Outside The Box Office

Success with Facebook Ads, pt. 2: Preparations

In keeping with the Facebook Ad theme this week, I’d like to dive into the key components you’ll want to have prepared before launching a campaign:

1. Creative

If you’re a Mad Men fan, you’ll see them testing several different creatives at various points prior to launching a campaign — they’ll ask each others’ opinions, do focus groups, present something(s) to the client, and sometimes even take multiple variants to market. One of the most beautiful things about the Facebook Ad Platform is that it takes care of all your testing for you, and continues to do so throughout your campaign. It will serve a small percentage of impressions of each of your creatives on a rotating basis through various targeted audiences. It will then (really, pretty constantly) monitor where success is coming and double-down on those matches of creative and target.

In other words, Facebook is working constantly to make sure your ads are performing as well as they possibly can. That’s is an incredibly powerful concept, and not an empty promise. But a testing algorithm is only as successful as its inputs allow. Garbage in, garbage out. But how are you supposed to know what’s garbage?

You’re not. You test. Use any possible variant on creative you can think of (three components: headline, image, copy) and let Facebook’s magic algo-sauce go to town. The more options you give the algorithm, the more precise its targeting can be, and the lower your costs per click will be.*

Components of good creative:

• Headline: Should be strong and straightforward, it’s your primary call-to-action on the ad.
• Image: Say what you will but the numbers don’t lie — pictures of attractive women outperform other images in ads by a 3x factor. I’ve seen it time and again. You want to stay consistent with your brand, but if it’s appropriate and relevant, you’re likely to see positive results. In the case of All Smiles, the ads just with Jim’s photo outperformed those with the album cover by between 2-5x per day over the course of the campaign. 
• Copy: Be descriptive, be excited, and most of all, explain why your brand is relevant to this person in particular. Will this mean a bit more work up front? Yep. But you’ll see results in spades, both because your message will be clearer and because you’ll have to put extra thought into how your target audience is likely to respond to you.

2. Budget

Make sure you understand how much you’re willing to put into the campaign and how much you’re willing to pay per desired action taken (eg - click, email address subscribed, etc). I say willing because you should be monitoring closely, optimizing as you go along (we’ll talk about this tomorrow), and if things aren’t working out be willing to pull the plug and invest the money in other channels.

There’s also a tricky balance in terms of timing and budget. I have yet to establish a baseline on whether it’s better to do a large spend in a short amount of time or spread it over a somewhat longer period. Theoretically they should be equal, but no algorithm is that perfect. I would love to know if anyone has any experience in this area — please leave a comment.

3. Targets

Selecting your targets is arguably the most important part of any campaign. Yes, Facebook will help you by optimizing to the best performing targets if you target too widely; but as I mentioned in the creative section, the creative needs to match the targeting, so you could easily wind up having to make an insane number of creative variants (a set of each of each different headline and image combination with different copy targeted at each group). 

Targets should be defined by whether or not they would be interested in your brand and why. Again, this requires you to get into the heads of the audience and be brutally honest with yourself: would these people have any reason to take interest in my music? Even if I have a really great looking ad, will they actually become my fans after they click through? Why would they?

If you start your scope small and incredibly focused, Facebook will help you expand to a more reasonable list to get the number of impressions you’re looking for. They do so by analyzing their own data on similar “likes” (among other things) and finding similar, real taste profiles (instead of a list of bands you come up with in your head that you want to sound like but sound nothing like). I may just be a nerd, but I trust data over my own preconceived notions on just about anything.

4. Landing Page

Minor detail: you need somewhere for all these newly interested people to go! Not only that, you need to have a rock solid idea of what you want them to do when they get there, and you need to make it as easy as possible for them to do it. 

The actions you want them to take should be directly tied to the goals of the campaign. I’m a big fan of a primary call to action asking for permission to market to them and a secondary call to action providing context in case they’re not sure yet. This generally means an email for media widget as the primary call to action and a streaming player as a secondary call to action.

Making it easy involves reducing clutter to an absolute minimum while still making them feel comfortable giving you their contact info (be it email, Facebook, Twitter, SMS, etc). A professional, clean design with as few elements as possible will prove most effective.

Our landing page for All Smiles has 5 elements: an email for media widget, a short note from Jim, a streaming player streaming the whole album, a Facebook Like button, and a small link to the rest of the website. The background is white, the title and text are black. The calls to action are the only colors on the page. Very simple, but incredibly effective.

Once you have your page prepared, your targets and budget defined, and your creative designed, you’ll be ready to dive in and start your ad campaign. We’ll pick up next time with some thoughts on execution.

*What I would really love to see, and what I think Facebook is trying to build, is optimization based on conversions. They can’t reasonably do a cost-per-conversion payment plan (too easy to game based on variants in implementations on the advertiser’s side), but optimizing around conversions instead of clicks would allow for far and away the most efficient ad platform ever.