service of any sort on a budget, it would behoove you to think about whether running ads on Facebook could help move you toward your sales goals. But before you jump to buying ads, do you know the key factors that will cause a Facebook user to click on your ad?

One of my more successful blogs (which I sadly haven’t updated in a while) is The Frugal Couple. For a while, I saw it as a vehicle to promote another online business of mine. (BTW, there’s a great post at Copyblogger that deals with the fact that to make money with a blog, you must think of it as a business.) I paid for ads on Facebook and was surprised by the results.

There are four things to keep in mind when creating an ad for a social media platform such as Facebook. Who you will target, what image(s) to show, your emotional connection, and the offer.

Targeting

Facebook offers advertisers several filters to reach a group. The goal here is to get your audience to between 10,000 and 40,000 users and then to test, test, test.

In order to test, I ran several instances of the same ad, but targeted at different age ranges and at men and women.

Next, I targeted users who used phrases in the interest section of their profiles such as: playing with my kids, spending time with my husband, spending time with my children, etc.

Graphic

The graphic was simple, but it was effective. I believe it clearly got the idea across that the site being promoted was about money for couples.

Emotional Positioning

The Ad Copy: “Become a Frugal Couple. Sign up for the free newsletter and start getting tips on living frugally and keeping your marriage healthy.”

It was a simple ad. There were two sentences in each ad. Each sentence told the user what to do. The main call to action (Sign up for the newsletter) was followed with the result the user could hope for if they signed up for the newsletter (keep your marriage healthy).

This appealed to users’ deep down desire to keep their marriage healthy (or in the case of some subscribers, to save their marriages from the dangerous paths they were on).

Offer

Although the stated offer was a free subscription to a newsletter, the real offer was to save readers’ marriages from turmoil caused by financial disagreements. Again, the offer was more than a newsletter. It was free information that would help the audience meet an emotional need.

Results

At the time when I ran the ads, the average ad on Facebook had a click-through rate (CTR) of 0.04%. That’s 1 out of 2,500 ads shown.

Depending on the targeting, my ads had CTRs of 0.10% (1 out of 1,000) to 0.25% (1 out of 400). In addition, I converted about 20-25% of the clicks into subscribers. I was paying $0.20 per click on average, giving me a cost to create a newsletter subscriber of about $0.80 to $1.00.

My ads had CTRs that were 150% to 525% higher than the average ad. Needless to say, I was happy with the results.

What are you offering?

Not all offers need to meet an emotional need. As much as I dislike it, many advertisements are successful because they promise to meet a desire such as greed or lust, or they boost the ego of a consumer.

As you begin a campaign, think about who you want to target and what that group’s psychological needs are.

Have you created ads that have been less successful than you would have liked or expected? Or if you’ve had great success, let others know about that as well. What worked?

To set things straight, I am comfortable in the Ubuntu terminal, but by no means am I a LAMP guru. I follow directions well and learn quickly, but often need to refresh my memory before attempting to update or adjust something.

I am slowly consolidating about a half-dozen hosting accounts to my provider of choice: Linode. Linode offers root access in a virtual machine, meaning lots of freedom, but also lots of tweaks. I’ve been able to do everything I’ve wanted with Linode, except for one: email.To be fair, I haven’t put much time into setting up email on Linode. I followed directions posted in the Linode Library, but never got things to work. After talking with a trustworthy developer, I decided to skip hosting email at Linode and let Google deal with the mail.

Setting it up was pretty straight forward. I visited Google’s Standard Apps link (here) and set up an account. I verified my primary domain (davidpodley.com) by creating a file with some Google-defined text and gave it the Google-defined file name, uploading the file to my server in the domain root (davidpodley.com) and asked Google look for the file.

I did this for each domain I wanted to associate with this account. To add other domains, I clicked DOMAIN SETTINGS, then DOMAIN NAMES, then ADD A DOMAIN ALIAS.

Finally, I followed Google’s directions and set my MX records through Linode’s DNS Manager for each domain Google was going to handle email for.

Google’s MX records were found through this link. If you use Linode’s servers for DNS management, you can follow these directions to change the MX records:

• Log into the Members section of Linode (linode.com/members) and  open the DNS Manager at Linode by clicking ‘DNS Manager’ at the top. This should take you to a page that has at least one domain listed.
• For each domain you want Google to handle mail for, you will need to do this. Select the domain. This will take you to a screen that lists several options. Mine lists SOA Record, then NS Records, MX Recods, then A/AAAA Records. If there are any MX records, remove each one.
• Once there are no MX records listed, click ‘Add a new MX Record.’
• You will be taken to a page with a form that says, ‘Add/Edit an MX Record’. In the ‘Hostname:’ field, enter the first address provided by Google. In the ‘Priority’ field, enter the priority provided by Google. You do not need to enter a subdomain unless you have reason to. Click ‘Save’. Do this for each record in Google’s list (again, found here).
• You will be taken back to the previous page. I click ‘Render’ at the bottom. I don’t know that it does anything, but it doesn’t seem to hurt.

You should be all set. For some reason, when I did this, I had to log into the Google Apps system one more time and tell it to activate the email account. In the Dashboard, it showed that the account hadn’t been activated, so I clicked ‘Activate’ next to the account name and all seemed to work.

Alternately, if you’re interested in becoming a Linode user, please use this affiliate link before creating a Linode account.

# cd /etc/apache2/sites-available

If you run the command

# ls

You will see the sites you’ve defined so far. You will need to create another site titled xyz.YOURDOMAINNAME.com, where xyz is the subdomain of your choosing. You can do this by either copying an existing, working site definition or by creating a new file. I usually copy an old site using:

# cp example.com xyz.YOURDOMAINNAME.com

Now enter the command

# nano xyz.YOURDOMAINNAME.com

You should see a box with text similar to:

<VirtualHost *:80>

ServerAdmin admin@YOURDOMAINNAME.com

ServerName YOURDOMAINNAME.com

ServerAlias www.YOURDOMAINNAME.com

DocumentRoot /srv/www/YOURDOMAINNAME.com/public_html/

ErrorLog /srv/www/YOURDOMAINNAME.com/logs/error.log

CustomLog /srv/www/YOURDOMAINNAME.com/logs/access.log combined

</VirtualHost>

Change the following lines as such:

ServerName xyz.YOURDOMAINNAME.com

ServerAlias www.xyz.YOURDOMAINNAME.com

DocumentRoot /srv/www/YOURDOMAINNAME.com/public_html/xyz/

It should now look like this:

<VirtualHost *:80>

ServerAdmin admin@hiredhired.com

ServerName xyz.YOURDOMAINNAME.com

ServerAlias www.YOURDOMAINNAME.com

DocumentRoot /srv/www/YOURDOMAINNAME.com/public_html/xyz/

ErrorLog /srv/www/YOURDOMAINNAME.com/logs/error.log

CustomLog /srv/www/YOURDOMAINNAME.com/logs/access.log combined

</VirtualHost>

What we have done is created a new server that notices that a user is coming to xyz.YOURDOMAINNAME.com and it now points the user to a different directory (note xyz at the end of the line in the DocumentRoot section).

Exit out of that screen by typing control-x. You will be prompted to save the file, so click Y for Yes, then hit enter when it prompts you with the file name (it should be your subdomain.domainname.com).

Now we create the directory the subdomain will read from:

# mkdir -p /srv/www/yourdomainname.com/public_html/xyz

Now we need to tell Apache to enable the site. Type:

# a2ensite xyz.yourdomainname.com

Then refresh Apache to know what’s going on:

# /etc/init.d/apache2 reload

You’re Done!

It may take anywhere from an hour or so to a day for the subdomain to propagate through the DNS system. I errantly assumed it would be immediate because DNS was managed by Linode, but it took about an hour for it to respond to my browser.

Help Keep This Site Going
If you found value to this article, please consider donating a dollar or two to help cover the cost of managing this site.

Alternately, if you’re interested in becoming a Linode user, please use this affiliate link before creating a Linode account.

If you use Twitter or Facebook, you’re probably familiar with services that shorten a URL, turning something like

http://www.website.com/this-is-a-test-and-it-is-a-lot-to-type-into-a-browser

into

http://pod.li/abcd

For many years, a site called TinyURL.com managed this quite well, and as far as I know, it still does. But as Twitter has taken off, the number of characters in link has become more important. Twitter’s 140 character limit on posts has driven people to use sites such as http://ur.ly or http://bit.ly for their shorter domain names.

I recently bought the domain name http://pod.li and installed an open source domain name shortening program on it. (In case you’re curious, the .li in http://pod.li is based out of Liechtenstein.)

Over the last year I’ve bought a handful of books that help with Facebook application development. Although I’ve yet to release an application into the wild, I’ve had fun putting some applications together for my own use.

When I came across Essential Facebook Development by John Maver and Cappy Pop, I figured I’d give it a try because it’s the newest book in this field.

The book is well written and leads the reader through a clear path toward not only building an application, but understanding how an application is built. It uses quite a bit of object oriented PHP, so for someone new to programming, it might not be the best resource, but overall it’s a great book.

The bulk of this post is about two other issues, though. First up is the speed of changes in Facebook’s application development API. The second point is the responsiveness of authors and the value it adds to readers in niche markets.

The ever-changing nature of Facebook’s API must be a thorn in the flesh of authors. Case in point: Essential Facebook Development has a sample application in the book. (Like other good programming books, the code is made available at a site dedicated to the book.) Rather than hand-code the examples, I put them into an application I set up on FB and I got an error. In searching for the solution, the book uses a method to post messages to a FB user’s news feed that Facebook no longer supports. In the last few weeks, Facebook solidified a new way for applications to post messages and communications to users’ news feeds, walls, and other areas of interest.

I reached out to the authors on their site (essentialfacebook.com) and I had a response within 10 minutes from John. His response:

Yes, the stream API was in beta for a long time, so we didn’t focus on it in the book, but of course it was released shortly after the book was finalized. We will be doing a set of blog posts on using the stream API to make sure our readers are kept up to date.

This speaks to a problem we’ve seen in other areas, namely that traditional books are obsolete out-of-date the day they hit the shelves. As this happens more and more with developing technology guides, authors will need to be more and more responsive to their readers.

I read about two books a year. For some, that seems like a lot of books, for others, it isn’t. Nonetheless, our home is running out of space to store our books. If you figure an average book is 2 inches thick, that’s about 4 feet of books each year I add to our bookshelves.

When Amazon released their Kindle, I was skeptical at first, but when Jeff Bezos announced an updated version of the Kindle, I picked one up and I’ve absolutely loved it. The idea of taking a medium-sized library with me was exciting, and the instant delivery over a cellular network quenched my thirst for immediate gratification.

The first Kindle that showed up had a stuck volume button. (For those who don’t know, the Kindle can double as an audio book reader for Audible books, as well as simply play MP3s.

I called Amazon and explained the problem. A day or two later, Kindle #2 showed up. For the next six months, Kindle #2 and I had a torrid affair. We spent time together in bed, on the couch, and at the beach. But a few days ago, my dear old #2 decided she had had enough of me and gave up the ghost. #2’s display stopped working. (Image of broken display at the end of this post.)

I called Amazon again and explained the situation, and they sent out K#3. #3 and I are getting used to each other, but like a lover scorned, I hesitate to really become attached to #3.

While I was on the phone with Amazon, I asked about adding an extended warranty in case #3 fails like her sisters. I was told that because the warranty service is offered through a company other than Amazon, they did not make exceptions to their rule that warranties must be extended within the first 30 days.

I’ve got to admit that I’m disappointed with Amazon’s quality control. The first Kindle shouldn’t have made it out of the factory. My biggest fear is that #3 will pass away after its warranty period and I will either a) need to buy K#4 in order to read my purchased books, or b) give up my library.

Here’s the image of the screen. Note the white lines where the display simply has stopped working.

I recently read an article by a gentleman (article found here) named Charles Warner. Mr. Warner has held some significant marketing positions, has written a textbook on marketing, and now teaches at The New School in New York.

The article is a long conversation on how to show a potential purchaser of advertising space (in this case a supermarket chain purchasing ads on a radio station) the return on advertising investment.

It’s a detailed article, but it confuses revenue with earnings.

To sum up the article, the advertising salesperson shows the supermarket executive that if he invests $1 million in advertising and another $500,000 to increase staff at the stores to handle the increase in store traffic, he will see an increase in revenue of $2 million. Mr. Warner states that this is a 50% return on the advertising investment ($500,000 net increase on $1 million in advertising).

What’s Missing?
What is missing is that earlier in the article, the author points out that the grocery store has a gross margin of about 3% including advertising. That means that on $2 million in sales, the grocery sees pre-tax gross profit of about $60,000.

I’ll give him the benefit of the doubt and say that the grocery chain’s marketing busget is 20% of sales. This would mean that the margin before marketing is about 23%. On $2 million of sales, the margin would be $460,000.

A profit increase of $460,000 before accounting for a marketing expense of $1 million and the increased staffing needs of $500,000 is worse than never running the campaign.

Even if the increase in sales for the campaign last four times as long as the campaign itself does, the grocery chain still loses money.

Thoughts?

I got a new MacBook Pro the other day and I went through the process of installing my software. I have a couple of Apple’s pro media suites (Final Cut and Logic Studio). I installed Logic, then Final Cut. I compressed some video, played with Logic a bit and all was good.

Then an odd thing happened when I opened Final Cut Pro. I got the following message:

You cannot open the application “Final Cut Pro” because it is not supported on this architecture.

It seemed odd. The same day I had done a Software Update and one of the items on the list was a fix to the NVIDIA video card, so I figured there might be an incompatibility.

Long story short, I was wrong. Apparently the order in which you install Apple’s Pro Media programs matters a great deal. Kind of like mixing acid and water (acid into water, not water into acid, by the way, which is a funny story how I learned that one).

The correct order is Final Cut Studio THEN Logic Studio. Don’t ask me why.

I called AppleCare to make sure I was right about this, and the nice young woman on the phone said, yes, I did it wrong. I should also reinstall the operating system if it’s not too much of a hassle because uninstalling the Pro Apps is near impossible.

I’m now backing everything up with SuperDuper! (a great program, by the way). I’ll soon have everything worked out. Hopefully.

If you were an advertiser, what would you pay to reach 1,000 viewers of the 2009 Super Bowl? How much would you pay to reach 1,000 viewers of American Idol?

This is a question advertisers ask themselves as they go about their media buying planning. A 30-second commercial for the 2009 Super Bowl will cost advertisers $3 million. Assuming this ad will reach about 92 million viewers, the cost for each viewer is about $0.0336 (3.36 cents) or $32.61 dollars to reach 1,000 people.

The same commercial costs advertisers about $700,000 on American idol. On average, American Idol is watched by 30 million people. This breaks down to a cost of $23.33 for each 1,000 viewers.

(For a point of comparison, a commercial running on American Idol at the same cost per viewer as a Super Bowl ad would be $978,300, or almost 40% more.)

I know a lot of beer ads are run on the Super Bowl (Budweiser is running 9 ads for a total cost of $27 million), but I haven’t paid much attention to American Idol ads. I know Coca-Cola is a sponsor of the competition, and I would guess lower-cost cars are advertised on Idol, but beyond that I don’t know what is sold on the show.

My guess is that part of the 40% premium is a higher value on the Super Bowl’s demographic, and another part is that the Super Bowl is watched by many for the ads rather than the game. If a company develops a great commercial, they can get much more value out of the commercial than the airing of the commercial. Think about it. How many other times during the year do people come to the office and talk about the commercials they saw the day before?

Sometimes I’m an early adopter and sometimes I wait a while. With the iPhone, I’ve waited and I’d love to pull the trigger and get one in the next few weeks.

I currently have a Blackberry through T-Mobile that is out of contract and I have a plain old Motorola phone with Verizon. The iPhone would replace the Blackberry and its service.

The pros of the iPhone: Well, it’s cool and it has some great software on it.
The cons of the iPhone: AT&T service isn’t so great in my town, it’s almost twice as expensive as the Blackberry service I have now, and there’s an initial outlay of cash of $200 to $300 (depending on the model) for the phone itself.

As I calculate it, the iPhone would cost me about $2125 over a two-year contract (including the cost of the phone and taxes). If I keep the Blackberry, it would cost about $800 for the service. This is a difference of $1325 over two years, or $55 per month.

The primary question I have to ask myself is, do I get $55 of productivity over the Blackberry out of the phone each month? The Blackberry costs me about $30 per month, or a dollar a day. I can sort of justify the cost of the Blackberry because I get email throughout the day and it has helped me a number of times to save a trip back to my computer to get some information I forgot to bring with me.

Would I get another $2 per day of value from the iPhone? It’s hard to say. I don’t know that there’s one application on the iPhone that makes me say, Yeah, that would save me lots of time/money. I would love to think that there is, but I just don’t see it right now.