Fail and Fail often.
Having launched hundreds of websites in my life, one thing is certain; I have no idea what works. Of course, you wouldn't expect anyone who makes a living developing online businesses to admit that -- but that is the truth. Nobody knows exactly what is going to work. The internet has proven to be a great medium where your business can fail miserably. We often think to ourselves that we can replicate previous online business models, and cash in like all the other internet entrepreneur legends.
Truth be told, I love it when I fail. Failure is the tactic in which I find out what works best.
You should be failing as much as possible.Of course when I say failing, I really mean testing. We should be testing everything that we do with the goal of optimizing our products, to provide a better service to our customers. As a business owner, your primary goal should be failing on a very small scale, with as little risk as possible.
Testing and optimizing your product is the essence of great marketing.
Market research falls into one of two classifications; problem-identification research, or problem-solving research. From there we go into subcategory market research types like buyer decision research processes, brand association research, ad tracking, price-elasticity testing, distribution channel audits, etc…
Despite all the methods and types of testing, we're going to start thinking like smart business people and dumb-down our questions to only the important stuff:
The Big Four1. Am I making money?
2. Where is my money coming from?
3. How can I make more money?
4. How should I use my money?
Your user interface, technology, branding, and messaging are all critical to your success. However, what matters most is that your business makes money. Testing is the way to do exactly that -- make more money! By continuously testing and analyzing your results you will understand how your audience is reacting to your content, and correct it accordingly to drive more sales.
Steps to TestingHere are the components of testing intelligently:
1. Build a system designed to capture behavioral data and other business metrics.
2. Define what questions you have.
3. Know who you are asking.
4. Identify and mitigate risk.
6. Analyze data & draw conclusions.
7. Adjust accordingly.
Types of Questions to Ask:
I'm going to categorize the types of questions you should ask as the following:
1. The basics 2.The Optimizers and 3. The Drivers.
The Basics are questions you should be asking, and metrics you should watch everyday:
How many people visit your site?, How many people are on your Facebook Business Page?, How much content do push out a month?, How many subscribers do you have for your email list?, What are your online sales? How many clicks did that banner get? etc..2. The Optimizers
The Optimizers are questions that you should be asking to help discover trends, and make the most of your efforts. These are a little more in-depth, but can still be implemented by yourself, with many FREE online solutions available.
What days of the week are you generating the most sales? Are people more likely to click on an orange button, or a red button? What time of day are people most engaged on Facebook?, What type of posts solicit the most engaged reaction?
These questions are absolutely critical and are now becoming the industry standard for every online business.3. The Drivers
There are higher level questions that should not only drive your business decisions, but define your business. These questions are more related to human emotions, and become far more difficult to quantify:
What will the needs of my audience be five years from now? Can we become more efficient in delivering value to our clients? Why do people buy from our competition and not us?
The questions of why people buy, and how you can motivate people to buy your product are easily answered by simply asking them. Questionnaires that are promoted with incentives are a low cost solution to engage audiences, and get information. I do caution that we must be always careful to not trust what everyone says, but what everyone actually does.
This concludes part one. Part two includes information on the metrics that matter, types of testing, and systems to test with.