Public

Plans for Analytics in 2012

Lithium Alumni (Retired)

analytical-coffee.jpgBack in the early days of Lithium, social media analytics was generally not a hot topic (no pun intended). The primary measurements people were interested in were traffic and membership-related: page views, registrations, and so forth. These metrics gave our customers a way to assess the health of their communities and were simply time-based reports. It made sense back then.

 

However, as social media exploded onto the scene and our community platform became robust with interaction features, we realized we needed to devise better ways for our customers to measure ROI. And thus the Community Health Index (CHI) and a host of algorithms were born to help our customers better interpret and manage their communities. In order to create them, we had to scale up and build a data warehouse to combine metrics in new and interesting ways. This was circa 2008.

 

Today, social media has matured to the point where analytics are now the hot topic. Everywhere you go, companies are claiming to offer the best in analytics. However, if you look closely, you'll realize that they're still just simple reports with simple metrics. There is no relational data analysis. No clever insights into how to better manage your social media programs. Even our Lithium analytics fall short here because we currently do not drill down to board-level user metrics and mobile metrics. On top of that, the 2008 data model we created does not easily allow us to add new metrics, sustain an evergrowing amount of community data, and cross-map various metrics to create new reports for our customers. I'm allowed to say this because I am the product owner of our analytics product line. All eggs and tomatoes may now be aimed in my direction.

 

BUT. We're going to start changing all that this year. Last month, we made the difficult but ultimately right call to begin work on a new data architecture that will deliver to our customers the level of granularity, the scalable reliability of a growing data set, and the solid foundation upon which new metrics can be added easily in an everchanging social landscape. We're calling this "V2 Architecture". As this is being built, I am visiting with customers, scouring the Lithosphere for reporting requirements, meeting with our CSMs, and devouring social media analytics books to build a better analytics product for our customers. Together with Dr. Wu and our talented analytics team led by Tim Wong, we will build these requirements into the roadmap once the V2 Architecture is complete. If you as a customer ever wish to talk analytics with me, just send me a message and we can discuss your business needs. I always welcome insight into how our customers analyze and act on their data.

 

What this means though is that working on the bugs and issues coming from the 2008 data model may need to take a backseat until we're done with the V2 Architecture in July 2012. Essentially, in investigating our current architecture, we learned that the vast majority of the issues will be better solved in V2 (e.g., differences between Metrics and CIC, slow load times in the CIC, metrics missing from the product, etc.). My mindset is that if we're going to do something right, we need to do everything right. Now, this doesn't mean you shouldn't report a bug if you come across one. Please do, because there are cases where we can fix them in the current data model, and in each case, I'll be involved to triage them appropriately. 

 

With that being said, I just want to thank you for being a Lithium community owner and for having patience while we invest in our time and energy into the V2 Architecture. Our goal is to get everyone, including customers and our own metric-minded folks, the very best analytics possible to make the right decision, at the right time, and at the right place.

 

Thanks,

Stephen Hsu

Product Manager - Analytics and SMM

6 Comments
Khoros Oracle

Very cool to see this little history, as well as some hints on future plans, Stephen.

 

One thing that CSMs have considered as we prepare to help discuss these metrics imrovements with our customers is the how to work with a potential delta between old metrics and new metrics, should a delta like this arise.

 

To wit; if we (hypothetically) were calculating 'minutes online' differently in V2 Analytics, and a customer has been measuring 'minutes online' for a number of years already, will they have a formula on how re-calculate past metrics, or will we just bake in the new formula and have to run metrics from years past all over again?

 

Hopefully I articulated that correctly.  If not, and at the risk of putting too fine of a point on it, let me share a scenario:

 

The ACME Box company wants their community users to spend as much time on the community as possible.  Asking questions, reading blogs, reading TKB articles, etc.

 

The ACME Box company measures their communty's success (partially) by looking at how many 'minutes online' their top 100 users have spent on the community.  In fact, the Community Manager's compensation plan even pivots a bit on this statistic!

 

If the 'minutes online' metric ends up being different by about 10% in V2 analytics, will Lithium have a formula or some other kind of change-management technique to help explain this kind of delta?

Respected Contributor

This is wonderful news. For years, different product managers have asked "what are the hot topics?" These people care about a small subset of the community. The desktop guys don't care about the printer boards and vice versa. And even within the boards there are specialties -- the ePrint printing guys want to know about ePrint not setting up wireless print networks.

 

Because there are so many people with such different specialties in the company, we desperately need a deeper level of analytics. Monitoring tools can tell you the number of posts and the number of authors in a topic. BUT you can't get page views --- a critical metric when looking at hot topics.

 

The metrics you are talking about can put added teeth to showcase hot issues coming into the Forum and drive change in a company.

 

While I understand the complications of making this possible, it has been years of pain to get here. So -- I have decided to only throw a few tomatoes your way and will give you a towel to clean up afterwards.

 

Seriously, this is exciting news.

Honored Contributor

Thanks Stephen and Dr. Wu and Tim Wong and team for building us a new platform that will allow us the manage and understand our communities in the ways that are now critical for our businesses.  As you mention, we are all moving down the maturity path in these areas and looking for you to "bake" your thought leadership into your product.

 

I also had a suggestion taken from Hans Rosling's gapminder visualizations (http://www.gapminder.org/), if possible, the ability to create time lapses of some key weekly or monthly trends for a defined interval could really accelerate intuitive insight, explaining and showing executives and others in our organizations how our communities are evolving.

 

Stephen, eggs and tomatoes can make a nice omelette with salsa 😉

 

Looking forward to seeing this evolution in the Lithium platform, full steam ahead!

Lithium Alumni (Retired)

Jake, you raise a critical point that we are conscious of as we build out V2. In fact, part of the work Michael and I are doing involves an audit of all our metrics, both from the community and those that are loaded into the data warehouse. As you can imagine, this is a very time exhausting process, having to identify, investigate, and potentially redefine every single metric (hopefully not much of the last one though, but just keeping the possibility open since we're taking the mindset of translating community metrics into business / ROI metrics for our customers). So, to your point, we'll be keeping track of any formula / metric changes and only when we have a solid understanding of what we're dealing with, will we be able to properly assess any need for change management.

 

Siobhan, thank you for your support, understanding, and towel! As we achieve any milestones on our way to V2 Architecture, I'll do my best to keep everyone posted on progress.

 

Brian, thanks for the link to gapminder. I've shared this with UI/UX designer and will go over it with Michael in our many upcoming meetings around data visualization. I agree a time lapse feature would be telling story. If a picture is worth a thousand words, a video is worth um...well...I guess a thousand pictures depending on the frame rate. I'm also looking into dashboard chart visualizations right now. Have you seen: http://www.domo.com/ ? They haven't launched yet, but I'm intrigued.

Honored Contributor

Yes Domo looks pretty interesting, I will keep my eye on them, thanks for the tip 😉

Lithium Alumni (Retired)

This is great to see explained to all prospects, customers and employees some context and motivations behind the current product development efforts. This will come to fruition within the near and mid-term future (2012-2013) for benefits that will unfold over the decade.  Analytics are what enabled Lithium to crack the behavioral DNA of a community and its inner constituants (ambassadors, superfans, influencers, participants, etc.).