Subscribe
Sign in
Sherlock — Simple, powerful user engagement analytics
28
Replies
@dskaletsky @cpresc I'd love to crack the nut on unengagement analytics. To me, that's where the value is. Sure there are things you can do with highly engaged users. But the value of knowing when someone stops engaging and why is huge.
To get there, we can do a few simple things. The easiest would be to allow filtering by users who are trending down and sort by users who are trending down more than others. From there, it would be great if you could dive in and compare the difference between current usage and recent usage. For example, say you've got a tool that lets users order breakfast. One day you decide to redesign the tool to make it more powerful. Frank, a loyal user, doesn't understand the new 'improvements' so he stops using the tool and goes back to his old breakfast routine.
So if Frank isn't performing an important action, you'd instantly be able to see him in the negative trending users and see that he is suddenly not ordering breakfast. From there, customer success can reach out to him and solve the problem before he churns.
Hi PH'ers - Happy to have Sherlock hunted! We're very excited to release Sherlock - our latest offering for product teams looking to understand how their users are engaging with their product.
With SHERLOCK, you create a custom engagement algorithm for your product by scoring each one of your events based on its "engagement weight" and Relay delivers you a ranked list of your users ordered by their engagement. It's completely flexible and allows you to see how user engagement shifts over time. You can also dive into detailed profiles with more info about each user's overall engagement.
Early users are getting almost immediate value, so we're really excited to open it up to a wider base.
Happy to answer any questions...fire away!
@dskaletsky Very cool idea Derek! Love that you can use a slider to carefully weight each event, and that it uses Segment so it works with tons of analytics tools out of the box.
@imakestrides Thanks, Kyle. You bring up an important point - Segment. We are huge Segment fans and built this MVP off of Segment exclusively. So...it's an important point - you can only use Sherlock via a Segment integration right now.
Cool concept! How is this integrated in the product? I'm assuming we'd need to add your Javascript and then need to manually setup each event that we want to track?
@guy Hi Guy - great question. Right now, we're built off of Segment.com - which makes integration super simple - assuming you're already a Segment user :) If not, we will have other methods of integration soon. Would you prefer JS or a Rest API?
@dskaletsky Segment sounds like a good start.
I think having a JS call that you fire when there is a relevant event, would be good
2 cents on this: engagement is fairly easy to find and calc. The harder stuff to discover is retention. Modelling how often you want a user back vs how often they do come back. Finding example users who model that behaviour and enabling more to do so...
Ah, good start. It would be awesome if the product head could see which product works and which features to retire against conversion numbers 😄 Let's say if this tool can predict or conclude about Uber's aha moment "after 5 trips users become repeat buyer" or Facebook aha moment "after 4 friends connection in a week, users become loyal Fb users". Those segments just aligns with product goals. Simple and crisp.
Very cool Tool. It makes our life simple and easy. To track engagement metrics and send them to Segment (which is integrated with Sherlock), i think you can combine this tool with www.weelytics.com too. So Segment + Sherlock + Weelytics = Zero coding and a lot of benefits :-) Good job.
For the last few years I see that customer engagement calculation is automatically done/calculated by more and more SaaS services. I understand that you can "actually" define your engagement KPIs to have a summated metric - in fact a sum is really avoided and not suggested in metric world, as it's only a number and doesnt represent much value.
But the main problem still remains: Ok, I have a list but what am I going to do with all those users? Send them a push notification? Offer a discount? Provide a means to dig further by sending out drip marketing messages?
What are your plans with this issue?
@gorkemcetin Hi Gorkem - great comments. I agree. The raw number is fairly useless by itself. But it does provide the basis upon which you can compare users and get the ranking. Also, we are adding more time-based functionality soon which will show trending in the raw score which gives it more context.
But, totally, what do you do with this information is where the rubber meets the road. our main product - knowtify.io - is an engagement messaging platform, so Sherlock is a piece of an overall "User Engagement" play. For the Sherlock MVP, we allow you to export your ranked lists to CSV so you can at least have the data to take action on...
@dskaletsky Why are they separate if you don't mind me asking? Will they get a closer tighter integration in future?
@sheppyc Hi Chris - kind of a longer story - I'd be happy to connect with you offline on it (derek@knowtify.io). Short answer is that we thought that this take on analytics could deliver a lot of value to companies/teams that didn't need the messaging piece. They will all be tightly integrated shortly (we have another product called Relay which allows you to connect user events to Slack - https://www.producthunt.com/tech...).
Of course knowing specifically WHO cares most about our product is an important step on the road to knowing why and how to improve our targeting. Thank you for building this! We love it!
@samharris82 Thanks, Sam. Yes...WHO is the guiding principle of this product. We always say..."Who Matters?" and we hope this product will help you figure that out quickly and in a helpful way...
Hey Guys, great job on releasing this. I like the clean website. On my hand, we are considering using mixpanel for this. How would you guys be better?
@chubucko Hey Jean-Mark - thanks for the question. Btw - I would like to give props to @lenndizzle from our team who designed Sherlock from the ground up (UX&UI). He did an amazing job in a very short timeframe.
If you haven't started using MixPanel, you are in a great spot for Sherlock. MixPanel is a great analytics solution, but it's pretty heavy and complicated to get value from without considerable effort. Our goal is (a) to be a much simpler solution that will allow any company to get up and running and seeing value within a couple of days (depending on how active the users for your product are); and (b) to focus on user-level analytics - as @samharris82 mentioned above - we really thing WHO matters most. When everything is organized, primarily, at the user-level, analytics become much more actionable.
Sorry for the long-winded answer - does that help?
@dskaletsky Thanks for the answer. At the moment, we are looking for a simple solution as we dont have heavy usage yet. btw its Jean-MarC ;)
@chubucko whoops...sorry, marc. let us know if you need any help getting set up...
IssyBot Wordsearch