Your websites and online ads are a treasure trove of data that can be mined for new business value. If you’ve never dipped your toe into analytics or data mining, you should make it a point to try!
Today, I’m going to introduce you to a few of my personal techniques for mining a business’s online data for new marketing ideas. I find that the deeper your understanding of your business’ metrics, the more likely that you’ll make decisions based on hard facts instead of speculation.
My focus is going to be on your website, but you can apply the same principles to direct mail and offline sales.
Step 1 – Start creating your data streams
Everything starts with creating various streams of metrics that you can look into.
My first and best recommendation for an analytics tool is Google Analytics. Google Analytics is an amazing tool that tracks an incredible variety of user behavior and sales on your website. It’s incredibly easy to set up for your website, and very comprehensive. When you start using Google Analytics, ask yourself:
– Which online channels is my web traffic mostly coming from?
– Which are my top performing web pages?
I also recommend mining your financial/accounting data for information. This is an oft forgotten set of data that is hiding some good nuggets of insight. When you use your financial data, ask yourself:
– Who are my biggest customers?
– What is the lifetime customer value of my customers?
– What are my top performing products?
– Have any major customers stopped ordering (and can I win them back)?
Try out some great online tools. There are a number of powerful tools that can audit your website and give you more data to work on, ranging in your comfort with online marketing data.
I’m a big fan of analytics. Here’s what I figure: the more data I have access to, the more potentially valuable nuggets I can find.
This is a step that most people skip, and one I find very valuable.
One of the things people don’t fully grasp about data mining is: most of the data you’re going to look at is boring and/or useless. You have to “walk the entire landscape” before you know what’s interesting to look at.
I recommend that you spend time to really get to know your tools: spend a few hours to click through EVERY page, menu option, axes in the graphs, and columns in the tables. It’s not until you’ve walked through every nook and cranny of the tool and thought about it, that you really understand the few graphs you really need to use.
Next, I urge you to understand each of your analytic tools’ limitations and to look beyond your obvious tools.
Data are your materials, and the analytics tools are literally tools; they have strengths, weaknesses, and hard limitations. When novice data analysts gain access to an incredibly comprehensive tool like Google Analytics, they begin to think that the tool is omniscient. These novices become close minded: “since it’s not in Google Analytics, I can’t answer that question.”
Here is an example I witnessed recently: Google Analytics doesn’t track profits on your orders; it only tracks revenue. Because the profitability of this product segment was so valuable, it didn’t matter that he compute the information by hand. The analyst literally wrote down every order number that made it into Google Analytics’ report, computed the profit manually, and generated the new report via Excel.
The presentation was a success, and could not have been done via Google Analytics by itself.
Step 3 – Focus your analysis (know what you’re analyzing)
There are a plethora of things you could be analyzing for when you get into more advanced analytics.
However, I’d start with these 3 things:
– Analyze your pages (once they’re on your site)
– Analyze your acquisition channels (what brought them to your site)
– Analyze your products/transactions (what they’re buying)
With so many pages, graphs, and charts, it’s easy to get lost. Focus on the main objectives!
Step 4 – Try out my favorite metrics (know where to look)
I admit—I play favorites!
I have 3 basic metrics that I favor when it comes to reporting:
– Sessions (traffic volume)
– Conversion Rate
By focusing on these 3 metrics, (and a solid understanding of your sales funnel) you can quickly generate a basic understanding of your business.
Of course, you have to bring in all the other metrics to tell a more complete story! As you become more experienced, you’ll find more and more insights by considering the other metrics.
Step 5 – Reading between the lines (know what to do)
Having a bunch of graphs isn’t very useful if you don’t know what to do next!
In the last section, I highlighted a few metrics I recommend that you focus on. But what are some of the common things people do with those metrics?
Here are a few general ideas on things that you can do:
Invest more into your top performers (i.e. products, ads, advertising channels, and pages), or create similar variations. If something’s making money, you should try pouring more into it! Perhaps you can replicate your strategy and add a brand new revenue stream to your business.
Shore up (or cut!) your worst performers. Remember: not all business ideas are profitable. Try to improve things, but cut if you believe they’re a lost cause.
Keep an eye out for stragglers or outliers that may be the start of a new idea. I often find that as I generate more data, I’ll come across some free radical that sparks a new idea or a new line of questioning. It could be a lone visitor from a website that I hadn’t seen before; this could spark a new line of thought about new web channels. I’ve also investigated web pages that was getting a lot more traffic than it should be, only to find traffic sources I had never considered.
Step 6 – Slicing and dicing (with segmentation)
The first 5 sections are good for the basics of data mining.
The last topic I’ll cover today is segmentation, which allows you to cut and reassemble your data so you can view your data from new perspectives and uncover new insights. It allows you to “zoom in” to specific segments to look at, or “zoom out” to see the summary view.
To explain segmentation, take a look at the following screenshot (click to zoom in):
Conclusion – Keep asking questions
I’ll finish this tutorial with one final thought.
I’ve found that the best way to “look for insights” is to constantly ask questions about your business. Ask yourself broad questions like, “I wonder how effective my email campaign is doing”, or “how much money is coming in from being ranked on Google”? And also ask yourself specific questions like “what pages are performing best on Google rankings”.
One of the key tricks of data mining is: insights are only there if you’re looking for them, and you can only do that if you’re asking the right questions.