When you run a small Facebook campaign, things are simple, variables are few, and measurement is easy. But as you scale up, there are layers of complexity that are invariably added to the mix. Some people would recommend just setting up a small campaign and increasing the spend; and that’s an option, but there is a lot more potential here. If you handle a big spend wisely, you can be drastically more efficient and effective.
1) Target Target Target
Targeting the right audience is always critical, for big or small campaigns. But it’s actually easier to target a small niche audience than a larger niche audience. If you’re looking for campaign reach into the millions, you need to re-think your targeting. Ideally, you want your target audience to be at least 400% larger than the total number of people you plan to reach. Since Facebook only serves ads to active online users — instead of total existing users — you must be sure there is a large enough pool of active users to get optimum efficiency out of your buy.
2) Test Small Before You Go Big
Testing is not just important, it is critical. I recommend setting aside about 10% of your budget for testing. But also set aside time for testing. If the full campaign needs to roll out in 30 days, testing needs to start next week. Plan a few weeks for testing from day one. A month of testing is even better.
In my experience you cannot get a pulse on a concept with less than a week of this measured experimentation. Two weeks is better. I recommend a “wide net” initial two-week test; refine your test campaign and test for a second two weeks. Then pick the top performers and launch into the full campaign. Keep in mind that, even in the full campaign, you never stop testing and optimizing.
3) Lots of Combinations
When you go to test, include numerous creative combinations and options. You will occasionally be surprised which options will be the top performers. Have multiple sets of copy and plenty of images. And do not be afraid to try every available permutation of text and pictures, or at least the ones that make sense.
If your final campaign will consist of 10 ads, I recommend testing no less than four options for each of the 10 — so 40 total options. Ideally, you will test more. For the first two weeks of testing, you will have 40 options; for the second two weeks, you will optimize down to 20 options, and then you will launch the campaign with the best 10 options. You can pair down further as needed.
4) Naming & Organization
Do not, I repeat, do not go with the default campaign and ad set names. Create descriptive and detailed names for everything. When you set up a campaign, you may think that you will remember what was in ad set 1 and ad set 2, but after a week of running tests and data your brain will be scrambled. Label and even date each campaign and ad set. Then when you go to report on things, match those labels directly to your report. The report at the end will be so much easier if you labeled well at the beginning.
5) Allow Time For Auto-Optimization
When you set up an ad, you can choose the action that Facebook should optimize for: post engagement, website clicks, purchase conversions and so on. I recommend selecting and setting up the appropriate optimization parameters.
Keep in mind it is in Facebook’s best interest for your campaign to succeed, since they want to avoid you taking your budget elsewhere. So if you tell them what you need to optimize for, set up the appropriate tracking codes, and give it some time, they usually do a very good job at it. This instantly sets things going in the right direction and there are long-term benefits.
I usually see auto-optimization make significant improvements over the course of weeks, not days. But keep in mind that auto-optimization only helps by a few percentage points, so you need to work to be sure you are in the right range before setting anything on autopilot. Also, this never takes the place of manual optimization in terms of creative, target audience, and timing.
Big Facebook campaigns are part science and part art. Plan for both, and win big.