Rise and Shine

Dramatically Improve Marketing Results with Event Processing Analytics

Google Intelligence EventsAlmost everyone that runs a website or blog, or handles online promotion in any way uses some type of analytics software. There are lots of benefits that can be derived from measuring your progress, and it’s so ingrained in online marketing that users give little thought to how it fits in their strategy.


Some more savvy marketers are now figuring out how to show verifiable ROI through measurement. Believe it or not, doing this has never been easier. Software now records more data points across more platforms than ever before. In a large organization with hundreds of business objectives, it would be hard not to generate some sort of ROI with an integrated online campaign; even a poorly run campaign. Have you ever heard that there are three kinds of lies? Lies, damn lies, and statistics.


In smaller organizations, I have saw professionals recount methods of proving ROI that were actually more creative than the marketing that preceded it. For web entrepreneurs the use of analytics usually equates to a compulsive need to check page views and traffic as often as possible. Showing ROI and getting a handle on the different metrics that are measurable are important, but both practices disregard the most powerful use of modern web analytics toolkits: improving your campaigns.


New Analytics Software Spells it Out for You

When it comes to tracking your campaigns and reacting to signals, software like Google Analytics has boiled the whole process down into a simple feature called Complex Event Processing (listed under Google Analytics as Intelligence Events). With Google Intelligence Events, the software arranges all types of operational intelligence gathered in a list sorted by perceived importance. You still need to tie all the information together, but after making a few deductions and amendments to an existing strategy, the inherent benefits will be obvious.


It won’t tell you how to improve your campaigns, but instead alerts you to identified trends that indicate opportunities as well as dillemas. Google Intelligence Events doesn’t differentiate between internal and external influences, that’s your job. It does however lay the foundation for a software that can, and it’s already currently available in your Google Analytics account by going to Home, and then selecting Intelligence Events in the left hand column.


Event Processing Cycle

Google Intelligence Events takes care of measuring and interpretation in one step.


Google’s Intelligence Events is only a basic use of event processing engines. Some experimental Event Processing engines are already handling tasks that were previously assigned to humans such as driving a car or flying an airplane. Incorporating pattern recognition, relationship detection and modeling information hierarchies, the software takes simple data and infers more complex events that have taken place, and can even be programmed to predict events that are likely to take place in the future. I don’t have to tell you how useful it would be to know about decreases in revenue or traffic before it happens.


The current Intelligence Events feature can look at events that are tagged on a daily, weekly, or monthly basis – as well as a combined overview.


Several Existing Benefits, Unlimited Potential

After applying values to the importance and magnitude of each event, Intelligence Events generates a list of the changes that impact your marketing the most. Sometimes they’re positive, sometimes they’re negative, but they always indicate an underlying cause that should be considered if not addressed.

The reports that can be automatically generated with Complex Event Processing represent the pinnacle of analytics software for a variety of reasons.

  • It’s the most basic output possible


  • It’s a way to manage big data and reduce the likelihood of missing an important change


  • Each result represents an actionable event that can be targeted for amplification or mediation


  • By extrapolating compound events, a user can make several times as many deductions and information requests as a user who is forced to take extra steps to arrive at the level of operational intelligence the software achieves in an instant. This can lead to an exponential increase in results


  • Through inference and identification of the abstract relationships that each data point has to each other, Complex Event Processing can mimic and sometimes replicate human thought patterns


  • With expanding datasets and larger databases being developed all the time, Complex Event Processing will only become more powerful


Using Your Data

There are two primary uses for knowledge you gain from analytics. The first is to inform your marketing strategy, and the second is resource allocation. When you get a signal like a 20 percent increase in traffic over a month, there was either an external or internal stimulus for that increase. If you can identify causation within the corresponding marketing strategy you engaged in during the last month, it’s a very good reason to increase the resources that you dedicate to that specific strategy or technique.


Automatic Intelligence Event Alerts for SMS

Recently I noticed that the bounce rate and time on site for one of our top performing articles on Instagram was lower than the site average. I made a few layout and content edits to the article. Google Intelligence Events has alerted me to the success of my changes.


For example if I ran a website that sold country flags, and during the month of June my revenues i ncreased by 300 percent. By taking the jump in sales at face value and not investigating it, I may be missing out on a much bigger opportunity, and could also risk losing the increased revenues in the future.


Upon further inspection, I see that all of the flags sold through my store saw steady sales, except the Indian flag, which saw a massive jump in sales; upwards of 1700 percent. After checking general traffic statistics, I later find that the site saw only a 10 percent increase in traffic. However, a small pin board style social network based in India became one of my site’s top referrers.


I track the influx in sales to a user pinning a graphic of the Indian flag from our sales page onto their personal account. Depending on what my current marketing is comprised of, I may use resources to establish a presence on that network specifically for an Indian clientele, or I may target Pinterest instead. The question I need to answer is whether a visual marketing tactic sells flags, or if consumers in India are more likely to purchase flags.


The Reality Already Rivals the Possibilities

With Complex Event Processing, it would be possible for my analytics software to simply report a 300 percent increase in sales, and publish a suggestion that I develop my business in India with a report of signals and external data that align with that suggestion.


Google’s Intelligence Events isn’t quite what we consider Complex Event Processing; it’s a valuable tool that uses some preliminary functions of CEP, but doesn’t assume the role of advisor or decision maker except when it comes to ranking the importance of each event. The output you receive is sorted with the actionable data at the top, opposed to the general data you’d see on a traditional analytics report. First it would have identified a large increase in users from India visiting the website, a higher rate of conversion among Indian users, increased sales among Indian users, and a massive influx in traffic from a source that previously referred no traffic. Those alerts would be sorted to the top of my Intelligence Events list. Under those, I may see increases in overall conversions, overall traffic, and overall sales, which of course are by-products of the jump in sales from India. If my data set is extremely large and the increases aren’t as apparent, the new opportunity could have more easily been overlooked.


This leads me to the whole idea of analytics: Identifying potential opportunities and deficiencies to inform your future marketing strategy and areas of concentration.


It works in the opposite direction as well. Instead of positive signals, I could be seeing a large drop in organic search traffic or a drop in Facebook referrals. These signals could be attributed to an external influence, such as a Google algorithm change or a drop in engagement on the Facebook platform. They could also be attributed to an internal influence, such as a decrease in my Facebook ad budget or a trial shift in my posting schedule.


The only difference between an external or internal influence is how I handle the event. If Google changes their algorithm, I may need to react by editing older posts, removing backlinks, or altering my ongoing SEO strategy. If I lose a large source of referral traffic because Facebook is less active, I may need to pursue alternate channels. Where are the Facebook users now? If I see a small increase in Pinterest referral traffic, a network that I dedicate very few resources to, it could be a sign that I need to allocate some of the resources I spend on Facebook towards Pinterest. The great thing about Complex Event Processing is that all these signals will identify themselves, and in the future could even identify both causes and possible routes of action.


Add Google’s Operational Intelligence to Your Insights

Complex Event Processing is the most efficient output of your analytics data. It allows one person to track the most important trends in a very small amount of time. Before recent advancements in data sorting software, it would have taken a larger staff much more time to extract the same insight from web statistics, even for a small operation.


Intelligence Events Alerts for Aug 23

Intelligence Events issues alerts when metrics exceed a predetermined limit. On August 23, I was alerted that both returning visitors and US visitors, both key demographics, responded favorably to ‘Is Your Content Joshua Bell on a Subway’. This finding has led me to alter our content strategy a great deal to better suit our key demographics.


The amount of data that we’re able to accurately measure will only increase over time. In the future this technology will be able to compile complex visitor profiles, draw on a large external database or network of databases, consider semantically similar data and identify solutions.  


Right now with Google Intelligence Events you can quickly identify the most noteworthy events, compare the relationship with other noteworthy events, and more quickly identify causation, sources, problems, and other influencers, both internal and external. By quickly identifying variables that are influencing your business it’s easier to exploit marketing channels that are particularly productive and address channels that seem to be wasting your resources.


Have you been using Google’s Intelligence Events? What types of implementations can you see for Complex Event Processing in the future of digital consumer marketing?


One comment

  1. S_holverson@yahoo.com /

    Excellent article, Justin with invaluable information. Thank you. SonjaSwissLife

Leave a Reply

Your email address will not be published. Required fields are marked *



Read previous post:
Editorial Calendar
Creating Your Social Media Editorial Calendar

The Social Media Editorial Calendar is your blueprint to a successful content marketing/ social media marketing campaign. Do you ever feel...