Call (+1) 650-212-1212

Blog

Prescriptive Analytics and Social Network Marketing

Sean Salleh 04 Jul 2013 Modeling methods

Are prescriptive analytics ‘the Final Frontier’ of analytics? This apparently is how market research firm Gartner sees this area, melding this third phase of business analytics with images of Star Trek. Prescriptive analytics are now increasingly presented as the evolution from predictive analytics (what will happen), which in turn is the extension from descriptive analytics (what happened so far). It’s also being mashed up with Big Data and being ascribed capabilities that sometimes almost defy belief: for instance, the power to pre-empt customer complaints and reverse negative brand image by scouring social media for relevant data. Are these descriptions of prescriptive analytics fiction or reality?

Social marketing and social business Image source: flickr

Traveling at Hype Speed

Continuing our space saga analogy, it seems like prescriptive analytics is currently benefiting from positive public relations. While this new spate of interest may herald constructive developments, it has to be said that prescriptive analytics (PA) has been with us for a while. Also known as decision analytics, PA has been the mainstay of companies such as Amazon and their automated sales recommendations to customers based on (you guessed) what the customer bought last and what new, hot products have recently been released. Amazon being a big data universe all by itself has lots of data to work with, and – an important point – the power to structure how that data is generated.

Amazon prescriptive analytics at work Image source: flickr

‘It’s data, Jim, but not as we know it’

Too right, Mr. Spock. There’s numerical big data, estimated to account for about 20% of all the data out there. Then there’s all the rest – unstructured text, images, videos and so on. In social media, customers are not always obliging enough to express themselves by ticking boxes, answering multiple choice questions or entering numbers that fall between pre-defined bounds. However, they are often enthusiastic about exchanging unstructured text strings (opinions, comments) with their social media friends.

‘Ye canna’ change the laws of physics!’

And yet, Scotty, you of all people should know: Star Trek did this all the time, with warp factor speed, tele-transporters and much more. So in that case, we should at least be able to twist unstructured data around, pre-digest it, sort it and feed it into a business analytics model that then ranks the relative desirability of different actions. Some vendors are now offering solutions to scan social networks on behalf of firms and their brands to pick up relevant feedback that can then be used to assess the current situation (descriptive), make projections into the future (predictive) and formulate recommendations for immediate action (prescriptive or decision analytics).

Data cleansing process Image source: blog.vovici.com

Just Plug It in Here

All the component technologies exist. There is no reason why even something currently as vague as ‘social network marketing’ should not benefit from more precision and (gasp!) paths to real results, like sales, instead of just ‘Likes’, ‘Friends’ and ‘Followers’. Marketers will better know not only how much churn they’ve had and why, and much they can expect in the next months, but also the actions to forestall that churn. The challenge will be to make a process that works end to end to pull in data from the likes of social networks, clean it, transform it, analyze it and action it in a way that is efficient and effective.

If you’d like to know how Analytica, the modeling software from Lumina, can help you with prescriptive and decision analytics, then try a thirty day free evaluation of Analytica to see what it can do for you.

» Back

Sean Salleh

Sean Salleh is a data scientist with experience in guiding marketing strategy from building marketing mix models, forecasting models, scenario planning models, and algorithms. He is passionate about consumer technologies and resource management. He has master's degrees in Operations Research from University of California Irvine and Mathematics from Northeastern University.

Leave a Comment