Photo via Flickr (Creative Commons License) by: ella_marie
Today is Thanksgiving Day, which is a United States holiday with a long and varied history. The most consistent themes remain family and friends gathering together to share a large meal and express their gratitude.
This is the fourth entry in my ongoing series for expressing my gratitude to my readers for their truly commendable comments on my blog posts. Receiving comments is the most rewarding aspect of my blogging experience. Although I am truly grateful to all of my readers, I am most grateful to my commenting readers.
“Data quality issues probably occur on some scale in most companies every day. As long as you qualify what is and isn't a data quality issue, this gets back to what the company thinks is an acceptable level of data quality.
I've always advocated aggregating data quality scores to form business metrics. For example, what data quality metrics would you combine to ensure that customers can always be contacted in case of an upgrade, recall or new product offering? If you track the aggregation, it gives you more of a business feel.”
“Back when I was with the phone company I was (by default) the guardian of the definition of a 'Customer'. Basically I think they asked for volunteers to step forward and I was busy tying my shoelace when the other 11,000 people in the company as one entity took a large step backwards.
I found that the best way to get a definition of a customer was to lock the relevant stakeholders in a room and keep asking 'What' and 'Why'.
My 'data modeling' methodology was simple. Find out what the things were that were important to the business operation, define each thing in English without a reference to itself, and then we played the 'Yes/No Game Show' to figure out how that entity linked to other things and what the attributes of that thing were.
Much to IT's confusion, I insisted that the definition needed to be a living thing, not carved in two stone tablets we'd lug down from on top of the mountain.
However, because of the approach that had been taken we found that when new requirements were raised (27 from one stakeholder), the model accommodated all of them either through an expansion of a description or the addition of a piece of reference data to part of the model.
Fast-forward a few months from the modeling exercise. I was asked by IT to demo the model to a newly acquired subsidiary. It was a significantly different business. I played the 'Yes/No Game Show' with them for a day. The model fitted their needs with just a minor tweak.
The IT team from the subsidiary wanted to know how had I gone about normalizing the data to come up with the model, which is kind of like cutting up a perfectly good apple pie to find out how what an apple is and how to make pastry.
What I found about the 'Yes/No Game Show' approach was that it made people open up their thinking a bit, but it took some discipline and perseverance on my part to keep asking what and why. Luckily, having spent most of the previous few years trying to get these people to think seriously about data quality they already thought I was a moron so they were accommodating to me.
A key learning for me out of the whole thing is that, even if you are doing a data management exercise for a part of a larger business, you need to approach it in a way that can be evolved and continuously improved to ensure quality across the entire organization.
Also, it highlighted the fallacy of assuming that a company can only have one kind of customer.”
“I recently attended a conference and sat in on a panel that discussed some of the future trends, such as cloud computing. It was a great discussion, highly polarized, and as I came home I thought about how far we've come as a profession but more importantly, how much more there is to do.
The reality is that the world is changing, the volumes of data held by businesses are immense and growing exponentially, our desire for new forms of information delivery insatiable, and the opportunities for innovation boundless.
I really believe we're not innovating as an industry anything like we should be. The cloud, as an example, offers massive opportunities for a range of data quality services but I've certainly not read anything in the media or press that indicates someone is capitalizing on this.
There are a few recent data quality technology innovations which have caught my eye, but I also think there is so much more vendors should be doing.
On the personal side of the profession, I think online education is where we're headed. The concept of localized training is now being replaced by online learning. With the Internet you can now train people on every continent, so why aren't more people going down this route?
I find it incredibly ironic when I speak to data quality specialists who admit that 'they don't have the first clue about all this social media stuff.' This is the next generation of information management, it's here right now, they should be embracing it. I think if you're a 'guru' author, trainer or consultant you need to think of new ways to engage with your clients/trainees using the tools available.
What worries me is that the growth of information doesn't match the maturity and growth of our profession. For example, we really need more people who can articulate the value of what we can offer.
Ted Friedman made a great point on Twitter recently when he talked about how people should stop moaning about executives that 'don't get it' and instead focus on improving ways to demonstrate the value of data quality improvement.
Just because we've come a long way doesn't mean we know it all, there is still a hell of a long way to go.”
Thanks for giving your comments
Thank you very much for giving your comments and sharing your perspectives with our collablogaunity. Since there have been so many commendable comments, please don't be offended if your commendable comment hasn't been featured yet.
Please keep on commenting and stay tuned for future entries in the series.