среда, 7 января 2015 г.

Milk With Crossed Barcodes Is Not Stupid, It's Marketing Genius

This good old case of milk cartons with crossed out barcodes (one, two and google for "milk with crossed barcode" for more)... A dairy company located near Moscow, Russian Federation crosses barcodes on its products with two diagonal red lines to counter "number of the beast", hereinafter referred to as "bad number", which it claims is present in every barcode...

People hear it and are immediately pissed off. They accuse Russia of leaving in Dark Ages and company owners being lunatics.

First things first. Is there any "bad number" in the barcode?

Milk cartons carry EAN-13 barcodes - plain boring barcodes used for labeling groceries. Evil Wikipedia (pun intended) to the resque! Every such barcode consists of a number of vertical bars and there're three groups of bars that are slightly longer than the rest of the bars. Those groups are designed to serve as alignment markers to help barcode scanners proper identify where the barcode is and they don't carry any data in them.

It just so happens that those alignment markers have bar width identical to how number 6 can be encoded (but this is not the only choice). The claim is therefore because each barcode contains three such markers those are three sixes and that's the "bad number".

Well, not so fast. Read Evil Wikipedia (pun intended). "Six" can be encoded various ways depending on where it is located in the barcode. There're three ways to encode each of the ten digits and they are called encoding schemes. Different barcodes will use different schemes depending on what the very first digit is.

Since the dairy company in question is located in Russia the first digit will be "four" and so according to Untrustworthy Evil Wikipedia (pun intended) the barcode should use "LGLLGG RRRRRR" encoding scheme which means that some digits are encoded using scheme L, some are encoded scheme G and some are encoded using scheme R.

Now schemes G and R indeed encode "six" with two closely located narrow lines but scheme L doesn't do so - it uses a very wide line combined with a narrow line to encode "six" instead. If a scanner finds those two narrow lines where a scheme L must be used it will produce a read error - those two lines will not be treated as "six" because a "six" represented this way at these positions is impossible.

So just looking at the bar is not enough to say if it is a "six". Position also matters. How about the alignment markers? Which scheme is used to decode them?

None.

Just in case you're not good at reading, I'll repeat - NONE. The alignment markers are not to be decoded and so they don't have anything encoded in them. There're no digits in them and so there're no "sixes" in them and so there's no guaranteed "bad number" in every barcode.

Done with this. Let's proceed to the marketing side.

No matter what analysis you carry out many people are so stupid they will be unable to understand it. They will trust anything if it comes from a source they wish to trust. If that source claims there's a "bad number" in EAN-13 barcodes - they will trust it.

So, suppose you're a dairy company and for whatever reason you want to show that you're uncomfortable with EAN-13 barcodes because of "bad number" in them. What can you do?

Let me think... You don't like X and X is usually printed on milk cartons and you produce milk cartons...

How about not printing the barcodes? No barcode - no "bad number" and everyone is happy.

No, then you're not happy because you cannot sell you milk anywhere - retailers won't accept goods that cannot be scanned fast on the counter.

So what do you do? Ditch the idea?

No, you'll cross the barcode and claim that the "bad number" is now neutralized. Just be careful - you need the lines to not interfere with the scanning process otherwise no retailer will buy the product.

So you carefully design a crossed out barcode and print it. You claim the "bad number" is neutralized and everyone is happy. Pure marketing genius. Should have been nominated for Red Dot Communication Design Award.

Is the "bad number" neutralized by such crossing? It is likely not - the barcode was designed for scanners and scanners scan it in their evil scanner manner (pun intended). If they still can read the barcode it means they still see it reliably and can still read all the evil stuff presented in there.

There's no "bad number" in there. If it was there then crossing the barcode would either have not neutralized it or it would have damaged the barcode and make it useless for retail.

Is the world approaching its end? Yes. Is that because of barcodes? No, it's because so many people are really stupid.

1 комментарий: