Most of us have grown up with the cartoon image of Donald Duck's uncle, the super-wealthy Scrooge McDuck, taking a greedy dive into his enormous "pool" of hoarded gold. It's a pitiful picture because we know that he could benefit so much more from his wealth if he would only use it for more than a hidden treasure.

Now imagine that hoarded gold as existing translation memory (TM) data and Uncle Scrooge as a translation buyer. The similarity is striking: translation buyers use and re-use the enormous wealth of data that they have amassed for their own translation processes instead of putting it to work for them in so many more productive and beneficial ways.

According to the LISA 2004 Translation Memory Survey, companies own an average of more than five million translation units after five or more years of using TM technology. Depending on how well they were able to source their translation vendors and use technology, they will have paid somewhere between 50 cents and $1.50 for each translation unit. This makes the cost for the average TM database of a long-term TM user between $2.5 million and $7.5 million. And while this money was paid for the sake of translated products rather than a database, the TM database is likely the asset with the longest shelf life -- in fact, the only asset with a true shelf life. Software products, cars, medical devices and computers with their translated materials and/or interfaces will be sold, used and replaced —

but the data stays
and grows
and stays
and grows . . . .

Exactly how many other assets of this magnitude lie dormant in the typical corporate treasure vault? Not many, if any, especially among assets whose value is well known. After all, an understanding of its value is what keeps this data in the vault. Does the Uncle Scrooge reference become clearer now?