Do you think most organizations consider the data pyramid in their business intelligence strategies?
Sort by:
I used to be at a company where they never deleted anything. They had all the data they ever created and none of it was particularly useful. There was a lack of data integrity, lack of ownership and governance around it. No data warehouse. Point solutions on reporting all over the place. I think I counted 7 or 8 different reporting solutions. It's that age-old question: you've got data, but is it information? Is it actionable? Is it useful?
Part of my background is manufacturing, and in that space, there’s the operation technology side which grabs data from all the production machinery. Those are small, tiny bits of data, but without added context, they’re meaningless. Without information about time, place, which machine it came from, what product was being produced, etc., it's useless. The same is true in all the business information that we have. Unless you have real context related to that business information it's useless.
To some extent, but there isn’t a single blueprint and this becomes an area of focus that is deprioritized in the competing list of revenue drivers when trying to build and grow a company; until you try to scale and realize the data challenges to have piled up along the way. I think what we're creating isn't any more sophisticated than any other type of data cloud but one that's built for a specific marketplace and a very unique business problem that we're trying to solve for in the connected operations space.
I came to Stanford to implement Salesforce, which was the largest implementation ever at that time, with a quarter million users on Salesforce. I actually saw papers that were stored, not digitized. But we needed the history of those stored papers. So we worked through the massive task of scanning those, linking them with the right Salesforce records and so on. Then that data became information but it stopped there. It did not go further up the data pyramid.
Now I’m at a startup that's been in business for about 20 years where data's never been deleted. There were point-to-point connections and then we went IPO—that was a real nightmare. In a very short cycle I had to exclude those point-to-point connections and put a data bus in between. The data governance learnings from Stanford really came into play in terms of how to monetize that data and therefore harvest it. So I did not stop at information and tried to elevate it to wisdom. There's the point-to-point connections, APIs—everyone says it doesn't work like that. But that's really the challenge.