In a startup’s never-ending battle for new users, data is king. When the decision to put that shiny signup button down here vs. up there can mean the difference between 40 percent of new visitors signing up instead of 20 percent, good data analysis can be what puts food on the table.
YC-backed analytics service Heap wants to make analytics better. They want to help you to code less, but grow more.
Heap’s approach to analytics is, in a sense, backwards from what many web developers might be used to.
Check out their demo video, below:
With many analytics tools (whether its something built in-house, or something like Google Analytics’ event tracker), the decision to track a new metric can take days to implement. First, you’ve gotta decide what you want to track. Then you write and test the code — or, in the case of a bigger company, you wait for one of the engineers to write and test the code (a task that’s probably somewhere around 200 items deep on their to-do list.)
Then you wait. Since you weren’t capturing that specific data before, it’ll be a few days before enough data trickles in to be meaningful.
Because of that torrential stream of data constantly being logged, you’re able to come up with new questions and have answers immediately without writing a line of new code. So you made a surprise appearance on Reddit’s front page yesterday and want to know how many of those new visitors tapped your page’s drop-down “Share” button? The data is already there, ripe for the perusin': just tell Heap which DOM element is the share button, then tell it to count the clicks based on Reddit as the referrer.
User groups (like, say, everyone who signed up after coming from Reddit) can quickly be bundled into “Cohorts” while “Funnels” allow you to measure metrics across a certain series of events. Want to know where you’re losing the most would-be users in your multi-page signup process? That’s what funnels are for.
There’s a matter of user privacy that’s worth discussing here, but it’s a tangled enough topic that it’s probably worth saving for another day — or, as it’ll probably get brought up down below, for the comments. As users, we all expect certain pre-set actions (page views, signups, etc) to be logged for later analysis, but isn’t logging everything just a bit… much?
As you might imagine, capturing this much data requires a pretty heavy amount of storage, so this thing ain’t free. The price scales based on the number of unique visitors a site has. Got 100 users? That’ll be a buck a month. 2,500 users? $25 per month. 200,000 users? $1,000 bucks. 500,000 users? That’ll be $2,000. You can tinker with the pricing scale right here.