The revolution triggered by BIG DATA is about to hit the ferroalloy industry and things will never be the same. The actual transformation of the industry began several years ago when relationship buying was replaced by price buying. No longer was a long-term relationship between and buyer and seller was valued. It was just: show me the cash.
This still persists to this day, but the heavy analytic tools are still lacking. I recently reviewed the offers from a recent RFQ. I quickly calculated that the bids the steel mill received had a minimum of 150 options. There were: fixed or index prices; a combination of fixed and indexed; the ability to pick the index provider(s); the delivery and pricing periods; caps and floors, delivery terms (delivered, c.i.f., f.o.b.), and various financing options far beyond the 2 net 30 of the bygone days.
To really figure out the best bid to take requires more than a spreadsheet. To be really effective, the mill has to plot each option against the other options and give weights to the various options, i.e., is price the highest value option?How much weight each of the factors should be determined by examining all the past data running the same calculations.
The mill should also determine if the seller exhibits any tendencies—late shipments, poor payments—and give a weight to every one of those.
As the data expands from the present to the past with digitation of history, the number of data points increases exponentially. In the end, you have BIG DATA solely on objective terms. And, this doesn’t even include data that correlates directly with the markets and the examination of the ultimate buyers of the mills’ steel.
I can easily envision that awarding the RFQ would run through millions of data points.
In the end, the buyer ends up as an IT guy who then passes the information on to someone authorized to say yes. It could be buying walnuts or ferrochrome.