3 Refreshing Ways To Look At Marketing Analytics




  • It’s not just for data scientists anymore. Columnist Brian Massey shows us how to get creative and explore analytics from some unexpected sources.




    analytics-technology-ss-1920


    My numbers are down. After spending my last few columns in deep dives on report sampling, tracking account holders, and measuring video, I thought I would return to a more motivational topic for this column. Sharing on these very important columns has been down, with less than 400 shares each. By comparison, my Google Analytics Checklist column has garnered almost 4,000 shares.


    So, I’m going to attempt to expand my audience, and talk about the fun you can have with analytics. When we stop thinking of analytics as a discipline of data scientists, we free ourselves to make good decisions with data from unusual sources.


    1. Don’t Take Things So Seriously

    The opening statement above is an illustration of data-driven decision making that goes counter to much of what I teach here. I’m citing a metric — shares — that only hints at my real metric.


    My sample size is minuscule. The entire effect could be due to a seasonal downturn.


    My message for you is, “That’s OK.”


    I have formidable powers of analysis, and I wear a lab coat at work. Yet, I’m not going to feel guilty if my conclusions are only a step away from a coin toss.


    Why? Because the decision was not a coin toss. I drew a conclusion based on the data I had, and for which collecting more information would not have been worth the effort.


    You never have perfect information when trying to predict the future. All you have to work with is a preponderance of evidence. Let’s have a little fun “getting our preponderance on.”


    With my next few columns I will test my conclusion that you don’t want deep analytics articles. If you want more articles like this, then please share. I’ll be watching.


    2. Expand Your Definition Of Analytics

    The first lesson to be learned here is that analytics are everywhere. A repeatable, statistically significant, variable isolated split test is hard to beat for making good decisions. It’s overkill if you’re trying to figure out what to order for lunch.


    Instead, you gather some data. You ask your lunch date what he or she has liked in the past. You ask the waiter what’s most popular or most praised. You look around the room at what others are eating. You reference your memory of similar dishes you’ve had at other restaurants.


    Then you order lunch based on the analytics.


    You can do this same thing even if you work in the best-funded online marketing department on the planet. A few ad hoc data points may be sufficient to dismiss a minor idea for your website.


    One of our clients recently came to us with a hypothesis. Their lead generation form included an optional comment field that wasn’t essential to the sales process. The details put into this field will come out during the sales call.


    We’ve seen that removing fields consistently increases form completion rates. Why not take it off? A split test would answer the question once and for all. But where else could we look for some data to support our hypothesis?


    Would removing this field increase conversion rates? Is it worth testing?

    Would removing this field increase conversion rates? Is it worth testing?


    We decided to pull a sample of records from the lead database and see how many people were using this field. If few filled it out, we could feel confident deleting it without wasting a test cycle. This is analytics.


    Our analysis showed us that between 35% and 50% of visitors were filling out the field.


    These visitors are probably more likely to get through the next step of the process after spending time on this field. We can’t just yank it.


    We then looked at some session recordings. We found that the field didn’t seem to stop anyone who didn’t fill it out. Videos of people using the site is analytics.


    This information took an hour or two to collect. It’s not statistically significant. It’s not rigorous. But it’s enough for this decision. It’s analytics. In the end, we decided to leave the field rather than waste weeks on a split test. We looked elsewhere for a good hypothesis to test.


    Other Fun Sources Of Analytics


    Analytics are everywhere. Here are some other sources to consider.



    Look at the places people click using click-tracking software.


    Go talk to some of the salespeople about what questions they answer over and over.


    Call some of your customer support people to see if callers seem to support your hypothesis.


    Download the live chat transcripts to figure out if your copy changes are worth testing.


    Pull all of the ratings and reviews for a product line to see what makes people give low or high ratings.


    Use your sales Customer Relationship Management (CRM) system to support hypotheses about geography or job title.


    Look through the reports in your analytics package to see if new clues can be found from different angles. Browser type, time to purchase, and navigation flows are just some examples of interesting ways to slice your data.


    Have your accounting department pull the daily returns to see if you’re setting proper expectations with your buyers.


    Count the comments on your blog posts to see which topics are hot.


    Do a study. Temporarily measure the minor actions that your visitors take — a click on an image, the number of error messages generated by a form — to get data you need without testing.


    Look at the email click-through rates of different email offers to see if your ad copy ideas will work.


    Pull a bunch of data into Excel and see if you can pivot your way into answers.


    Many times, these data sources won’t reveal Earth-shattering insights. Yet, with hints from two or three of these ad hoc sources, you may be able to toss out a hypothesis, or you could find the support needed to justify a full split test.


    What unexpected sources of analytics have you discovered? Let us know in the comments.


    3. Analytics Is Like A Scavenger Hunt

    Once you’ve started to explore this new world of “unalytics,” you will get very good at asking, “Where could we get more data about this?” It starts to get fun.


    You’re no longer a marketer struggling to make sense of analytics data. You’re on a scavenger hunt. You have to be creative about where you’ll find the little nuggets of information that complete your confidence list.


    Scavenger hunts are fun. And so is playing data detective inside your company. It will save you from unnecessary tests, strengthen your ideas, and make you look very, very smart.


    You’ll also get better at making decisions in everyday life. Go ahead and order the fish tacos.


    Listen To This Article

    Play (Duration: 7:46 — 7.1MB)



    Some opinions expressed in this article may be those of a guest author and not necessarily Marketing Land. Staff authors are listed here.




    About The Author







    Brian Massey is the Conversion Scientist at Conversion Sciences and author of Your Customer Creation Equation: Unexpected Website Forumulas of The Conversion Scientist. Follow Brian at The Conversion Scientist blog and on Twitter @bmassey


    (Some images used under license from Shutterstock.com.)

     


    Marketing Land – Internet Marketing News, Strategies & Tips

    (197)

    Leave a Reply