Khoros Atlas Logo

What does Lithium + Klout Means to Me (Part 2—Future)

Lithium Alumni (Retired) Lithium Alumni (Retired)
Lithium Alumni (Retired)

Last time, I explained why I think acquiring Klout is such a smart move from a data science perspective. And it’s primarily because of the powerful enablers that come with this deal:

  1. The consumers who are willing to opt-in their data
  2. An agile, flexible, and scalable big data infrastructure (both hardware and software)
  3. The data science talent who can use the infrastructure to derive insights from the consumer data

 

I also expressed my personal excitement about this acquisition, but I don’t think I did justice there, because I haven’t had the chance to talk about how we plan to leverage these enablers. Today, I’m going to put on my product hat and try to give you a flavor of the products we could build with Lithium + Klout. Keep in mind that this is not a roadmap, but they are ideas we’re exploring. Since the possibilities are endless, I will describe 4 in total—2 that are immediate on the top of my mind, and 2 more that is further out in the future.

 

Klout+Lithium=shared value2.png

 

1. Scores for Everything

Klout’s algorithm was originally developed to score people’s influence on social media. However, with Klout’s big data infrastructure, we can score other quantities of interest—engagement, interest, expertise, popularity, reputation, contextual influence, etc. Moreover, we can score these quantities for other entities in addition to people—products and services, brands, even places with mobile. These entities can come from a product catalog, extracted from entity extraction algorithms from conversations, or manually entered for tracking.

 

It boils down to developing new algorithms that use different signals to scoring these entities, since the big data infrastructure is already in place ready to crunch the numbers. With the talented data science team at Klout, I’m confident that we can develop algorithms to score anything. And with the rigorous validation framework I developed at Lithium, we’ll make sure that the scores are predictive and mean something substantial.

 

In addition to the signals that Klout is currently using to score their users, now we can augment them with social signals from the community. This provides more context, specificity, and interpretability to all the new scores to come. These signals can be derived explicitly from ratings and reviews of product, people, and contents they create, or implicitly from the conversations among members of a community.

 

2. Gamification with Portable Reputation for All

gamification brands+consumers.pngLithium has a long history of success applying gamification within the community context. With Klout’s cross network profile, we can level up the game even further. This means reputation earned within a community is no longer limited to that community. It becomes portable and visible to the wider social networks, which is more relevant, personal and valuable to the users. This is how reputation operates in the physical world.

 

Since we can score different entities for different attributes, what’s more exciting is that products, brands, and even the place may have portable reputations just as people do. For example, we can answer questions like, which speaker has the best sound quality, and which is most desirable? Which brand has the best warranty, and which has the best customer service? Which park is most popular during Easter, and which during the week of your birthday? These new scores are all context specific. That means the same entity (whether it’s a person or speaker) may have a different score depending on the context.

 

I must emphasize that the reputation score of a brand (or product) is not the same thing as treating the social profile of a brand (or a product) as a person—which is currently done. They are scored with a completely different algorithm using different social signals from the community and the broader social networks. Such brand reputation scores provide a transparent benchmark for brands to see where their customer centric efforts stand among their competitors. The implication of this is profound. It means not only can we gamify consumer behaviors; we can gamify brands behaviors. And the consumers are the ones deciding what brand behaviors are desirable, what needs improvement and what’s acceptable.

 

3. Analytics as a Service

Once we’ve developed enough algorithms for processing different types of signal and scoring different kinds of entities, these calculations could be offered as a service.

 

In the data as a service model, you specify precisely the data you want from a set of available searches and filters, and the service providers will provide that data to you. Likewise, in analytics as a service, you specify precisely what processing and calculation you want to perform on the data from a set of available composable analytics modules, and the service provider will do the heavy number crunching and deliver the results.

 

data cloud.pngTo understand analytics as a service, we’ll consider an illustrative example in influence scoring. For example, you may have the communication data from your internal enterprise social network, and you’d like to find the top 10 influencers from that data set. Since this data is not public, Klout couldn’t have access to it. But if you provide the data (in a pre-specified input format) and specify the computation you want performed (i.e. score influence), then analytics as a service enables you to use the existing infrastructure and algorithm at Klout to identify the top 10 influencers within your enterprise social network.

 

Since these analytics modules are designed to be composable, you can chain them up by specifying a sequence of desired computations. For example, after you scored people’s influence to find the top 10 influencers, you can feed the results into an expertise identification module that tells you the expertise of these 10 influencers. As we perfect more algorithms, they will be added to the set of available analytics modules. So the number of new analytics you can create by chaining up pre-existing modules is virtually infinite. With enough analytics modules, analytics as a service provides brands the ability to perform arbitrary data processing tasks to address unique business inquiries.

 

4. Shared Value Network for Brands and Consumers

Since we score both the consumers as well as the brands (along with its products/services) on public social networks, we achieve full transparency on both ends. Brands can see consumers’ digital reputation, but consumers can also line up the brands (or products) and decide which could best meet their needs with trusted information from their peers

 

Consumers who are not shy to show their passion for a brand will be recognized and rewarded. They may get a Klout perk or a badge initially—which are extrinsic rewards. But as they level up through gamification, they could earn a digital reputation that is recognized wherever they go, by their friends, by the brand, and by other consumers—an intrinsic reward. As they continue to share their passion, brands can learn more about these passionate consumers and serve them even better.

 

On the other hand, brands that serve their customers well—customer centric—will also be recognized and rewarded. Consumers will give them more digital love and rate them higher among the competitors, which translate to more business. Customers will be more loyal, which means longer term business. Consequently, not only are consumers incented to share their experience with brands, brands are also incented to be more customer centric, so consumers have a good experience to share.

 

Traditionally, it is believed that the financial market is fairly efficient—prices on traded assets will reflect all available information in the market. However, consumer markets are far from efficient, because there is an asymmetric access to information on consumer transactions. Brands typically keep all the data on consumer transactions in their CRM system—completely inaccessible to other consumers. With Lithium + Klout, we can create a 2-sided platform where this data can be made public should any consumer want to share it.

 

shared value network_1b.pngAs more information is shared, the shared value to both the brands and consumers increase; and the consumer market becomes efficient. Thus ensuring both parties are getting the service and value they deserve in any transaction. This provides a natural counterpart to CRM that has a flavor of vendor relationship management (VRM) , where customers are empowered collectively with access to data on other customers’ experience with brands.

 

Conclusion

These 4 exciting possibilities are some of the most obvious from the Klout acquisition. These are not going to happen tomorrow and there is lots of work ahead. But it’s fun and exciting work. As we watch and learn more about the market, these ideas will evolve and new ones may present themselves, too. But if these are obvious, what could be some of the less obvious outcomes? Who knows? Who could’ve imagined that Google will be developing a driverless car 10 years ago? Having the consumer data, the big data infrastructure, and the data science talent can open up a whole new future that we can only begin to imagine.

 

As Peter Drucker once said, “the best way to predict the future is to create it.” With Lithium + Klout, we have the necessary ingredients to create something truly extraordinary—something extremely disruptive and yet transformative. So let’s take our combined expertise and assets to create a better future for brands and consumers. That is why I’m so excited.

 

Alright, what’s next? When I started writing about this acquisition, I wasn’t planning to write 3 parts. However, as I was writing part 2, I felt something was missing—the human side of the story. I’ve already shared my reactions to the Klout acquisition from a data science perspective and from a future product perspective. Next time, I like to share something more personal. That’s the final part coming soon.

 

Until next time...

 


 

Michael Wu, Ph.D.mwu_whiteKangolHat_blog.jpg is CRM2010MKTAWRD_influentials.pngLithium's Chief Scientist. His research includes: deriving insights from big data, understanding the behavioral economics of gamification, engaging + finding true social media influencers, developing predictive + actionable social analytics algorithms, social CRM, and using cyber anthropology + social network analysis to unravel the collective dynamics of communities + social networks.

 

Michael was voted a 2010 Influential Leader by CRM Magazine for his work on predictive social analytics + its application to Social CRM. He's a blogger on Lithosphere, and you can follow him @mich8elwu or Google+.

About the Author
Dr. Michael Wu was the Chief Scientist at Lithium Technologies from 2008 until 2018, where he applied data-driven methodologies to investigate and understand the social web. Michael developed many predictive social analytics with actionable insights. His R&D work won him the recognition as a 2010 Influential Leader by CRM Magazine. His insights are made accessible through “The Science of Social,” and “The Science of Social 2”—two easy-reading e-books for business audience. Prior to industry, Michael received his Ph.D. from UC Berkeley’s Biophysics program, where he also received his triple major undergraduate degree in Applied Math, Physics, and Molecular & Cell Biology.
2 Comments
Khoros Guru
Khoros Guru

Regarding this whole Lithium + Klout story, i think 'exciting' is a true understatement @MikeW ! Smiley Happy

 

Thanks for your so inspiring article ! 

 

Context specific score is clearly something that resonates when, today, you pay attention to how much topics can be different for one single person..... like for me (social, community, software, photography, rock music, UI and..... Klout !!). Your influence can't cleary be the same on such various subjects. Same for brands with diverse product lines for instance.

 

So, among so (too ??) many other possibilities, being able to refine the model in this direction sounds really interesting (but i could have said 'extraordinary promising') !!!

 

Lithium Alumni (Retired) Lithium Alumni (Retired)
Lithium Alumni (Retired)

Hello @ArnaudL,

 

Thank you for taking the time to comment and share your enthusiasm.

 

I just like to share a few thought to peek your interest in the topic of context specificity, b/c the context specificity in any measurement has its strength and weakness. The strength is, of course, it's more personal, more relevant to individuals, and people will have more utility for it, and so people will care more about it.

 

However, context specific measurement automatically means that we can't compare them, even in cases that we want to. So in some sense, any context specific score or measurement becomes meaningless if we cannot compare it to something else.

 

Another problem is that context specific measurement is actually not a well defined problem, because everyone person has a different context. And context is typically specified by the problem that people are trying to address. There is no limits to how specific we can get, because the types and the number of problems in this world is infinite. This makes it hard to enumerate all the context, let alone defining it.

 

It is a challenging problem, but a fun one that I'd like to work on. 

 

Thanks again for sharing your excitement. Let's continue the conversation as we develop this idea in the future. I'll be looking forward to hear your thoughts...

 

Until next time...