Sitting in a Las Vegas bar and talking to Old MacDonald about his Farm would be an ideal way to start a newspaper article about Farms in the 1970s. We’d probably talk about the latest tractors and what amazing mechanical gadget can be attached to them:

We are living in 2014 and not 1970 and we know that the newspaper industry is in decline so maybe we should start a webex with Old Macdonald and discuss the use of technology on his farm. He’d probably tell me that his latest tractor has internet for the weather, speed, direction and coverage stats. He’d probably also tell me that he can launch the ploughing job automatically and monitor it from his iPhone.

Strangely, Old Macdonald would recognise the same situation with conversion farms. In the 1970s it was all tapes, human operators and mechanical connections. Today’s conversion farms are all files, APIs, watch folders and data centres. Simply launching jobs and returning status is no longer enough. To get a good view of the efficiency of the farm we need to be able to mine the performance data and run some analytics. At NAB 2014 AmberFin showed a web interface with a selection of analytics to get feedback from our customers on the operational parameters that they wanted to see.

It was interesting that when we talked about the control of the farm and launching jobs, most customers had similar views about APIs and watch folders. The feedback and analytics, however generated a much more diverse set of responses. This is probably because our customers represent a very diverse set of businesses and business needs. Simply measuring the throughput and pass/fail does not give an accurate reflection of how well the farm is doing. In a very non-scientific survey we asked a number of customers if they measured the cost of rejects. After counting all of our non-scientific responses we came up with the answer … zero.

This is strange because some rejects are cheap to fix (e.g. a network error) and some rejects are very expensive to fix (e.g. bad caption file needs re-authoring) yet simply counting pass/fail cannot help you with these analytics. The only way to get this information is to have a good supply of metadata such as that stored in a good MAM. Processing that metadata and using it with a good web based analytics monitor.

Knowing that this combination of great metadata and great transcoding can generate real business value makes me very excited about the great, innovative benefits the partnership of Dalet’s MAM and AmberFin’s transcoding can bring. We’ve been a single company for a little over a week. You’ll have to wait a few more until these great ideas ripple through.