Sunday, June 12, 2011

The secret weapon of marketing war (2/3)

Battlefield: Slovakia

Now the reason the enlightened prince and the wise general conquer the enemy whenever they move and their achievements surpass those of ordinary men is foreknowledge.

Note: If you have for whatever reason missed introductory part, please follow this link first. 

Before a military battle, the battlefield  is usually mapped and studied in great details. Marketing battles are bit different in that they take place in the minds of consumers, so on a terrain for which it is hard to retrieve electronic map in few seconds from popular search provider and to make it even worse – landscape can change rapidly like dunes after a sandstorm. The question is, how can such place be mapped or studied?

The answer is s that it is possible to construct such a map by assembling bits and pieces of data from reliable sources. Valid source can vary from any user forum, social network to specialized portal. In this demo, articles will be used from top 5 portals specializing in mobile phone market news coverage. Again only (mostly incomplete) information published via RSS channels will be subject of the analysis, to conform with the IP rules. Despite considered as more direct and therefore better source for market analysis – user forums or social networks will not be used simply because none (in Slovakia) is publishing something like RSS/Atom feeds. Of course, at dawn of the war of life or death for your company, this is not something that would stop you, but here we can accept that consumer's mindset is strongly influenced by what he/she reads or hears from the media. Our example consists of 532 articles harvested from the following sources:


Of course, simple aggregation of news articles alone is useless – but this is exactly the point where secret weapon of our choice - text analytics (TA) –  can save the day. As we have seen in the previous demonstration, TA can be used to extract important domain topics and their relationships. So as a first step, we will run all articles through the analytic engine and from the results we can learn, that there are three mobile network operators active in Slovakia – Orange, T-mobile and Telefonica O2 (green frames on the picture #1) and also three major mobile phone platforms  - Symbian, Windows Phone, Android (red frames on the same picture).  Of course there might be also other platforms, but these were not discovered during the first run because of the domination of those mentioned above. The biggest cluster (291 articles – blue frame) also tells us that we have a category of companies in the data set (and strong presence of “operators”) with a close relation to "mobile phone(mobilny telefon)". So the first results are quite promising because we can see that data collection is full of information about the key players operating on the analyzed market.

Picture #1: Result of first run

To process this result, we will create an element representing mobile phone/smartphone category and as we go we will add elements representing particular devices. We will also create another element representing operators and will add elements representing particular companies. After few more cycles, our virtual battlefield map will consist of operators (picture #2): 

Picture #2: Operators
as part of larger group of companies (picture #3) mentioned in the data. On the next picture we can see, that the original list of operators can be enriched with  mobile OS/platform vendors and device vendors:

Picture #3: Companies

and to make it complete, we have a mobile OS/platform category and mobile phone/smartphone category with respective devices.

Picutre #4: Devices

As we move on, deeper research reveals additional topics - additions to previously defined categories of platforms, devices and companies. For example on the next picture (picture #5) we can see emergence of apple/iphone company/device:

Picture #5: Emergence of apple/iphone from the data

Another interesting characteristics that can be read from previous picture, but is even explained in more details in the next one, is that information written about mobile phones is closely related to specific device features like display, camera or processor (picture #6). 

Picture #6: Display/camera features of mobile phones frequently mentioned in data collection

After few more cycles, we can stop and see how a virtual topic map of our battlefield including companies (vendors, operators) devices and key features can be tested by performing a search for example for HTC company (picture 7). On the picture we can see related products and devices:

Picture #7: HTC search showing related topics

It is very important to mention that numbers in brackets after the element names represent number of articles matched under each topic. Numbers are absolute, not relative to our current search.

So far we have seen how text mining/analytics can be used to create a virtual model of some specialized market by aggregation and analysis of generally available information on the net. In the next part, we will use it for deeper analysis and will discuss the findings in the context of marketing warfare strategies. For example let's look back on the updated picture (picture #8) of companies:


Picture #8: Updated list of key players related to mobile phone market in Slovakia

It follows from the picture that device vendor Nokia is (still) clear leader (when considering frequency in the news alone) in the group of device vendors (HTC, Sony Ericsson, Apple, Samsung). It is also clear, that player number #2 (HTC) is catching up. Using categorization introduced by the Marketing Warfare book, we can view Nokia as the player still dominating the market, HTC and Sony Ericsson as the players with increased market share and the rest (Apple, Samsung) of companies can be labeled as "profitable survivors". Such categorization of companies can help us to predict which company is most likely to adopt which strategy/style when fighting on the battlefield. Leaders are advised to choose defensive styles, strong pursuers will most likely adopt open offensive moves and the rest of the market will most probably try to compete in a less aggressive style. To prove that our strategy (market analysis via news aggregation and mining) so far meets the reality, you can check article on Mobilmania, that presents official data from T-mobile with similar findings to what we have discovered so far.  
Posted by at
Labels: , , , ,

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)