In terms of price sensitivity, current vendors are charging $100-$150 a label, so the subscription plan is either an average deal, or a great deal if they have 5 labels.
The subscription model stems from a problem in purchasing an individual label. If vendors purchase say one label, there is a possibility where they might have made a couple of mistakes and they have to fix their label. The thing is I don't really know if they are fixing the label, or generating an entirely new label. At the end of the day my goal is to be as hands off as possible.
Of course, I'm open to ideas though.
Or you could do like it is done with barcodes, and charge a low fee per label generated. Say $20 for every label with no corrections. Just state it very clearly in there. Twenty dollars is low enough that they won't try and scam you out of things. Plus you can also sell them a package deal where for $99 they can get 6 labels (5+1 free). It will make sense in their heads because the $99 deal seems to be the better value.
So:
$20 per label or $99 for 5 labels plus get one FREE.
The idea came about from a pivot from a failed app (http://fs.howcookingworks.com/). In that app, users can take pictures of food, and get nutrition info back. It turns out that the need for nutrition analysis is needed much more in the enterprise space.
The validation of the idea came when pitching my local restaurants if they needed nutrition facts labels. It turns out that one restaurant did. After that, I hit the road visiting all my local farmer's markets and have been and gathered my first set of clients from those visits.