Can the problems of today’s digital media analytics tools be fixed?
Digital media analytics is metaphorically just a toddler. Even though it can feed itself, it has grown little since birth. There is a lot of growth and change ahead of it and the future is uncertain. Don’t get me wrong, there is much demand for a solution that will track digital media, whether it is social media, your local online newspapers or popular blogs. A clear articulation of this demand is the frenzy of companies that have sprung up that have made digital media monitoring as their primary business objective. Nevertheless, this also points to an interesting business environment in which it is unlikely that one company can satisfy the entire market demand. Thus, to this day, there is no media analytics service provider dominating the market.
The path that digital media monitoring and analytics is taking reminds me of what web analytics went through between the years of 2000 to 2007. During that time, there were many vendors selling data solutions. It was a crowded space with a few big vendors like Coremetrics and WebsideStory in addition to mid-market vendors such as indexTool and ClickTracks. But the market really boomed when Google rolled out Google Analytics allowing anyone and everyone to have access to good quality data for free.
One reason why there are so many social media companies popping up is because the perceived value digital analytics could potentially provide is understood by practically everyone. Digital analytics can assist organizations in crisis and opportunity monitoring, customer engagement, trend forecasting, market research, brand management and lead generation. Individuals, on the other hand, can with a single glance at a dashboard, assess their level of influence, engagement, demographics and related sentiment. Indeed, this is impressive stuff! Nevertheless, digital analytics in this nascent phase does have to overcome a few hurdles. Let’s take a look at those now:
The first thing that comes to mind is the amount of junk information that your average service will net through today’s digital monitoring solutions. A staggering 85% of twitter is useless information and over 50% of the data that you would see in your dashboard would be irrelevant. Unless you are spending hours and even days trying to refine your results, the information you will receive will not be actionable. Imagine if you are dealing with 1500 mentions a day through blogs, newspapers and social media! Having to look through every mention can be a tedious task, and even if you hired an expensive team of analysts to sift through the data, much of it would be tidbits of information without actionable triggers.
The core of digital media monitoring solutions is extremely simple. The tool/service first starts with keyword configuration, some use Boolean logic, and only after the data is extracted from the web based on those keywords does the heavy lifting analytics truly begin. It is at this stage the information is turned in to knowledge. It is only through understanding what the mentions mean, that actions can be taken to capitalize on opportunity or protect against disaster. Here is where the promises exceed the ability to deliver for many companies. Unless you include human analysts in the solution – which is often a top-tier service that costs up to 3x as much as lower tiered services – you are most likely to get a dump of mentions and little else.
Inaccurate sentiment analysis:
Sentiment analysis is probably one of the key selling factors of elite monitoring service in the market today. If users ascertain whether certain trending topics are positive or negative, they can often capitalize on opportunities and protect against threats pre-emptively. That is to say, before threats materialize or opportunities vanish, users can have a form of precognition of events before competitors. Unfortunately, sentiment analysis maybe the most inaccurate component in today’s digital media monitoring packages. It is for this reason, that users are often given the choice to refine their results because it is most likely that only through deep inspection will the sentiment classifications be accurately assessed. There was one company in Australia that claimed 100% accurate sentiment analysis. I soon found out that it is because they have a team of analysts that go through every result and mention and then specify the sentiment. That is simply not a scalable solution and one that is cost prohibitive as well.
Region specific data:
I am sure that anybody who has read this far into the blog would know about Google Alerts. It is adequate free tool, but using it comes with its own share of headaches. If I type in McDonalds for example, chances are that I will be getting results from Canada, the US and who knows where. The return of results that are out of context is a problem that maybe the easiest to tackle relative to the previous problems but few if any have tackled it effectively. Online news sites and popular blogs like the Huffington Post are especially problematic in this regard. Alarmingly, this problem is only exacerbated when one makes the foray into deep social media. With each information source along the spectrum of “traditional” digital media to social, it becomes increasingly difficult to accurately capture information that is in context and relevant.
Time consuming/the human factor:
Last but not least, most of the tools today require the user to sit for hours a day and go through piles of data if they want to ensure they are not missing any opportunities whilst protecting against all threats. In this scenario, the user will need to filter spam, read ALL the data, get rid of irrelevant data, refine the dashboard, and correct the sentiment to get at the big picture.
These five points briefly outline some of the problems existing in today’s offerings. But what is more intriguing is what today’s solution are not offering… There is a deep lack of practical knowledge in this industry sphere. A lot of the progress today is driven by possibility-innovation and not enough emphasis is placed upon customer-driven innovation. There is still a lot of room for increased sophistication, greater intuitiveness and reality-based solutions that are effective whilst not being too costly; we just have to have to raise the bar for everyone else.
Our goal is to create a tool that will not just rid you of all these hurdles to effective digital media monitoring, but that will also raise the bar by offering solutions that the market make have envisioned, but thought only of being a possibility far into the future. We don’t want to overwhelm you with fancy graphs or piles of useless information because we understand that more is not always better. Our goal is to make you spend less time monitoring and analyzing data and more time doing what you do best. We’ll talk more about that in our forthcoming blog posts.