Data Quality at the Adobe Summit

This year, I had the opportunity to run a hands-on lab about data quality at the Adobe Summit in Las Vegas, and I’ll run the same lab again in London on the 10th.

The lab was mainly about the test framework I also write about here a lot. My hope was that I could introduce the framework, then get people to experiment with it, and ideally return to their offices with a test description file for their very own web site.

Moon shot, right?

I was very surprised that in an audience of more than 80 people, only 5 said they had a data layer on their web site. I was clearly projecting an image of the maturity of our colleagues in the US that was too bright!

After some great feedback from the first session, I changed the second one very slightly, and I thought it went very well. Noone, I believe, went home with a test description, though.

Now here’s the thing: how do I run this in the UK?

More of the same

Given the outcome, I was very surprised to see that both my sessions got very good marks. My first one saw 53.7% of the attendees fill in a survey. And they didn’t smite me, no, they still gave me an above average score: 4.19 for the lab, 4.22 for me, out of 5. The second installment was even better received: 4.56 for the lab, 4.72 for me.

So do I do the same session?

Do I go in, check everyone knows what they’re in for, then plant some seeds?

That’s what it is, right? Planting seeds. Making people aware that something can be done, should be done, then letting them walk away, mull it over, and eventually some of them will water the seed, help it grow, and eventually reap the rewards.

(Sorry, I’m reading Iain M. Banks “Culture” books, and I reached the last one, which is absolutely full of very beautiful, stylish, and elaborate language. Rubs off, even when I cannot claim to even come close to the late Mr Banks.)

Radically different?

I wouldn’t have time to redo my lab book, to be honest. But could I, even with the lab book as it currently is, make the lab radically different? Turn it around so people can actually come away with some tangible result?

Having thought about that since Las Vegas, I have to admit that I don’t quite know how I would do that.

Would it work if I sat in the front, asked people to give me web sites, then build tests based on those sites?

Wouldn’t that be exceedingly boring for everyone but the person in question?

And wouldn’t I run the risk of turning a relatively technical session into extreme nitpicking?

Somehow I don’t like it.

I had a couple of other ideas, but they where so bad, I won’t even publicly dismiss them here.

Every little helps

So my conclusion is that I will not change the session, but change the tools.

Inspired by and with a ton of input from my friend Jim Gordon (of Tagtician and DTM Cheat Sheet fame), I wrote a web front-end for the test description, i.e. a page that lets you assemble a (simple!) test description without having to edit a JSON file directly.

My hope is that this tool will be enough to get people a bit further than those who came to my sessions in the US, and that it also helps those who shy away from opening a JSON file in an editor.

As a side effect, it should make the lives of my lab assistants substantially easier, too.

I look forward to seeing some of you in London, and I hope some of you will walk away with a usable test description this time.

For those who have been in my sessions in the US: try the rule builder! And don’t forget that you can download the lab books, both the US format (letter) and in A4 for the rest of the world.

What do you think?

One thought on “Data Quality at the Adobe Summit

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.