The Perfect Campaign: Test, Optimize, Repeat
Take a hundred dollars and turn them into a thousand. That’s the basic principle in marketing, right?
No matter if you work with clients or an employee in some company’s marketing department, at one point, we all hear something along the line of “I want you to get me 10 actions every day right from the start.” My answer is usually: “There is no magic formula. Marketing is A/B testing”
And that’s exactly my point. Marketing is all about data. You make a decision based on data you get from clients, purchases, leads, conversions or educated assumptions. It may be an assumption regarding your target audience.
Let’s say you are launching a jewelry store online. Who is your target audience? To asses that, you need to get input from the designer: Who is the jewelry made for? Then, you can create a sample targeting group that matches the demographic information the designer provided. This will be your first test. Once all the creatives are done and live, you are able to measure the results by a few factors.
• Click-Through Rate (CTR): The average between the number of impressions and the clicks the ad got.
• Quality/Relevance Score: How relevant are the creatives and landing pages for the targeted audience? This number is usually rated one to 10.
• Conversion Rate: What percentage of the users who clicked on your ads actually purchased?
• Time On Site Or Bounce Rate
By assessing the results for the different creatives (at least three to four for the first test), you can choose the best-performing one for that audience. Or you may realize that you need to adjust the targeting if, for example, the quality score is low (below six), the CTR is OK but the conversion rate is low and the time on site is short. That will tell you that this is not the right target audience. However, if the quality score is good, the CTR is good but the conversion rate is low, you need to analyze the traffic performance on the landing page and optimize it.
This is just one example of how you can test an assumption and optimize based on that. When you are trying to create a marketing strategy, always question your assumptions. Never promise anyone a magic campaign or huge results with a strict budget.
So, how do we set the right budget for the audience? What I like to do is look at the target audience’s estimated size. Most tools will also give you the exposure-per-set budget, so you can choose the right amount of exposure per day or overall for the test period. There is not a minimum budget to run with, but I never go for less than a few thousand impressions per day. I find that this will give you the best exposure for the test. It may be $20 per day or even $200 per day, depending on the system, audience, competition or keywords.
Many clients expect to spend very little or as little as possible and get amazing results. Never make promises you can’t keep. And until you have tried a target audience and a product, you never know how it will work. Even if you work on a product that already sells, you will be expected to show better results than before, and this is the exact same thing as working on new objectives.
The best campaigns likely had many variations before they saw amazing growth and success. Make sure you always have a few approaches ready to go. The whole point of A/B testing is to test two or more different assumptions. This may be a combination of different creatives, audiences and landing pages. By doing this, you allow your budget to be spent on two different assumptions and optimize the one that works better. Sometimes, you even get good results with the two approaches and it’s a win-win situation.
The thing that never changes is that you need to base your decision on data. You begin with an assumption or past performance data, you go into real-time performance optimization, and then you optimize based on results and do it all over again.
A campaign has to be optimized all the time. Treat it like a little baby — it needs care and love or it will die.
Looking for funding? Apply here.