When it comes to digital or social media analytics, companies often take two distinct approaches: quantitative analytics and qualitative analytics. Various analytics software products provide quantitative analysis and these are in abundance in the market. Qualitative analytics is expensive and rarely standard. Each in isolation isn’t sufficient to solve business problems. Hence a blended approach is needed which can be called “consulting with analytics”. (I tried blending the two to form something like “conalytics” but that didn’t sound right.)
Quantitative analytics products: These are based on the approach of automating the analytics process. Typically the assumptions here are: the data is too big to analyze, the data discovery, extraction and indexing process is tedious, generating an output will take time etc. These are all realistic challenges which companies solve using automated tools and technologies. The analytics products here may include counting software (those which count keywords and phrases, links to your blogs, web traffics, no of clicks, click-through rates, conversions, likes, dislikes, views, etc.).
Although there can be advance levels of counting such as establishing funnels, understanding responses to various web-page designs, comparing campaigns on various platforms and what not, counting in itself is not analytics. These software usually do insufficient justice to the qualitative aspect of data or user behavior. Advance textual analytics software or natural language process systems also address those issues based on pre-developed taxonomies and ontologies but their accuracy is a significant obstacle and however advanced they claim to be, these rarely match human intuition.
Qualitative analytics services: What machines cannot do, human beings have to do. No machine can understand and interpret data better than a trained human mind. Such an offering can be based on number of hours billed, number of resources employed, their skill set, location of work etc. Human analysts read through the data, qualitatively interpret and analyze it, categorize it, tag it and consolidate findings. Such approach is best to determine qualitative analytics such as conversation themes, product preferences, reasons for liking and disliking, trends, opinions, recommendations, views on features and comparative analytics etc. The biggest issue here is scalability. Although highly accurate and of a high quality, such analytics output cannot be scaled.
Hence the blended approach: quantitative and qualitative analytics together to produce a coherent meaningful analytics output.
Even with this there is no guarantee that actionable insights will be evident from the analytics. For effective problem solving and providing meaningful actionable insights, far more is needed; correct diagnosis of the problem, domain expertise, learning from historic examples, experience in solution design, understanding of the macro environment and ability to mobilize resources and implement solutions. In short, consulting skills are needed. Those, only human being can do.
If you want to solve business problems based on analytics, don’t just buy software or hire or rent people. Find a vendor that does it all; provides analytics software, provides human analytics services and provides actionable solutions.