User Statistics 101

Introduction to User Statistics

The Value of Statistics

I write this intro coming from a background in mainstream where big businesses spend big dollars on all manner of statistical tracking, analysis, and trend prediction. "Online" centric companies devote enormous amounts of resources towards learning their end users' patterns and habits. The resulting statistics are used to craft virtually every page of a website with the single intention of making them sell, whatever it is they're selling, well. Be it a product, a subscription, a service, or an idea everything producing revenue is selling something. When we start talking about selling a subscription we automatically infer an even greater amount of statistical tracking that should be employed. It's one thing to sell a one time product to a visitor it's another to retain that customer for an indefinite amount of time. We'll get more into that later for now just keep in mind that every major online player lives and dies by the numbers for a reason. They are the only accurate insight into your customer base that you have. Everything else is an educated guess. I don't care if you've been in the industry for 30 years. I don't even care if you just came from XXXX site which is an identical model. It's a different site, different traffic, different user base, different content. However similar it is the numbers can be completely different. You don't know until you capture and test.

Capturing User Data

Everything starts with capturing data. Until you have the necessary mechanisms in place to begin capturing and reporting information about your users' habits you have no insight into the most important aspect of your business, the consumer. Without this insight you are firing off educated guesses in the hopes that something sticks and when it does you'll ride it out until you notice a decline and then repeat the processes. Accurate statistics and reporting allow you to break this cycle of mediocrity in favor of a system which allows you to make informed decisions ultimately resulting in the ability to maximize your site's potential to convert.

Before we get start talking about what types of information we should be capturing and how we should be reporting it, let's talk about broadening the scope of who we're capturing information on. One serious oversight made by the adult industry is the true value in a non paying visitor. In the adult world it's convert or die. I envision buckets of cash being dumped onto a table whenever I stop to think about just how wide spread this problem is in the adult space. You are streaming tens of thousands of interested customers into your store and unless they're willing to open their wallets you're willing to just watch them walk away. Every study has shown that porn is an impulse buy but that doesn't mean there aren't potential customers in the ones you just let walk away. It's only logical that a percentage of those users will convert over time... when offered incentives that percentage can be drastically improved. Combined with a subscription based model you're looking at some serious dollars. How many customers did your site let walk away yesterday, last week, last month?

The next logical question is how to do we capture these users and market to them? If they aren't paying for an account they can't sign up. Well that's a serious problem but an excellent place to start! User's aren't about to start throwing their personal information at you however so you're going to need a little incentive to seal the deal. Why not offer a mini free user's area that contain only selected FHGs? They're all in one place and offer you a chance to show off your member's area while upselling every step of the way. Mainstream has been nice to us in this respect as it has trained users to feel "over abundant upsells" are ok for "free" areas. If you aren't proud of the format or design of your member's area then that should set off a red flag for you; address it because it's only hurting your rebills and certainly won't help you sell any new subscriptions.

What to Capture?

Demographics - Now that we're capturing data on a broadened scope of members it's time to start looking at what to capture. We'll start off with simple demographic information about your users. Any demographic information you believe will help you make a more informed decision about what to show any particular user to better sell your product should be captured. Be careful not to throw a ton of questions in your registration or users will be more likely to bail out however, some simple standard questions can go a long way; date of birth, gender, location. Beyond the basics you'll need to create some additional ways to get your users to answer questions about themselves. Be as creative as you'd like and in the adult industry you get some really fun options like... Blondes or Brunettes? Mix it up and don't bore the user!

Applying Demographics to Marketing - Gender and sexual orientation alone can go a long way towards helping you create more effective marketing on your site. Was it a horney housewife searching for that large cock site or straight male? The real question is whether displaying sexier girls or larger cocks will appeal more and/or differently to either market segment. What if one of your affiliates is sending you traffic very heavily focused on a particular demographic? Should they not be informed and a special default landing page assigned to their traffic? To determine if the two market segments mentioned above would benefit from a more targeted landing page we need to call out one of, if not, the most important tool in the statistical toolbox.

A / B Testing - Wikipedia defines A/B or "bucket" testing as "a method of marketing testing by which a baseline control sample is compared to a variety of single-variable test samples in order to improve response rates". The landing page or pages you use right now are you baseline control sample. You likely have data dating back to the start of your site which shows how well that page performs, it's conversion rate. The most simple A/B test you can perform is to swap out that landing page with a completely different one about 50% of the time. You can not completely replace the current landing page as traffic is continually different and you must have your control group to compare against. From here you could analyze both conversion rates over any given period of time and determine which page is most successful. This is a great start but what you should be asking yourself at this point is, what about this page makes it convert better? Once you start breaking a page down into its separate elements, (e.g. images, copy, layout, color) you can begin to run a myriad of comparisons each in turn with its own valuable set of data. The object here is identify those improvements which consistently show improved results; thus creating the highest performing most optimized page at your disposal.

I could write for days on just this single topic and it certainly deserves your additional attention as a business owner and/or webmaster. There is no replacement for good A/B testing and the more time you invest the better return you'll see. If you ever reach a limit where you feel you can not optimize a page any more than you should first try hiring a completely different designer for a page or two. Add something fresh into the mix and see what happens. Of course this doesn't apply to landing pages only. Any page in any registration or conversion path should receive the same treatment. This should be an ongoing process throughout the life of your business.

Reporting
Throughout the lifetime of your company, as you continuously add additional tracking points, the need for more clear and concise reporting will become necessary. As the bank of information grows you'll undoubtedly want to have new ways of overlaying and comparing data sets. If you have the capital, invest in a good solution now. If not, keep it in mind as you grow your business and plan for it.

Having easily accessible reporting available to anyone who needs it is invaluable and many large tech companies open the reporting up to all employees. I've seen individuals not involved in the decision making process spot trends others, who were directly involved, missed. When you're looking at hundreds of different data points any new set of eyeballs is welcome. It's easy to get lost in the seemingly endless sets of numbers but it's important to remember that everything should be proven with simple A/B testing. Only then will your hypothesis be proved or disproved and only then can you begin leveraging your new found knowledge backed by hard numbers.

User Surveys
One final piece worth mentioning that you may wish to incorporate is user surveys. These are tough sells for any site no matter what the product and can usually only experience good results from heavy bribery on your part to the user. That being said, exit surveys for member's cancelling their subscription can be invaluable to you as a subscription based model. If you aren't seeing a good survey completion rate offer to cancel their membership but extend the date they can access the site by a week or so. It's an extremely small cost for such profitable information.