- Some good quality coverage during PR Phase 1, although not as extensive as we would have liked
- Very close to 1,000 subscribers total
- 13M net photos added (i.e. taking into account deleted photos from accounts leaving trial)
- Significant product features released
The goal for this phase was to get limited quality coverage to start getting the Everpix name out there, but associated with a clear explanation of our vision and positioning on the market. The PR contractor we hired set up a mini press tour in NYC and the quality of the coverage matched our expectations e.g. CNET, VentureBeat and SmartPlanet (part of CBS Interactive).
However, we also use the 12/18 milestone to do a more regular PR outreach where the PR contractor sent a press release about Everpix and our latest “big” news (new Windows Uploader, updated iOS app, Everpix.com on mobile...) to a number of tech focused websites. That part didn’t work at all, probably due to a combination of factors from improper messaging to bad timing.
We reached out directly to some French Mac websites with which we had a previous relationship which resulted in good coverage. Unfortunately, we also missed an opportunity of coverage on TechCrunch because of timing issues and concerns of them not necessarily honoring the press embargo.
The conclusion of this first PR phase is that (unsurprisingly) we have a product that requires some education since it’s ahead of the curve compared to the photo platforms people are used to today. If we sit down with the journalists (or other important people for that matter) to explain our vision and execution, they typically get it and are genuinely excited.
As usual, you will find the full list of press articles at https://angel.co/everpix#press.
As a reminder, PR Phase 2 is focused on getting widespread coverage, both regional and national.
We are currently deciding among 2 PR firms and plan to sign the contract and start the work by mid-January. To piggyback on PR Phase 1 conclusions, the SOW is mostly getting 1:1s with key journalists.
We had about 1,650 signups this December and are now very close to 1,000 subscribers total. We sold quite fewer subscriptions than during the previous month but this was likely due to the Cyber Monday promotion of 11/26. We expect many would-be-subscribers to have subscribed late November instead of later in their trial in December which skewed the metrics.
You will find 2 new graphs as the end of this report:
- The first one tracks the number of Everpix users as well as “adjusted” AWS expenses (this means infrastructure expenses strictly related to our production servers i.e. infrastructure variable costs).
- The second one shows our net revenues (i.e. excluding fees and refunds) from Stripe (the payment processor we use on www.everpix.com) and Apple. The “Adjusted Total Net Revenue” curve shows what our monthly revenue is using the “sales basis” revenue recognition method (i.e. spreading out yearly subscription revenue over 12 months). Note that in QuickBooks, until this is fixed by our accountant, we are still booking the entire income from yearly subscriptions on the month they are paid.
You will notice the average infrastructure cost per user appears high at $1.40. Keep in mind we have a one-time cost for every new user who imports photos because of our image analysis. Our initial adopters also import more photos (10,000 on average) than what casual users would later down the road (likely around 6,000), which increases this initial cost even more.
The projected infrastructure costs computed last summer were planning on a $1.00 one-time cost for a new user (with around 6,000 photos) followed by a $0.50 / month cost i.e. about $1.50 for the first month, then $0.50 afterwards. If you input our December numbers into the equation and adjust for 10,000 photos, you get a result higher but close to the actual infrastructure cost. The difference is because of us having more AWS capacity provisioned than we actually need.
No matter what, we’ll want to reduce this infrastructure costs, which means focusing on 2 things:
- Increasing the trial-to-subscription rate to recoup as much as possible of the $2.00 an “average” trial user costs us.
- Optimizing the way we scale in AWS and taking advantage of various cost-saving features like EC2 spot instances.
We don’t have visible improvements yet. We know what we need to act on, but December was really busy wrapping up a number of end of year features for the PR Phase 1.
Also please ignore the Active Mac/Win Uploaders KPI as we realized the formula is just not meaningful and we need to revise it (as defined now, it will continuously go down no matter what).
We spent a lot of time testing and optimizing www.everpix.com for mobile browsers and now our website should be fully usable on smartphones and tablets.
This work was required to improve the experience of Everpix users when accessing Everpix on the go, especially considering a number of “advanced” features are only available on the website. It will also be some time before we have a native app for Android so this helps not shutting off completely Android users until then.
Everpix Windows Uploader 1.0 has been released to everyone as part of our 12/18 announcements. About 300 users have been using to sync close to 2M photos so far. The crowd that tries Everpix is still predominantly Mac based, so these numbers will take some time to grow. The good news is that there have been no significant bugs reported.
The next step is now to bring the Windows Uploader up-to-speed with its Mac counterpart i.e. add support for direct camera import, Picasa and Lightroom. Although Picasa is the equivalent of iPhoto on Windows, it might unfortunately not be possible to support it as recent updates have changed its database format and there’s no official interface to it.
Also part of our 12/18 was a brand new version of iOS app (1.4) bringing 2 key features:
- A single universal app instead of separated iPhone (Everpix) and iPad (Everpix HD) apps. For historical reasons, the two apps had a lot of code that was not common, so it took several releases to unify everything - trying to do it all at once would have prevented other progress on the apps. A universal app is better for our users and our exposure in the App Store.
- Location based sync: a big problem with iOS apps that sync the camera roll to the cloud is that, since iOS apps are not allowed to run continuously in the background, the user needs to remember to launch said apps regularly. If they don’t the latest camera roll photos don’t sync, which affects the overall experience of the product. In version 1.4 of our iOS app, we took advantage of geofencing: from the app settings, users can save “locations” (e.g. “home” or “work”). Whenever they visit these locations, iOS will automatically launch the Everpix app in the background, which will then sync the camera roll if there’s anything new. Because it’s not the typical use of geofencing APIs, we weren’t sure the App Store reviewers would approve the feature, but they did, and we believe Everpix is the first app to use this elegant solution.
It’s worth noting there has been significant changes among what we considered to be our closest competitors when we started more than a year ago:
- Gush: shut down in December 2011
- ZangZing: shut down in July 2012
- Snapjoy: acquired (acq-hired?) by Dropbox in December 2012
- ThisLife: acquired by ShutterFly in January 2013
- PictureLife: still going, raised $4M in November 2012
- Adobe Revel: still going but recently switched from subscription model with 30 days trial to a freemium model
We didn’t notice either many new entrants on the playing field apart from MyShoebox which got some repeated coverage on TechCrunch.
Last but not least, we will be running out of funding on June 1st assuming our current burn rate (which is unlikely to change). Technically there would be around $130K left on June 1st, so enough for 1 more month, but this is already budgeted for PR costs and some margin for unplanned expenses.
We will therefore start working on the next financing round this February as it will most likely take 3-4 months to complete.