Very excited to be the one that gets to hunt this!
I remember sitting down with Kyle and Ryan in a cozy little bar in the dogpatch back in 2011 to talk about a new project they were working on called "Schmetrics". I was working at Olark at the time and I think we somehow met on Twitter. Over the years we've stayed in touch and it's been really amazing to see that little seed of an idea grow into something so awesome and big.
So as a user (we at Hardbound <3 Keen) and a friend, congrats to the Keen team on launching this great update!
I'm curious if tracking public content is the right use case for Keen . Assuming a document sharing public website where one of the metrics its users get is a number of views. One document goes viral with Ms of views. Does that mean that we'll need to upgrade to the highest pricing plan? Or is there a more cleverer metric to measure in that use case?
@nikitakorotaev Hey Nikita, you've astutely touched on a challenge of our current plan structure that we're looking to address in the near term. We absolutely can support this use case! A key Native Analytics feature Dashboard Caching means that once you know which metrics are valuable broadly, we can optimize them for you
I'm one of the makers of Keen IO.
We're super proud of Native Analytics! I'll be hanging around here all day with some other members of the team, as we are super eager to get feedback from the PH community.
I thought I'd share a little back-story.
We started the company as a "build whatever you want with it" API for Analytics, where we handle the storage/uptime/computation, but are not prescriptive about how you use it. We think of ourselves as Lego for analytics: we create handy building blocks that fit together nicely, which lets the builder (software & product teams) create whatever they can imagine.
Over the years, we've found that a lot of companies are using us for white-label use cases (where some of the analytical features they're building on our platform are for _their_ customers). This is a bit like how, when you text your Uber driver or call your Instacart shopper, you're using Twilio under the covers. Similarly, there are a lot of SaaS and mobile companies whose products have analytics/reports/insights tabs, and under the hood those are being powered by Keen IO.
These white-label use cases create some unique requirements for our platform (for example, while load times for employee-facing visualizations can vary, load times for customer-facing visualizations need to be lightning fast; so we created Dashboard Caching, which pre-aggregates the data for view much like the way we handled the load time problem in my time at Google Analytics).
At the end of last year, we decided it was time to create a full-featured product for white-labeling Keen, informed by our learnings from having seen hundreds of these use cases.
That brings us to today, where we're thrilled to be launching Native Analytics!
Looking forward to your feedback :)
@dorkitude So happy to see this coming out right now, we were discussing just this morning how existing solutions were coming short for this precise use case. Consider me a fan!
@kevin_granger thanks, Kevin! Means a lot coming from you. The care, thought, and attention to detail y'all put into your work at Algolia is evident from your homepage to docs to API design itself, and down to every piece of marketing material I see. Kudos!
@rdardour@dorkitude thanks Romain! I remember chatting with you years ago, back when we had not yet a single paying customer. I was a newb when it came to modern web development, but you didn't hesitate to chat , tell me about your project, and give feedback. It helped me learn and begin to understand the developer community. I will always appreciate that!
Love, love this. Keen definitely built a kick AaaS model here. Less code for things that aren't your core is the way to go. If your not an analytics co but need to provide them, Native Analytics is awesome!
@mikerowan Haha thanks, we set out to kick AaaS and chew bubblegum and well... we're all out of bubblegum
Very excited to see even more folks collect data via the SendGrid integration and then feed valuable insights back to their users.
Congrats on the launch - big fan of what Keen's doing! But I've always assumed that this is what you would use Keen for, how is this different than the Platform?
Great question, @mylifeisshan -- and rather incisive. I have a philosophical answer and a specific one.
First, philosophical:
You're right that white-label use cases are definitely a part of how the Platform (i.e. our API) has been used in the past. In fact, the demand we've felt from the market is precisely why we decided to spend the first half of 2016 working on Native Analytics.
For reference, it's only one of the five key use cases we've seen for the Platform:
- White-label analytics built directly into your software product
- Feature utilization analysis (think like a build-your-own Amplitude or Mixpanel)
- IoT device analytics (for instance, we're used to analyze smart sprinkler systems, in-store retail beacons, the John Deere smart tractor, the NASA Mars Rovers)
- end-to-end funnel analysis (a bit like a build-your-own Google Analytics or KISSmetrics, but one in which there is no bias toward "pageview" events, because everything in the Keen IO Platform is a custom event)
- Automation (like a build-your-own IFTTT on analytics data -- for instance, see Pushpop for Slack which we launched right here on PH! https://www.producthunt.com/tech... )
Our theory on Platform vs. Product is complex, but the gist of it goes a bit like this: the Platform-maker's perspective is often "we should not build features for a given use case when we can instead build a more-abstracted version of that feature, thereby covering a a broader potential space of use cases." Platforms are strong because of `breadth` in the `future`. The Product-maker's perspective is "we should learn as much as possible about a _finite_ use case, and purpose-build things to make that use case better and better." Products are strong because of `depth` in the `present`.
Keen IO has historically been a pure Platform company, but Native Analytics represents an evolution of our culture, in which we are adding the Product-maker's discipline. Expect to see more Product launches from us in the coming years :)
Now, for the specific answer:
In terms of substantive product innovations, this guide written by my colleague @michellewetzler is probably the best place to drill in -- https://keen.io/guides/native-an...
@mylifeisshan Hey Shan, thanks for the kind words! Over the years we've seen many uses for the Keen Platform - internal product teams, cross-company workflow tools, A/B testing, customer success management, and more.
Keen Native builds on top of the core functionality to emphasize the needs of customer (or consumer) facing analytics. They have less patience and there are often hundreds or thousands of them! Not only do you need blazing, consistent performance and beautiful visualizations, but in order to deliver Native Analytics successfully you will need features like Project Provisioning and Custom API Keys to manage the onboarding of new customers seamlessly instead of manually.
@dorkitude@mylifeisshan@peternachbaur I'm a bit conflicted.
As someone who has been using Keen for this use case for nearly two years, I'm glad to see this embraced, philosophically. The part I'm miffed by is that, from what I can tell, the features listed have existed for months, so this feels like repackaging more than a product release.
I'm biased but hopeful that this change will be followed by tangible improvements that make this use case even better.
Hey @hijonathan -- good to hear from you & thanks for posting. I understand why you'd feel that way, but let me explain:
It's very much our nature and style to make new innovations available to our customers as soon as they're production-ready, because we want to make an impact on our current customers ASAP. Philosophically, this is because we tend to prioritize making an impact on our current customers ahead of the sort of marketing boost of the "build it all behind closed doors, then unveil it all at once" approach taken by many other companies.
So it is indeed true that much of the features in our Native Analytics suite have been available to you and other Keen Pro customers ahead of the commercial launch.
Two ways you might view this launch in a positive light:
1 - As a long-standing customer, you got access to these features before today, which is good
2 - We are planting a stake in the ground that says "this is the first use case we are doubling down on in such a big way". As this is our first true Product (as opposed to Platform) launch, you can be sure your use case at Appcues is top-of-mind for a significant chunk of our product management & engineering staff, which means indeed, there will continue to be tangible improvements to make this case even better :)
@hijonathan@dorkitude@mylifeisshan Hey Jonathan, that's a fair observation. I think in general, our ability to communicate will lag behind our ability to execute - we've been driving really hard on this use case and are now taking the opportunity to talk more broadly about it. We're eager to tighten that loop moving forward.
Speaking of which, we are just getting a new feature for Native Analytics across the finish line - we're super excited to shine a spotlight on our new Access and Permissions Control soon. This API will enable users to create Custom API Keys that have much more power than the current scoped keys. For example, the ability to automatically add properties on write, as well as the very important ability to manage and expire your keys programmatically.
Would love to hear your thoughts on where you see rough edges in delivering Native Analytics, what sort of tangible improvements we can make!
Awesome work, Keen Team!
I've been a long-time user of Keen at various startups and now at Suto. Glad to see Keen supporting the analytics-for-customers use-case as a core product. Great job!
@ngardideh Thanks for the kind words Nima, your support and feedback have been invaluable! Do you have thoughts on how to leverage Native Analytics at Suto? Looking forward to hearing how the new features meet your needs, and what can be better :)
@peternachbaur We're currently using Keen for our internal dash-boarding needs, but we're thinking about exposing some of the information we have about products in the world to our users (X people have purchased this product through us, Y people have reviewed it this way, etc.) to help them understand how our decision making process works.
Super excited to give this a whirl! :)
Having integrated Keen before, I'm excited for the new offering. I like how Kyle pointed out performance constraints and requirements for customer facing dashboards being much more demanding. I'm curious to hear how Keen evolves, especially in the IoT space.
@kicksopenminds glad you're excited, Taylor! Let us know if you have any questions about our new offering :)
We're also very curious to see how developers adopt and apply Keen in the IoT space! It's going to be a fun couple of years coming up.
@rywalker thanks for the kind words and recognition, and of course for being an awesome community member, customer, and supporter of Keen. We're excited to be working alongside you in this exciting space! 🙌📈
@kmircovich thanks a ton for your kind words and feedback! It means a lot to know that we're making people happy. Always feel free to let us know how we can turn happy into happier! 😃
Ship ship hooray! Love to see this doubling down on a use case that Keen IO is so well-built to handle. And I certainly wouldn't be the first to say something nice about the team :)
There is actually a market to make here when you think about it. How many products *could* be improved by customer-facing analytics? Almost every single one. But today only the best products have baked-in analytics because it's a full-stack engineering commitment. If Native Analytics lowers the barrier to entry far enough you could see analytics product features go the way social login / social features did: rapid expansion in a very short time.
I've helped the team a bit with this product and find it super compelling for any software product team that wants to offer in-app analytics to its users, but may not have the resources or bandwidth to build out an entire analytics infrastructure and user-facing product. Keen's Native Analytics product gives product platforms and PMs the ability to ship analytics to their users within their product very simply.
@binder Seth, you've been a huge help from the very start of this process. Thank you for all of your insightful questions and feedback along the way! We couldn't have done this without you and the Sequoia team!
@arieljalali a very humble lot of gratitude to you, @CrystalRose, and the rest of the Sensay team...for always being there as friends, actively contributing to the Keen community in fun ways, and of course being model customers and data-explorers-turned-discoverers 🙏
I have known the Keen team for years and I have been constantly impressed by the evolution of their platform. It is great to seem them branch out and build a product to address specific use case that I know a lot of their customers have been asking for. Well done.
From the first time I met Kyle via Startup Bus 2011, I knew he was going to do great thing! Keen IO and now, their Native Analytics is proof of his genius and tenacity in building great, bomb-proof solutions! Congrats!
@lksugarman I had a similar experience when I met Kyle, aka @dorkitude, while traveling through the Middle East with Geeks on a Plane in 2012. Our friendship grew over the years, and when I began exploring new professional opportunities, Keen IO - with all the people Kyle and friends had attracted - was clearly a place I could call home. I've been a happy member of the team ever since that realization in 2014.
It's always nice to hear other's stories like this. Have a wonderful Wednesday!
~Tim
You guys have come a long way, Kyle! I still remember the YC office hour and meeting you in person through Techstars/GAN conf years ago. Great job on persevering and finding the right fit!
We've been using Keen for our VoD platforms at AJ+ and love the platform's accessibility. Excited to see native analytics finally implemented as a full product. This will make it so much easier for us to share data with internal and external stakeholders.
Lex