Sentiment Analysis for Competitive Analysis

It’s important for any business to keep an eye on the competition.  These days a good way of doing this is to monitor your competitors’ social media.  This can tell you many things, among them:

• The share of voice that each of your competitors has on social media
• What people are saying about your competitors, particularly the things they do well and those they do poorly
• Who the influencers are for your brand and your competitors’.

Social media is also rich in insights regarding market preferences so you can better understand what your customer base wants. Social media can help companies understand what their strengths and weaknesses are, what differentiates them from others, and what new industry trends are shaping up. For instance, here are some tweets about airlines, but comparable ones occur for all industries:

• Pricing (“It’s hard to remain loyal to @<AIRLINE A> when fares are $100 or more higher than competitors for the same flight. Guess I’ll be flying @<AIRLINE B>.”).
• Flight booking (“Why is it so hard to book a flight with @<AIRLINE B>?””)
• Preference for certain flights (“I’m never flying <AIRLINE B> for a regional flight.”)

You can see these tweets show <AIRLINE B> is doing pretty well on pricing, but it’s losing out on flight booking and the quality of regional flights.

So you might think a few tweets don’t tell you anything very representative, but imagine that you can analyze a massively large volume of such tweets.  It would be impossible to do this analysis manually: it’s just not feasible or cost-effective.  There is no simple automated solution either because natural language is rich and creative and there are myriad ways of expressing opinions, likes, and dislikes.  Here is where an AI-powered solution comes in handy.

Sentiment Analysis: Getting a Better Handle on your Competition

Sentiment Analysis is a technology that analyzes natural language to identify positive and negative sentiments.  It’s especially applicable to sources containing a large amount of opinion, such as social media, blogs, reviews, and forum posts.

At a basic level, Sentiment Analysis identifies positive and negative language and sometimes associates it with items like the names of companies mentioned in the same document or in proximity. By seeing how often competitors are being mentioned close to where the sentiment is expressed, you’ll get a clearer view of your competitors’ market share. You can also see how your brand is doing. Additionally, a mention count will give you a good way of measuring your growth in social media buzz.

At an advanced level, using what is called Entity- and Aspect-based Sentiment Analysis, Sentiment Analysis identifies the specific things that the sentiments are referring to. For instance, travelers may love an airline’s frequent flyer program but complain about high prices or frequent late departures.  Given insights about a competitor’s shortfalls, you could design a new campaign highlighting how your company excels in those areas.

Sentiment Analysis can also handle Big Data. The more data you can analyze, the more confidence you can have in the insights.  With advanced Sentiment Analysis, it is feasible to process massive amounts of data concerning your competition in real time, which will make the sentiment information highly valuable.

Sentiment Analysis normalizes the sentiments it extracts. Sentiments like “I hate X” and “I despise X” will receive the same output format.  This makes it possible to aggregate the sentiment over a very large collection of content. The most popular way of viewing this sentiment is via a dashboard interface that presents multiple views of the sentiment. It uses different kinds of graphs and charts. Sentiment data can be analyzed in different ways, for instance by displaying the positive and negative sentiments for different aspects or categories (e.g., quality of cabin service or on-time arrivals). The user can drill down to the source text to see what the actual sentiment language is.  Sentiment can also be tracked over time based on the time stamps of the source.

In sum, advanced Sentiment Analysis can allow you to get a much clearer view of your competition – their strengths and weaknesses, all via their social media footprint.