Glass Cremation Urns, Paperweights and Memorial Jewelry

Remember Through Art

Facebook A/B Advertising Testing

business ofDavid Blake

Back when life was simple, one used their hands to work and advertised with a hammer, a stout tree and a piece of paper.  But now, there are SEO specialists, who are different than SEM specialists, who are different than XYZ specialist (ok, maybe not that last one but you just wait!)   The fragmentation of advertising channels and techniques is ever growing.  There's just an overwhelming amount of learn; best to start with the major players.

Today I'm going to focus on Facebook, which much to my chagrin seems to be taking over more of my life every day.  Facebook is very much a dual edged sword.  Very high usage, inline social proofing (comments/shared) and the ability to deeply slice your demographic.  On that last one, you can slice by interest, life event, even job title, allowing you to tightly target your ads.   

The downside is you can spend a lot of money on chasing down loose ends and Facebook doesn't make it easy to do complicated A/B testing (and I've yet to find a good App for that.)  For this reason, among others, Facebook is notorious for poor returns in terms of advertising.  With that said, I'm going to share a bit of how I set up a A/B test for a complimentary business of mine utilizing the wood pendants I use for keepsakes for elegant pill cases for men (TwistedWoodPillCases.)

I'm not going to get into the step by step process for creating an ad, Facebook covers that quite well via online help in their ad creator here.  What I will talk about is the thought process for deciding who to target ads for. 

My Twisted Wood Pill Cases are these little wonders.  About 2" in length, 3/8th inch diameter and hold 4-5 pills:

Pill Case Key Chain at www.twistedwoodpillcases.com

Pill Case Key Chain at www.twistedwoodpillcases.com

I came up with this product after my own challenges keeping a small aspirin or other pill on my person.  Often I shove the pill in my little jean pocket or in my wallet.  Not very sanitary and often just forgotten about.  I saw my keepsake pendants could double for this usage. 

In any case, I could just make an ad campaign for everyone, but clearly there are better ways to handle this.  But how exactly do I start building a demographic profile?

Warning; it is very easy to spend money on Facebook Ads and get very little market traction.  Everyone advertising on Facebook needs to be aware of not falling into the 'Like Trap.'  The like trap is where you get a lot of likes on your product, but no sales.  Bad - because without sales you eventually run out of money and then there are no more likes.  I suppose it's sort of like gambling or the fame game addiction, but with emoticons...

So back to figuring out demographics.  The first way, useful if you don't know who your market is, would be simply advertise to EVERYONE and then after a week looking at your demographics tab to see who expressed the most interest in your product.  Keep in mind who expresses the most interest MAY NOT be who buys the most.  We're simply starting out here so this information is valuable.

The second way is more appropriate if you already have a brand identity and know roughly who you want to market to.  In my case, I want to build a fun off-beat fan targeted to mostly men from 35-55; those who are old enough might need to take a pill every now and then but are still young enough to care about looking stylish doing so.  As such, I set my marketing to 35-55 men nationwide.

As mentioned, my brand is off-beat humor; as such here is one of my ads

So a bit more on demographics.  I'm going to target this to those who like beer OR hiking OR hanging out AND who also are making between $50K+.   Now here comes the tricky part.  Let's say I get results; which one of those demographics are responsible?

To answer these, we do A&B testing.  Whether you decide to do this under a single campaign or multiple campaigns is up to you (I'm aware of no upside/downside either way) but you'll want to make a different Ad Group (this is due to Facebook sets demographic targeting at the Ad Group level); in my case one will target beer, one hiking, and one hanging out. I still want to AND this with $50K+ income.  Generally I set the spend to be equal between the groups.  Additionally I use the same ad copy for each ad.  Like with any A/B testing we're trying to reduce variables to ONE - in this case demographic interest.

Now remember the Like Trap?  Again, very important to make sure you're optimizing for sales verses interest.  There are a number of ways to do this in Facebook.  Assuming you have a Facebook Pixel set up you can do a custom conversion on your cart.  I would also set a custom parameter such as ad=beer so you can also utilize Google Analytics.  Once these are all set, off to the races we go.

Now you have other options, such as scheduling, optimization options and what not.  Regardless of your choices, make sure they're the same.  I also would like this run for at least a week.  If money is an issue, I would spend maybe $5 a day BUT run it for at least a week to help take some of the daily variability out of the equation (you can test temporal variables another time.)

At the end of that week (or two or four) it should be pretty clear via your conversions what is working and what isn't.  You may find some demographics are less frequent purchases but the purchases themselves are larger, and conversely.   Rinse, repeat, and you'll surely get closer to your optimal ad mix.

by David Blake
Owner, Spirit Pieces
www.spiritpieces.com