Scout Labs, a Social Conversation Monitoring tool worth a look

I spent some time chatting with the folks at Scout Labs to learn more about their offerings. Jennifer Zeszut, the CEO, was kind enough to take some time to provide me with an overview, answer some of my questions, and provide me with a demo that I will be setting up, using, and sharing my thoughts on in the near future.

Some Background

Scout Labs has been hard at work for the past three years, working with beta customers the last two years, and officially launching early this year.  While Jennifer and I did not go into exact numbers, I do know that they have hundreds of customers, ranging from small restaurants, government agencies, and Enterprise customers like Coke and Disney.  Yes… You read that right…  Businesses can get into the game for as little as $99 per month.  Here are a few thoughts I wanted to share on their pricing:

• A key philosophical belief of Jennifer’s, Scout Labs, and of John Moore’s (yes, me) is that social can only succeed if it becomes part of the corporate culture.  To this end the pricing is for unlimited users.  Yes, anyone, and everyone, can set up an account and utilize the data you are gathering.  This is an important differentiator.
• Your cost will not increase as the level of conversation about your company increases.  With many solutions you incur extra costs if the amount of traffic goes up.  This is not the case with Scout Labs.
• $99 a month is still too high for some businesses.  I would love to see Scout create one more category for the smaller businesses, something that can get them jazzed up about the power of social conversations at a very low entry cost.

How will Scout scale?

The folks at Scout are techies, focused on building great technology to solve real-world business problems.  I like what I have heard about the technology choices made (Linux, great caching, data stored in their own data center and not in the cloud).  While I always build on the Microsoft stack they have gone with a Java-backend and Ruby front-end.    Time will tell how it scales and, if I have time in the future, I’ll dig in deeper on this front as I know I would learn a lot.

 We know you haven’t played with it yet, but tell us more..

Data is updated frequently, with blog posts generally captured within a few minutes and tweets in near real-time.  While the real-time nature of the data gathering is important it is even more important to note that social conversations are displayed by default based upon the “importance” of the conversation.  Essentially, the importance of the conversations is based upon the amount of traffic being generated about that item.  In others words, if you have a United Guitars type of issue bubbling up, it will quickly rise to the top of your view, enabling you to react to this item before it blows up in your face. 

Scout Labs appears to be doing some good work determining sentiment.  While it is presented simply as generally positive, generally negative, or generally neutral, there are a couple of key points:

• The sentiment of the conversation is relative to the entity being monitored.  In other words, if a post was extremely positive about Jennifer Zeszut by negative about John Moore that would be reflected as such.  In other words, conversations are neither positive or negative on their own, they are only positive/negative in relationship to the entity being monitored.  That simple statement is very, very powerful.
• If the solution mis-diagnoses a post users can easily correct the sentiment of a post and all other users in the company get the corrected information.  This information is also fed back to the engineers in the back room to further refine their algorithms.

A ‘cool” feature is called Quotes.  Quotes mine your data  to find conversation snippets that are categorized between categories such as Love, Hate, Wish, Compare, Recommend, issues, Caveat.. In the samples that I saw the quotes provided some good insights of how people are feeling about the terms being monitored, something richer than the raw numbers alone.

The areas where I see initial room for growth will need to be addresses soon:

• There is no ticket system of any sort today.  It is being built and is due out in the near future.  Social Conversation Monitoring solutions must have a basic ticketing system if they are going to work for mid-sized through enterprise companies.  If you don’t have this you are not really in the game in my opinion.
• Reports are comprehensive but rigid giving you all of the data without the ability to either see reduced subsets of information or automatically route certain sections to certain users.  For example, you might only want to send negative items to your PR and Customer Service teams while product feedback might only go to your engineering team.

That’s it for now.  I’ll share more as I set it up and play with it.  Stay tuned.

John