Artificial Intelligence applications for social data mining in Marketing

With ready and affordable availability of commodity hardware, the use cases of Social data mining within Marketing have grown way beyond the original applications of sentiment analysis and market research. Freely available natural language processing APIs, artificial intelligence capabilities, massively scalable servers, and dirt cheap storage capacity have all contributed to an incredible sophistication in how companies are using social media data for building competitive advantage.

In this post, we identify three use cases for social data mining with particular reference to marketing. All three applications highlight how companies can combine artificial intelligence capabilities with social media data to build innovative marketing applications that can increase both the efficiency and effectiveness of marketing function as a whole.

1-Chatbots and ‘Conversational commerce’

Chatbots are effectively messaging systems that can be installed on customer’s site or on other third party platforms (e.g. Facebook). The bots are equipped with artificial intelligence capabilities that allow them to identify meaning in conversation. Technically, this involves passing the conversation text to an NLP algorithm which then identifies data that can be used to look up a database and construct a response on the fly. In another category of bots, the conversation can also be shaped using a rules based engine that decides the text to be output based on past visitor data (e.g. number of visits, pages seen, engagement index, transaction status etc.)

The applications of bots in marketing are literally infinite. Think automatically pointing a visitor to a specific link based on his past browsing behavior collated from a tool like Segment.com or a DMP? Or offering an instant discount based on the user estimated propensity to buy? How about implementing an advanced bot that can convince trial users to get into paid membership or prevent account cancellations from users looking to walk out? Sounds sci-fi but certainly possible by creating overlays between anonymous click stream data and any available social data points (e.g. if the user has used a social login to register for the site).

2-Social ads

Social ads are any kind of paid content displayed to users of a social network. They can be simple banners, images, links in a Tweet, or even audio advertising (such as is the case with podcasts). The great thing about social ads is that they can be tailored and targeted to very specific customer segments which in turn increases the chances of conversion. For example, Facebook allows advertisers to create custom audience using first party data and then run hyper-targeted ads for these. It simply means that rather than use its own intelligence on which user to show which ad, Facebook will show the ads to members that meet an advertiser’s audience criteria. By combining social data (accessible if the user can be made to login using social login functionality) with other organic data points, advertisers can apply artificial intelligence capabilities to create a range of segments and all of which can be setup on a platform like Facebook to generate significant lifts in campaign performance.

3-Social CRM

Social customer relationship management (CRM) is what we get when the capabilities of social media management are combined with traditional CRM software. Instead of simply collecting customer data points such as emails, phone numbers, company information and job titles, social CRM allows companies to reach out and engage prospects and customers in much more meaningful ways through overlay of social behavior data, professional and demographic information. Imagine capturing the email of a visitor through social login and then collating his social interaction data from across platforms over a period of time. Artificial intelligence can then be applied to this raw data overlay to identify a number of signals including product propensities, feature preferences, likelihood to purchase, and purchase lead times. These signals can then be used to bucket prospects into hot and warm leads. For example, a user posting a question about products in a particular category. Or the same user visiting a product reviews section on a brand’s Facebook page. If the same visitor is also part of a leads database with known job title, company name then it makes sense to push this lead to Sales even though the visitor may have never initiated any direct interaction.

Conclusion

Progressive Marketing organizations earn their distinction by staying ahead of the curve when it comes to applying emerging technologies for solving specific business challenges. By building in artificial intelligence capabilities into their data processing technology mix, companies can build advanced customer interaction capabilities that propel them into a new era of marketing sophistication and customer profitability.

A ‘use-case’ themed article that outlines potential use cases for artificial intelligence in marketing. Branded content that targets top-of-the-line prospects looking for increasing the sophistication with which they can use data.

Technical data Analysts and Architects looking for advanced use cases to improve the insights that they get from social data.

A leading product company in social-tech space and with an enterprise-grade product focused purely on providing raw topic based data from popular social networks.