Interview

Tawlk About Social Media Monitoring (Interview)


Brands no longer have a choice on whether or not to ‘get social.’ The cold, hard truth is that all brands are now social as conversations are already taking place whether or not you chose to engage. The first frontier on the path of engagement is the emerging field of social media monitoring. Tawlk, a social search start-up, has stepped forward with a new platform and data-collection model. I was fortunate to get a few moments with Lance Vick, Tawlk’s Co-Founder and Lead Engineer, to learn more. 

This blog sits at the intersection of branding and social media. Tawlk provides brands with a tremendous opportunity to monitor sentiment and conversations around their brands on social media. Can you give us the elevator pitch?

Tawlk is a search engine that allows you to put in any keyword, and see where that keyword is being mentioned across the social web in real time. Additionally you receive important stats to get a feel for the overall sentiment, popularity, and amplification of a topic. Also for topics that reach a certain popularity threshold, an overall 0–100 “Tawlk score” is provided that acts as a kind of social credit score.

There are many social media measurement tools out there. What makes Tawlk different?

Some key things we do differently from most/all of the competitors in the social data analysis space include:

  • We offer free public search and analytics with no sign-up or credit card required.
  • Our “Comparative Wall” (below) shows how competing brands rank against one another in respect to their overall social sentiment, influence and popularity,
  • Our data and statistics are updated and made visible in real-time.
  • Our unique scoring algorithm factors in sentiment, amplification and popularity of public posts about a topic and returns a single number score for easy and quick interpretations.
  • Key usability features in our interface allow users to view pictures and videos without having to leave their search results screen.
  • We have open sourced our unique sentiment analysis and data collection processes, giving us access to accountability and improvements from a wider range of engineers than any one company could ever hire. This also helps us appeal to, and be more easily trusted by, software engineers who are often the decision-makers on how technology is developed in companies. This strategy has proven successful in other industries by brands like SugarCRM, WordPress, Android, and Red Hat.
  • Our overall platform is being developed in an API-driven configuration, meaning that future subscribers will be able to develop their own unique applications against our engine, much the same way one currently writes applications against Facebook and Twitter platforms.

Your hybrid collection model that powers Tawlk is pretty innovative. Can you explain how it works and tell us about your upcoming presentation in Dublin?

Most companies that do social data collection at scale, do so by utilizing many servers (often hundreds) to go out to many social networks and watch a wide range of topics. This expense is often passed on to customers via hundreds or even thousands of dollars per month in subscription fees. The primary reason for this is many services enforce strict limits on the amount of data a single server is allowed to consume. Being a startup we can’t afford such luxuries, and we also wanted core functionality that would scale economically enough that we could provide it to the public for free.

To solve this problem, we opted to push the majority of the data collection work to the actual web browsers of those using our service. This means instead of you doing a search, and waiting on our servers to go get the data from all the various social networks for you, your browser is silently given code that enables it to gather the data on its own. The actual paper that being published by the AAAI at the event is 4 double column pages, so this is obviously a very top-level description. Source code to support these collection methods is also available on our Github account.

Who do you see as Tawlk’s ideal user/customer?

Tawlk is for anyone who wants to better understand how the public at large feels about any given topic. This could range from academic researchers wanting to understand the social impact of current events, to brand managers wanting to see how they are matching up against their competition. This sort of data can also be very valuable to consumers who simply want help in making buying decisions, such as choosing a cell phone carrier.

How are you planning to monetize this?

We are currently developing systems to allow us to begin offering extended functionality to users for a modest monthly subscription. These tools will allow researchers and brand managers methods to monitor and understand topics of interest not just in real time, but also historically with intuitive graphing tools and alert systems. Example use cases include someone who wants to see how the sentiment of a brand changed after a marketing campaign, or to see who the biggest influencers were at any point in time.

The other brand that I’m curious about is your brand. What can you tell us about the ethos of the Tawlk brand that you hope to create?

We intend to provide intuitive tools to help anyone better understand and interact with social media data. This includes making free tools available to the general public wherever economically possible, and providing low cost tools for heavy use cases. We also are passionate about transparency, collaboration, and open source development. For example, we were unable to find any open source sentiment analysis solutions with any reasonable accuracy, so we built and open sourced our own (called synt).

We believe in the “open source (almost) everything” philosophy as defined by Github founder Tom Preston-Werner.

What’s next for you in the development and launch process?

The next phase in development will be systems that allow users to authenticate with their various social networks, so they can interact with data directly. Without leaving the interface, users will be able to do things like retweet, comment, upvote, like, etc. as appropriate. We are also working on features that allow one to interact with data from their own personal networks, in addition to the public data as seen today.

Exciting stuff! Thanks for sharing, Lance (Above-left with Co-Founder/Creative Director Hassan Seguias). Remember, you can give Tawlk a try right now and see what all of the buzz is about.

What tools are you using to monitor sentiment and conversations around your brand?

All Tawlk screenshots are from the upcoming release targeted for Q3 of 2012.