Editor’s Note: This post was originally published on Seattle 2.0, and imported to GeekWire as part of our acquisition of Seattle 2.0 and its archival content. For more background, see this post.
By Richard Luck
There is a curious little phenomenon observed in quantumphysics. Particles of matter popinto existence, then out of existence, or may even materialize in an entirelydifferent location from where they were previously observed—and all seemingly at random, without purpose or reason.
Now I’m not a quantum physicist—far from it. I’d be hard-pressed to even spell physicist correctly if it weren’t formy trusty spell-checker. But Ilove a good mystery. And whymatter can simply be then not be is a whopper of one. Physicists have several explanationsfor why this is possible. Most ofthem I couldn’t begin to explain here if my very life depended upon it. The only thing from all the episodes ofNova I’ve watched that has stuck withme is this: The simple act of observing the particle seems to causesit to come into existence. Whenthe observer is removed, the particle ceases to exist.
The whole thing reminds me a philosophy professor I had incollege who once stated: “Reality is perceived. If you change one the other will change as well.”
So what does any of this have to do with running astart-up?
The other day I was compiling data on our HeyPublisher service in preparation for ameeting with one of our advisors. Very briefly, HeyPublisher is an online service that brings togethershort-content writers with print and online publishers, providing submissionmanagement and other services to both parties. As more writers join, we should see submissionsincrease. As more publishers join,we should see submissions increaseas well. Consequently, the mostimportant measure of the success of our service would obviously seem to be the“submission” count.
Running the numbers, I saw a nice little graph that lookedlike this:
Granted, the growth rate of submissions flattened a bit overthe previous month, but user growth rate seemed to indicate that we were stilladding writers at a good clip, as the following graph would indicate:
The problem with these graphs is that they didn’t reflectreality. Or, more accurately, theyonly reflected my perception of reality. To understand what was really happeningwith our service I needed to change my perspective.
If you haven’t yet watched PhilLibin’s Founder Showcase keynote speech, you really should do so now. The video is a little over 20 minuteslong, but in it he talks at great lengths about how Evernote is making thefreemium model work, what they’ve learned from their users, and how theyapproached investors based upon what they discovered. The single thing that stuck with me when I watched it,however, is how Evernote has grouped users into cohortsbased upon the month in which the user signs up. More interesting is the observation that most of thecohort members will abandon the service within the first two months of usingit. Those cohort members who stay, however, stayfor a very long time. And thelonger they stay, the more apt they are to upgrade to the premium serviceoffering.
Taking this idea, I decided to slice our submissions data ina different way. I wascurious: What were the submissioncounts — by cohort — coming through the system? The following graph illustrates the results:
As you can see, most of the submissions coming through theservice in any given month are coming from writers who signed up for the service in that same month. In short — we’ve got a retentionproblem. Faced with this newreality we started digging furtherinto our data to see if we could figure out whythis was happening. As it turnsout, we discovered, the vast majority of these writers were signing up foraccounts and submitting their works directly from our publishing partners’websites. They were notcoming to our site to find a publisher — they had already found one. They were simply using our service tofacilitate the last step of the transaction. Given that, what reason did they have to come back?
Now we’re onto tosomething! A simple shift inour perspective and we suddenly have a new reality in which to operate. We now know that we have to do a muchbetter job of educating all of the writers using our service. And instead of focusing so intently ongross month-over-month submissions growth, we now have a new (more important)metric to measure: cohort submission growth. If we can grow this number, we know we have a service thatis of value to writers.
I’m certain there will be many more discoveries like this inthe coming months and years. One of the most existing (and challenging) tasks of running a start-upis trying to figure out what you don’t know. But if you keep your eyes open you’re likely toobserve the unknown become the known. Things only pop into existence when they’re being observed,after all.
Richard Luck is co-founder and CEO of Loudlever, aSeattle-based company which builds tools for writers and publishers. He enjoys talking about all ofthe crazy things he learns from Nova,Stephen Hawking, and/or old episodes of TheOuter Limits with anyone who will sit still long enough to listen.
