09 Jun Harnessing the Power of Data: A/B Testing
Posted
in RedGIFs Tips 'n' Tricks
In my previous piece on harnessing the power of data, I mentioned A/B testing as a method to confirm your conclusions about your audience demographics. But what exactly is A/B testing, how can you use it, and what should you keep in mind about this research method, which is specific to successful marketing projects? Read on to learn more!
WHAT IS A/B TESTING?
A/B testing is a very old statistical technique that has its roots in testing hypotheses in scientific experiments. Think of how scientists use groups to compare how effective a treatment is, and this is where the roots of the technique began. It is the most basic form of a randomised controlled trial (RCT). A/B testing was first used to evaluate advertising campaigns in the early 20th century and has since become a mainstay of marketing, accompanied by elaborate tests of statistical significance to demonstrate which offerings perform the best.
Fundamentally, A/B testing is a way to compare two versions of something to work out which one is better. Here is a simplified plan of how to use this method:
- First, start by deciding what you want to test; this is known as your hypothesis.
- Second, gather information and data related to your hypothesis to create a baseline. This is important to check back after your testing to see what has changed.
- Next, map out all the variables and factors that may be influencing your hypothesis. Using this information, you can then create your A and B versions, which you will test to see which performs better.
- Finally, test these versions against each other, and gather the data to see which one did better according to your original hypothesis.
- And as a post-script… then use the information gathered in your first test to design your next A/B test.
Fairly straightforward, but this can seem really overwhelming to try and put into practice! Do not fret; in the next section, let’s talk about how you can apply this to your own content creation work.
HOW CAN I USE A/B TESTING AS AN ADULT CONTENT CREATOR?
In my last blog post on harnessing data, I talked about A/B testing in the context of providing you feedback on whether you are using your own site data correctly. As I said before, this kind of testing helps provide you with (fairly) concrete feedback on what is or is not working. As with any experiment, it depends on the design, making sure you have all your variables and factors controlled (or at least are aware of their impact) and evaluating your results as accurately as possible. This requires a bit of artistic flair and feeling, not just knowing the process inside out. You know your data and audience best, so let this guide you to the adjustments you need to make.
For you as an adult content creator, a simplified version of A/B testing is perfectly fine when you start out trialing this method. You can always make your testing and results more sophisticated, so you have nothing to lose by starting off with a test of a basic hypothesis. Here are a couple of examples, based on the five steps I outlined above:
- Hypothesis: “My ass worship clips are more popular than my foot worship clips” or “more of my audience is live at 10am on Tuesdays than at 7pm on Tuesdays”. Decide on the one platform you are going to use to collect data (eg social media, clip site, fansite etc)
- Baseline data gathering: “Calculate an estimate of how many people view your ass worship clips vs your foot worship clips” – you decide what measurement you are going to use eg video views, twitter/X post views. For our other hypothesis, gather data from sites that provide that level of detail to see what your current audience levels are at those days and times.
- Create A and B versions to test: Depending on the details of your hypothesis, create an ass worship promo and a foot worship promo. Keep all other factors the same (wording in your post, time and day that the promo is dropped), changing only “ass worship” to “foot worship”. Same for our second example, keep all the factors in the promo the same, just change the time that the promo is released on the platform you are focussing on. The versions you create are not just about the content, but you need to be clear on the details of when, where and how the content is released.
- Test and gather data: Run the promo as you designed it, and keep track of the data associated with the test. Test one version first, and then test your other version.
- Design your next A/B test: After the test is run, take time to reflect on the test, and make changes for your next A/B test. Your “ass worship” content may have won out over your “foot worship” content, but in your next test, pair one of these with another of your niches to see if you can confirm the results. With the example of live audience times, run another test comparing another day or time to see if your results are confirmed.
As you can see, A/B testing is more of an art than a science! It is about being aware of your content, your audience and what you are trying to achieve. The goal is to use this research method to fine tune your approach to your content and promotion.
GOLDEN RULES OF A/B TESTING
Although A/B testing can be used across a variety of circumstances and to answer a wide range of questions, there are a few “golden rules” to follow to ensure you get the best results from your test.
- Let the test run its course: a testing run is not a couple of days; you need to be doing at least a two week cycle (ideally making it a month so you have 2 cycles in there). Time/day remains the same to test in the period, but you gotta let it have a good chance of collecting the data you need to make an evaluation of its success.
- Keep it simple with the metrics you are looking at: you could gather a bunch of different data points to make sure you get the best out of your tests, but the reality is that sometimes you can have too much information and this can muddy the results of your tests. Keep it simple and make conservative adjustments to what you will include or leave out.
- Always retest (at some point): this is to ensure the results were not a fluke, and to consider the changing landscape within content creation. Things move quickly in the digital realm, and revisiting and rerunning the test within a reasonable period will help keep you on track.
Rem Sequence is an Australian adult content creator, blogger, and internationally published alt model. She has a background in psychology, philosophy and political science and worked in health and sex education, youth work and trauma counselling for almost two decades. Now, she works full time in the adult industry, as well as indulging her passion for arts, writing and music in numerous side projects.
In my previous piece on harnessing the power of data, I mentioned A/B testing as a method to confirm your conclusions about your audience demographics. But what exactly is A/B testing, how can you use it, and what should you keep in mind about this research method, which is specific to successful marketing projects? Read on to learn more!
WHAT IS A/B TESTING?
A/B testing is a very old statistical technique that has its roots in testing hypotheses in scientific experiments. Think of how scientists use groups to compare how effective a treatment is, and this is where the roots of the technique began. It is the most basic form of a randomised controlled trial (RCT). A/B testing was first used to evaluate advertising campaigns in the early 20th century and has since become a mainstay of marketing, accompanied by elaborate tests of statistical significance to demonstrate which offerings perform the best.
Fundamentally, A/B testing is a way to compare two versions of something to work out which one is better. Here is a simplified plan of how to use this method:
- First, start by deciding what you want to test; this is known as your hypothesis.
- Second, gather information and data related to your hypothesis to create a baseline. This is important to check back after your testing to see what has changed.
- Next, map out all the variables and factors that may be influencing your hypothesis. Using this information, you can then create your A and B versions, which you will test to see which performs better.
- Finally, test these versions against each other, and gather the data to see which one did better according to your original hypothesis.
- And as a post-script… then use the information gathered in your first test to design your next A/B test.
Fairly straightforward, but this can seem really overwhelming to try and put into practice! Do not fret; in the next section, let’s talk about how you can apply this to your own content creation work.
HOW CAN I USE A/B TESTING AS AN ADULT CONTENT CREATOR?
In my last blog post on harnessing data, I talked about A/B testing in the context of providing you feedback on whether you are using your own site data correctly. As I said before, this kind of testing helps provide you with (fairly) concrete feedback on what is or is not working. As with any experiment, it depends on the design, making sure you have all your variables and factors controlled (or at least are aware of their impact) and evaluating your results as accurately as possible. This requires a bit of artistic flair and feeling, not just knowing the process inside out. You know your data and audience best, so let this guide you to the adjustments you need to make.
For you as an adult content creator, a simplified version of A/B testing is perfectly fine when you start out trialing this method. You can always make your testing and results more sophisticated, so you have nothing to lose by starting off with a test of a basic hypothesis. Here are a couple of examples, based on the five steps I outlined above:
- Hypothesis: “My ass worship clips are more popular than my foot worship clips” or “more of my audience is live at 10am on Tuesdays than at 7pm on Tuesdays”. Decide on the one platform you are going to use to collect data (eg social media, clip site, fansite etc)
- Baseline data gathering: “Calculate an estimate of how many people view your ass worship clips vs your foot worship clips” – you decide what measurement you are going to use eg video views, twitter/X post views. For our other hypothesis, gather data from sites that provide that level of detail to see what your current audience levels are at those days and times.
- Create A and B versions to test: Depending on the details of your hypothesis, create an ass worship promo and a foot worship promo. Keep all other factors the same (wording in your post, time and day that the promo is dropped), changing only “ass worship” to “foot worship”. Same for our second example, keep all the factors in the promo the same, just change the time that the promo is released on the platform you are focussing on. The versions you create are not just about the content, but you need to be clear on the details of when, where and how the content is released.
- Test and gather data: Run the promo as you designed it, and keep track of the data associated with the test. Test one version first, and then test your other version.
- Design your next A/B test: After the test is run, take time to reflect on the test, and make changes for your next A/B test. Your “ass worship” content may have won out over your “foot worship” content, but in your next test, pair one of these with another of your niches to see if you can confirm the results. With the example of live audience times, run another test comparing another day or time to see if your results are confirmed.
As you can see, A/B testing is more of an art than a science! It is about being aware of your content, your audience and what you are trying to achieve. The goal is to use this research method to fine tune your approach to your content and promotion.
GOLDEN RULES OF A/B TESTING
Although A/B testing can be used across a variety of circumstances and to answer a wide range of questions, there are a few “golden rules” to follow to ensure you get the best results from your test.
- Let the test run its course: a testing run is not a couple of days; you need to be doing at least a two week cycle (ideally making it a month so you have 2 cycles in there). Time/day remains the same to test in the period, but you gotta let it have a good chance of collecting the data you need to make an evaluation of its success.
- Keep it simple with the metrics you are looking at: you could gather a bunch of different data points to make sure you get the best out of your tests, but the reality is that sometimes you can have too much information and this can muddy the results of your tests. Keep it simple and make conservative adjustments to what you will include or leave out.
- Always retest (at some point): this is to ensure the results were not a fluke, and to consider the changing landscape within content creation. Things move quickly in the digital realm, and revisiting and rerunning the test within a reasonable period will help keep you on track.
Rem Sequence is an Australian adult content creator, blogger, and internationally published alt model. She has a background in psychology, philosophy and political science and worked in health and sex education, youth work and trauma counselling for almost two decades. Now, she works full time in the adult industry, as well as indulging her passion for arts, writing and music in numerous side projects.
