We are finally where we wanted to be almost 10 years ago: packaged food scanning via app! The dream of scanning the barcode of a food with a cell phone and then getting a reliable result regarding our own tolerances is something we already had over 10 years ago. That was when the “nmidb – Scan” project began. We then invested hundreds of hours of work time, did research, received funding, won innovation competitions, but then also got to know the harsh reality of the business angels and start-up scene. After narrowly avoiding personal bankruptcy, the project came to a standstill until, thanks to the stuffmakers agency, “Frag Ingrid” saw the light of day. But still without a scan, “only” as a grocery list with our nmidb data and a unique and innovative community rating at the time. We wanted to use advertising revenue to create a budget to implement the scan. Unfortunately, this failed because you can hardly earn any money with apps and the GDPR made it impossible to earn money with advertising.
The only way to recoup the development costs to some extent at the moment is to create a fair subscription model. It looks like this: Ingrid remains completely free of advertising. You can consult Ingrid twice a day free of charge. If you want to query data more frequently, you have to take out a subscription. This means that you can choose either € 3.99 per month or € 29.99 per year (-37%). That’s not even 8 cents a day!
If we manage to recoup our costs a little, we can start expanding the data. This is also costly and time-consuming. I once made a podcast about this, which you can listen to here. You’ll be surprised how much private money you invest in such a heartfelt project only to get nothing in return.
We hope that you will remain loyal to us and support the app with a small amount (8 cents per day). If this works, we can continue to invest in research and development and make “Ask Ingrid!” better for everyone… for example with a …
Food scan in the app
Yes, the time has finally come! In this new version, you also get the scan with your subscription. Why did it take so long? Because the scan is so complicated… Let me try to explain…
The crux of the matter with data
You first need certain data, such as ingredients, labels or nutritional information, to be able to classify a food with regard to your own intolerance. As a human being, this is easy, but as a machine, i.e. an app, it’s not easy. You have to have the ingredients in a clean data set, which is not at all easy, because either it costs an awful lot of money or you get unclean data that is “community based” for free. You also need to know the compatibility of the ingredients. However, the tolerance depends on the quantity in the food and the amount consumed. A person can assess this intuitively, an app cannot. There are already a few apps on the market that say they can output the tolerances. However, as my tests have shown, these apps are not very reliable.
Basically, I’m not surprised, because the algorithms used are too simple. A functioning algorithm behind such an evaluation is extremely complex and as a developer you have to be very well versed in intolerances and food science. It was not for nothing that we received funding from the state of Tyrol to develop such algorithms. And even then, a residual risk remains and not every intolerance can be reliably issued. Histamine intolerance, for example, remains the classic case here: it is not possible to make a reliable statement about tolerance based on the list of ingredients.
Food scan as an app
We have now managed to correctly evaluate many foods using the algorithms from back then and the data from openfoodfacts. Especially for lactose intolerance and gluten content. Fructose intolerance also works very well, but not yet perfectly. As already mentioned, histamine cannot be seriously evaluated. But: We have the community rating, which already works perfectly with “Ask Ingrid” and provides very precise analyses based on hundreds of thousands of ratings from those affected. And that also applies to scanning. So you can rate food in the “Ask Ingrid” app and help hundreds of thousands of affected people anonymously. And you also get a meaningful analysis based on the ratings of others.
And most importantly, we wanted good results, so we spent a long time tinkering with the algorithms. It is important to us that they work. And we can now claim the same for ourselves, as I have described here in a test in comparison with other apps.
Thank you for your support and loyalty over the last few years. I am very pleased that we finally have a scan that works!
Please subscribe so that we can continue to develop for you. It’s only 8 cents a day!
iOS and Android
The scanner update is currently available for iPhone (iOS). There are still a few problems to solve for the Android version, but this version should also be ready soon. We will inform you immediately here and in the Ask Ingrid app.