Aggressive optimization has nothing to do with a hostile temperament, but it does involve being quick, decisive, and being willing to let go of any sacred cows. It is more comfortable to optimize over time, and often better from a scientific standpoint. But sometimes you don’t have time and need to produce results fast.
Data Is Your Friend
Whether you are optimizing display ads, search ads, Facebook ads, or even TV ads, data is your friend. The more data you have the more certain you can be and the easier it is to test. In fact, it is impossible to optimize without data. But what if you don’t have data? How can you make decisions fast?
2 Ways To Optimize Without Data
There are two main approaches, one involves aggressive guesswork. That can sometimes work but it depends solely on your experience and gut. But there is nothing scientific about that process. The other approach is aggressive optimization, which is more than skillfully handling limited data, it’s about pushing to create more data faster.
The three constants to all digital campaigns are time, budget, and goals. If I have a 90 day campaign I may be able to afford taking a couple weeks to optimize. If I have a 10 day campaign, that isn’t an option, I need to find what works and commit to it. If my first shot doesn’t hit the mark for the conversion goals I need to make changes fast.
Time Is of the Essence
First thing is first, you MUST have data, at least some data. If time is of the essence don’t take one shot at a time. Don’t create one ad set or one permutation, create five or ten and run them concurrently for a day or two. Optimize by cutting what doesn’t work and putting that budget into what does work or new tests.
If things don’t appear to be working quickly then don’t wait until you have enough data to be sure, create more ad strategies immediately and get them out for testing, create 20 more if you need to. Run them for a day or two and compare notes and instantly cut the poor ones and enhance the better ones.
Typically we test with a small part of our budget and then roll out more budget when we get things working well. If time is too short you may need to test with a disproportionately high portion of budget in order to forcefully create enough data to make optimization possible and utilize the remaining budget well.
Having a methodical scientific approach is always preferred, but the bottom line is at times aggressive optimization is needed and it may require putting some best practices on the shelf. Use your desperation to produce innovation.