User Metrics and Other Data

Continuing the discussion from Samsung Developer Conference (SDC) 2019 (October 29-30, 2019):

@tgauchat @ady624

This all depends on what data they are actually collecting. The source code is available to view.



They know the number of users and Pistons (and executions). I assume they know the distribution of number of Pistons created by each user, the rare of user growth, …

400,000: Gosh that’s a trivial number for product that’s been on the market for over 5 years.

The rest of the 31 million, with only 6 devices or fewer, hardly any Automation can possibly be complex enough to benefit from WebCoRE - And, frankly, such a small customers are also unlikely to substantially benefit from ActionTiles either.

Of course, a change in the market / consumer behavior could happen for many reasons unpredictably - but it would be reassuring to believe there are some possible triggering events on the horizon.


Remember that with the consolidation of the various Samsung apps over the last 18 months, that 31 million now includes everyone with a smart Samsung appliance or a smart Samsung television. :wink:

As far as benefitting from complexity, it’s not about the number of devices. It’s about whether or not they have stacked conditionals in the use case.

“When Charlie comes home between sunset and sunrise, turn on the entry light to 100% brightness unless Michelle is already home And is watching television in the living room, In which case turn the entry light on to 50% brightness.” Only two devices in that use case. :wink:

True - But the odds of a “stacked conditional” (or other sorts of complex rules) increase somewhat exponentially in proportion to the number of devices.

Most customers with few devices are probably likely to forego the learning curve and complexity of an external rules engine, even if it make their few automations “smarter” than they would be with the built-in options.

1 Like

For some more context, WebCoRE alone has 20,000 users with 100k automations that execute 1 million times per hour.


This data would be more exciting if correlated to the number of Devices per user, as well as a distribution curve of how many Automations per user (not average - but power users vs median vs average vs very sparse / light ones.

The 20,000 users alone is nothing to sneeze at, even without any of that other information IMO.

1 Like

:sneezing_face: :smirk:

All usage and sales statistics from any organization are subject to interpretation and analysis. That’s why corporate accounting is such a complex and important field… i.e., to ensure investors have a consistently accurate and comparable view of a company’s performance over time and over other companies.

While I have no doubt that WebCoRE has a non-trivial number of users (and that’s a large part of why SmartThings recognized them at this year’s SDC), as owner of a business selling a SmartThings tool (duh, ActionTiles), I have a particular set of data regarding user retention and attrition, etc…

Since I don’t have the detailed data about those WebCoRE numbers, I very curious as to the details, as this type of data about a set of SmartThings customers with particularly unique characteristics, would be informative and possibly actionable.

Not an apples-apples comparison, obviously, due to WebCoRE being free of license fees, but then again, we happen to have detailed data on SmartTiles usage; which was significant enough to be the only Community-birthed SmartApp mentioned in the keynote at the Samsung Developer Conference in April 2016.

SmartThings’s “successful developer partners” (not counting strictly hardware brands - especially since hardware tends to be cross-platform, whereas the “apps” tend to be SmartThings only) number far, far fewer than Alexa’s (as an arbitrary and perhaps irrelevant example). So each of us deserves to be proud of what we’ve accomplished. But from a market research perspective, I’d sure love to know more detail about each such enterprise.

The success story of WebCoRE is nothing to shake a stick at. I haven’t poured through the terms of service and ownership (?) details; and I guess this really drifts off topic, but I wish that, as a “Community Project”, the usage data would be open, not just the source code. Such data would be valuable to the … “Developer Community”.