The key piece of artificial intelligence that we use in Selasdia is called intention analysis.
It is the ability to look at a tweet and determine if it is a complaint driven by a problem, or an inquiry driven by a need for information, or an intent to purchase driven by a pressing need, or just a piece of information that we can safely ignore.
The ability to differentiate intentional tweets from informational tweets gives us interactive superpowers.
This is because intentional tweets normally require a response. These are tweets that are sent with the sole expectation that someone will do something in response to the tweet. They are intended to change the state of the world.
Informational tweets are fundamentally different because they are not meant to change the state of the world. They are meant to reflect the state of the world, or inform people about the state of the world. They are not meant to elicit a reply or cause someone to take action.
So, by ignoring the informational tweets and focusing on the intentional tweets, B2B firms can be far more customer oriented and can sell better, market better and serve customers better.
The superpower of intention analysis allows Selasdia to scan through the hundreds of thousands of tweets that go out each day and only identify those that are useful for marketing, sales and service.
Here are three examples that illustrate how intention analysis allows users to react faster than their competitors:
a) Tweets that complain about products.
Here are some real-life complaints picked up within the last 24 hours:
@m______ is your site down?
@m______ Is your server down?? Nothing is loading. Really hope I didn’t lose all that work..
@H______ is signals down? not getting notifications, this kills me!
These tweets alerted us to problems in competitors’ products long before their customer service teams could respond.
The customer service teams of the firms in question in all probability had to manually sift through tens of thousands of messages about their products, making it difficult if not impossible for them to spot complaints in spite of an enormous expenditure of manual labour.
This is an illustration of the customer service use case for intention analysis.
b) Tweets that inquire about B2B products.
Sometimes, customers request information about products. For example, the following tweet is an inquiry about marketing automation:
Where’s the expectation of privacy in @H____ Signals? Should you tell your recipients you’re using it?
The quick detection of the inquiry enabled the insertion of a response before the vendor at whom the inquiry was directed did, illustrating the marketing use case.
c) Tweets that reveal a need for B2B products.
Sometimes, customers reveal that they’re evaluating products. For example, the following tweet revealed a need and a possible ongoing evaluation of competing products and allowed us to pitch our product to the prospect:
Anyone have any live experience with using P_____? Has anyone done the comparison between P_____ and M______?
It is easy to see that the firm that the person who tweeted this works for is looking to buy marketing automation. So this would be the right time to approach this firm about a sale.
This is an illustration of the sales use case.
The following whitepapers are about intention analysis:
1. On gaping holes in text analytics and how to comprehensively convert unstructured text into a structured form: http://aiaioo.com/whitepapers/text_analytics_360.pdf
2. Applications of intention analysis to sales and marketing (analytics and improved targeting): http://aiaioo.com/whitepapers/intention_analysis_use_cases.pdf
3. Applications of intention analysis to making call centers more efficient:
Our research on intention analysis was published in a demo paper at the Coling conference in 2012.
You can view a demonstration of intention analysis at our research lab’s (Aiaioo Labs) demo website:
Share This Post With Your Friends …
How Separating Intentional Tweets from Informational Tweets Helps You Sell [Tweet This]