Question: How many of you social media practitioners and webmasters look and analyse your platform’s statistics on a regular basis?
I am guessing the answer would be not very many.
In my own experience, I think that many practitioners are focused on the concept, ideas and execution of many online projects, but few look really deep and hard into the desired outcomes of such projects and the resulting performance of the project.
I know what you’re thinking:
“OMG I’m so busy already trying to get the website up and running (not to mention this is the 3rd iteration after my boss’s boss asked for changes), and you want me to spend time on statistics?”
Well, the answer is yes. Resoundingly yes.
I think that the problem that most people face is that the tools are usually there (be it Facebook Insights, Google Analytics or any other reporting platform you own), but the complexity and features of such platforms usually boggles the mind.
In my mind, any form of statistics/analytics should be seen in the same way that intelligence (in the military context) is seen.
Direction
Based on the mission, the Commander (the boss in a business context) determines the intelligence/information that is essential that would enable him/her to develop plans and make decisions. The Commander directs the collection effort to find essential elements of information.
In the business context, what would bosses need to know? Some ideas:
- Awareness — how aware is the target audience of your brand?
- Opportunities — what new opportunities are there? new markets, trends, customer segments
- Effectiveness — how effective are our current marketing/publicity efforts?
- Engagement — is our audience being effectively engaged by our brand?
- Competitor Analysis — how is our competitor doing?
Based on your business objectives, you could then develop a collection plan which details what information you require and develop the metrics and indicators which would generate that sort of information.
Collection
The collection phase is pretty straightforward. You look at the plan and the relevant collectors (military: scouts, unmanned aerial vehicles; business: Facebook Insights, Google Analytics, online sentiment analysis tools) pump data back to your Headquarters.
In this phase, I think the important thing to note is that there is a lot of data that can be collected through various means (unlike military intelligence, where data is often quite hard to get). This may lead to data overload, and your human collectors need to be very focused on the key information that you require, and not be distracted by the mounds of data that is coming in.
Because of this problem, I believe that it’s important then, to create custom reports/dashboards/alerts, so you remain focused on a particular “mission”. Avinash Kaushik has a great post on how to produce actionable insights through custom reports.
Processing
In a military sense, there are four sub-sets under processing, namely:
- Recording
- Evaluation
- Fusion
- Interpretation
Recording, simply put, is reduction of information into key statements or graphics and arrangement of the information into categories. Most analytics software would probably take care of this for you, but I would spend time identifying the key metrics (e.g., hit rates, visitor demographics) that you want to capture and ensure that they are captured properly.
Evaluation is the examination of information to determine its value, as well as its accuracy. To be honest, at first I could not determine the business equivalent for this sub-step, because there is some overlap with the “determination of value” and the “setting of collection targets” in the Direction phase.
However, thinking a bit deeper, I realised that there were still applications to this sub-step, especially when thinking about verification of the accuracy of the information being received from external sources.
Most organisations engage a media buying agency or work directly with a media owner for advertising, and the agency/owner would provide the data on the success of the advertising campaign. I’ve always felt that there was a conflict of interest when a media buying agency or media owner provided the statistics of how well the ads performed – wouldn’t they rig the results? They could adjust the results of the ads performance to suit their business objectives (“these ads performed well/badly, so why don’t we buy more/buy a more expensive option”).
Note that the data received is still important, but because it comes from a source other than your own, the reliability rating would go down slightly.
What could help solve the reliability problem would be the Fusion step. This stage is to provide a current and fused picture of the situation by merging all the data available.
Taking the example of ad campaign performance, let’s say you want to determine how many users clicked thru on your ad. The data sources available could be:
- Report by Media Buying Agency/Media Owner
- Stats from URL Shortener (e.g., Bit.ly)
- Website statistics’ campaign tracking (like campaign tracking in Google Analytics)
- Surveys (Yes! Even though we’re talking about online performance, why not do some surveys of your target audience?)
By combining the data from above sources, you can get a reasonably accurate picture of how your ad campaign is working.
Interpretation is the process of looking at the fused information against information that was previously known and trying to develop patterns and deductions.
An analyst could probably look at the data on past ad campaigns and compare it against the initial results from the current campaign, to see if the data conforms to patterns or is going against the norm. He can then add contextual information (e.g., whether the time of year is better for sales – like just before Christmas) to form certain deductions.
The analyst could then make some postulations on how he expects the target audience to behave and the designers/developers could rework the designs in response to those postulations.
Dissemination
Dissemination is probably quite straightforward, but to challenge yourself, you could think about having different customised reports for the different stakeholders in your organisation (will elaborate further).
At the broadest level, you could produce an overview of the performance of your online and social media activities, similar to how the US Army does its weekly report.
In addition, because your organisation probably has various social media and web platforms, you might have to develop your own recording charts, something similar to what the US Army has developed for their Online and Social Media Weekly Report.
At the stakeholder level, you’ve got to figure out what they would be interested in. For example, if you had a team producing marketing videos on your YouTube channel, they’d be interested in the viewership and also the audience attention. The customer service and feedback team would be interested in the number of feedback queries submitted through your website and social media channels. The sales team would be interested in how ads are driving sales.
Hence, each organisation could work with the various stakeholders to develop customised reports and dashboards, and the important thing would be the generation of actionable insights. For the video team, they’d want to know what kinds of video styles are more engaging so as to increase audience attention.
Conclusion
What I’ve tried to do here was draw parallels between online and social media analytics and the military process for intelligence gathering. While the similiarities are a bit strained at times, I think the principles are generally similar.
What you can do as a practitioner is look at your business objectives, your collection means, your resources and modify this process to fit what you need. Perhaps you don’t need something very rigorous, and you only have a small budget for analytics – cut down the steps.
The key thing is that you never ever try to fight a war in an information vacuum, so why should business be any different?
It is said that if you know your enemies and know yourself, you will not be imperiled in a hundred battles
- Sun Tzu’s “The Art of War”