It looks like Vuforia will be focusing on Fortune 500 companies moving forward, which is probably what we should expect from a large company like Qualcomm. This is not a startup like Metaio or PointCloud. There are ROIs to meet, and pressure from the top to make it profitable.
Their market research team on this pricing strategy looks like they've decided to drop the indie developer/small agency/entrepreneur demographic. If that is unintentionally, then maybe they assumed the watermark would be acceptable, or perhaps they falsely deduced that developers make more money on apps than they do. Either way, the #1 AR SDK will unfortunately be reserved for developers with access to larger company budgets at this point.
This pricing strategy does not work for consumer mobile applications whose aim is to make money via the application. At all. It does work for large companies pushing their app out to large audiences, because they can afford to pay a few grand a month for their marketing strategy. AR experiences are, for them, only a part of their larger marketing plan, and don't need to turn a profit from the experience itself. Requiring $5/user ROI on an app just to pay off the AR SDK is ... literally laughable.
Now, if they did want to recapture the smaller agencies and indie developers, a more reasonable pricepoint would be something like:
- $99-$499 / yr. for unlimited offline 'recos' (Sure, limit your cloud services. Most indie dev's are basically looking for a glorified OpenCV plugin.)
- OR a one time charge like Metaio for a developer account (I would personally pay in the thousands for this).
- They could continue to charge monthly for 'cloud' recos, but it should be in the ballpark of $50-$200/50,000 cloud recos.
- Remove the ridiculous 'overages' - it's an appalling money-grab. It alienates everyone. I'd rather just get an email at 75% quota use and be warned that it'll stop returning results soon unless I upgrade. At least give the option to opt out of overages. Geeze. Surely someone at Qualcomm has made a mobile app and understood what 'virality' is. Download spikes (even from countries in timezones while you sleep) are a fact of life in mobile dev.
Again, Qualcomm doesn't want to do this, in all likelihood. They don't make enough ROI on the SDK from it.
As is, the core of this forum (and the majority of their current developer base) will likely be leaving within the next few weeks to find other alternatives. They've 'bait and switched' the majority of their AR dev's.
If nothing changes on Qualcomm's pricing structure (and they ought to send someone in here pronto to clear some things up...) then I encourage the community to look into OpenCV alternatives (perhaps AR Toolkit) and combine our efforts to get a solid open source AR SDK up and running. OpenCV has taken us pretty far, and Unity can do the rest. Most of us don't need 'Smart Terrain' or Vuforia's other features - we need simple coordinate locking based on a marker, then SLAM from that point forward to keep the enemy's gate down.