This way, your tabular / panel business data, their meta data (like column names), and analytic results related to these data remain confidential and only readable by yourself. Though, we and other analytics suppliers are able to derive meaningful patterns and forecasts from these masked data by means of statistics and machine learning. After you translating it back by your key and our script into clear data likely lets you discover new business insights.
In some exceptional cases, it might be helpful that you reveal some of the metadata. For instance, when it comes to the prediction of credit card fraud it would be useful to disclose the column names of customer IDs and credit card IDs to the analyst for easing business reasoning, but it doesn’t need their column contents decrypted. I this example, all other meta data would stay encrypted, so that column names remain cryptic to everybody except yourself.
Feel free to let your analytics suppliers enter in competition – e.g. by the Kaggle analytics portal – and just chose the best result. We can help you in setting up such competition.