Marketing research you’re leaving behind

You’re paying for enormous amounts of research you’re not using.

I can make this bold statement because I’m guilty of it, too. I’ve paid large sums of my own money for research I didn’t use as well as I could have. What is this research? I’m speaking of paid advertising.

Right now, I’m in the middle of marketing my new book. I’m running Facebook and Twitter ads to capture attention and build interest. To advertise, I have to write copy and select images to create the ads. Most marketers write up the ads, turn them on, and walk away. After the campaign is over, you shelve the ads and move onto the next campaign. What a waste!

You’ve just paid money for research. You’ve just paid money to find out what words, phrases, and images resonate with your audience:

Campaign_overview_-_Twitter_Ads.jpg

What should I pay attention to? Clickthrough rate tells me what’s working best with my audience. Above, the photo of me gets a higher clickthrough rate. There are ads not shown that have the same text but a different image. Those ads are performing half as well as the ones with me in them. The ads beginning with copy about marketing – a noun – are doing slightly better than the ads starting with a verb. I can see the beginnings of a trend here. In a week, the statistical validity of these ads should firm up and I’ll be able to develop writing strategies from them.

So how do you make use of the research you’re collecting?

First, establish statistical significance. I recommend Rags Srinivasan’s excellent Excel template to run the test.

  • Punch in your impressions in the first row for any two ads.
  • Input your clicks or conversions in the second row.
  • The calculator will tell you whether the result is statistically significant.

Screenshot_1_27_15__7_02_AM.jpg

Second, after you’ve established statistical significance, incorporate the results in other forms of media. Assume the above results were statistically significant. There is a pattern in which ads leading with nouns about marketing trumped ads leading with verbs. I have two choices for a blog post title: “Marketing research you’re leaving behind” and “Use the marketing research you’ve paid for”. Based on the research I’ve paid for already, I’d choose the former.

Third, look at your web analytics over time to see if blog posts you’ve written using prepaid research perform better than average. If they do, then you’ve increased the indirect ROI of your ads. Even if your ads didn’t generate great performance, you’ll have repurposed your findings to improve other parts of your marketing strategy. That’s a win!

Make the most of every dollar you spend on your marketing and advertising. You’ve already paid for the research — use it!


If you enjoyed this, please share it with your network!


Want to read more like this from ? Get daily updates now:


Marketing Blue Belt Preorder

Subscribe to my free newsletter!


Marketing analytics tools for non-marketing uses

There’s obvious professional benefit to mastering your marketing analytics tools, from data visualization to complex statistics. You can improve your marketing programs, grow your business, and make more money. However, often we just do the same rote things in our marketing analytics. What can we do to improve our skills? How can we get better at making our analytics tools work for us?

What if we measured something else?

One of the most powerful trends right now in data collection is around the quantified self movement. This is where you track lots of different data points. I own, for example, a Basis wristwatch that tracks things like steps, heart rate, calories, etc. The watch itself comes with a halfway decent web reporting system that gives you broad information about how you’re doing. This is akin to many marketing analytics packages giving you pre-defined reports and visualizations. It’s a good place to start, but like most marketing, what if you want to dig in?

Fortunately, as with any good marketing analytics software, fitness tracking software can let you export your data. Using a free, open-source package from GitHub written by Bob Troia (aka Quantified Bob), you can export minute-by-minute information about your life from your device. Certainly, you can’t process that amount of information as a normal human and glean any insight. What you can do, however, is feed it all into your marketing analytics tools, from simple Excel spreadsheets to Big Data tools like Hadoop, MapReduce, and Watson.

For example, here’s a simple visualization of calories burned versus heart rate (top chart), steps taken (middle chart), and air temperature (bottom chart):

Tableau_-_Book2

You don’t have to be a data scientist to figure out what you’re looking at. There’s an almost perfect correlation between steps taken and calories burned, which makes base logical sense. The more you move, the more energy you use, the more calories you burn.

Here’s a more complex example, asking IBM’s Watson what influences calories burned.

Watson_Analytics

Watson obviously picked out that steps matter most, the first part of the decision tree. It then picks out heart rate as the second factor that influences calories burned. What’s interesting is how the tree splits off there. For standard “office life”, where my heart rate is between 73 and 81, skin temperature matters. Being warmer is slightly better. For exercise periods, air temperature matters, and there, colder seems to be slightly better.

Are these causal? As with all correlations, the answer is not necessarily. Causality and correlation dine at separate tables, but now we have enough data to begin experimentation. Maybe the next time I work out, I drop the air temperature even lower than it normally is.

We can do other kinds of experiments as well. If I visualize steps to heart rate as a moving average, then look at a typical day, I can see that just getting up and walking around more brings up the calories burned. Even relatively few steps keeps the engine running, so an experiment would be to try to walk around much more frequently. Instead of once or twice an hour, get up and walk around every 10-15 minutes.

What’s wonderful about using marketing analytics tools for non-marketing uses is that it gives you the freedom to explore and learn your tools in a different context. You have much more control over your fitness and activity than you do the open rate of an email campaign. You can choose to get up from your desk much more easily than you can choose to create content to be retweeted. Thus, for learning how to spot potential causality and designing experiments to establish or disprove causality, fitness is one of the many areas you can practice with your marketing tools. Then, when you go back to marketing, you’ll have an entirely new perspective and lots more experience with your tools.

Give this a try – and remember, it doesn’t have to be fitness related. You can use any data set that’s in good condition, and chances are you or your employer have all the tools you need. You could put in your income and expenses, then use your marketing analytics tools to spot patterns in order to save money. You could put in sleep tracking and see what impact a good night’s sleep has. You could track growing conditions in your garden to see if you can manipulate environmental variables like soil pH and water to get a better garden. The only limitation is your imagination.


If you enjoyed this, please share it with your network!


Want to read more like this from ? Get daily updates now:


Marketing Blue Belt Preorder

Subscribe to my free newsletter!


The sunset of keyword-based SEO

In a tacit acknowledgement that (not provided) basically killed keyword-based optimization, Google Analytics over the weekend relocated organic keyword tracking, burying it inside the campaigns menu. Why the change? For years now, fewer and fewer keywords were being picked up by web analytics software. Encrypted search and mobile are the two reasons they’ve vanished from our radar; as you can see, just in the last 30 days, more than 90% of keywords are coming in as (not provided):

Organic_Search_Traffic_-_Google_Analytics

So what’s a marketer to do? A few things. First, be sure to set up Google’s free Webmaster Tools for your website. While you won’t get keyword lists per se, you will get the queries people type into Google for which Google displays your site:

Queries_-_Google_Analytics

If you think about this, this is what Google has said you should be aiming for; not individual words or tiny phrases, but the actual topics for which you’re relevant. With services like OK Google and Siri, search queries will continue to get longer and longer. In the last year, the number of words in search queries for my personal site has increased from 2.99 to 3.44, and the number of queries has exploded from 1,600 to nearly 5,000. Queries are getting longer and more diverse.

Here’s 2013 (scale adjusted to be equivalent) search terms by number of words:

Tableau_-_Book2 2

Here’s 2014:

Tableau_-_Book2

Did my site suddenly get more popular in 2014? No, but the diversity of terms that people used to find me exploded. There were more 4+ word terms in 2014 than there were all combinations in 2013.

So how do you take advantage of this trend? The reality is that you can’t do keyword stuffing and narrow-focused keywords any longer. You have to expand to focus on the topic that you want to be relevant for. The reason is that you can’t accurately predict what people are going to search for. By writing topically, rather than focused around just a handful of keywords, you’ll be more likely to show up in search for the longer, more complex queries.

Think human! Look at your own search history, as an example. Look at how you search for information that’s relevant, and then model your content based on how you naturally search. Use Webmaster Tools, Quora, and Trends to expand your topic horizons.

The narrow-focused keyword SEO of the past is fading away. Be ready for much broader search horizons!

Updated: Vincent Tobiaz pointed out in the comments that the original screenshot was wrong – keywords got buried in campaigns instead of being removed entirely. Thanks!


If you enjoyed this, please share it with your network!


Want to read more like this from ? Get daily updates now:


Marketing Blue Belt Preorder

Subscribe to my free newsletter!