Data Quality and the Web

Avinash publishes another blog article regarding data quality.

Its a good article, nicely written but I have to agree with Mickey

"This is a nice sentiment but nothing new in the BI area - another example of web analytics not learning from “old school” analytics.

Although operational databases are usually very accurate, of the 100’s of customer databases I’ve data-mined over the past ten years, I cant think of many that were over 95% accurate, with customers data being ommitted, duplicated, mis-inputted etc.

The sentiment is right, go for precision in the data that helps you make your decision"

I get rather frustrated with web analytics going over ground that we have previously covered. Aviniash claims to come from the BI world. Unless he was working off an operational dataset, dealing with problems of accuracy should have been his bread and butter and therefore this should be nothing new for him, or for anyone else in the industry. If it is, I'll again quote Mickey,

"but will this be a cultural change for most marketers? It shouldn’t be. If it is news for you, either your tech people have been lying to you about you data quality for years or you havent been collecting much data about the customer"

Linking to a new blog

Just a quick post linking to a blogger who's rapidly becoming one of my favs in this area, the Ad Contrarian.

And here's a good recent article

Enjoy!

Tracking twitter and using other data to enhance your web analytics

In recent posts 1, 2 & 3 2 of my colleagues have gone into great detail about tracking twitter.

The response to these posts and presentation of this material at our Omniture summit was apparently overwhelming - everyone wanted to know about how they could track and use twitter.

I felt these posts were fantastic at conveying how one could go about pulling this twitter data into omniture and making use of it. What occurred to me was that peoples interest was not some much about the fact that it could be pulled into SiteCatalyst, but that the data could be got at at all!

It strikes me that people didn't realise that APIs could be used to get this kind of data - hopefully Adam and Ben have shown getting the data is not only possible but a very good opportunity.

Once the data is in Omniture, it's not tied to any other behavioural activity, it really is just a trend of data over time for a given segment, keyword or set of keywords. This could just as easily be pulled out of the API and into your own database as it would be uploaded to Site Catalyst.

In order to "analyse" this twitter data you would want to compare trends side by side and see if you can determine any kind of causality between you online activities and the various "tweets" about your company (or vice versa).

Personally, I would love to do a proof-of-concept to run a "text mining" tool over the tweets and see whether they could be segmented in someway to provide more meaningful trends than just a general indication of volume. I think the text mining software may have some difficulty with the online vocabulary (lexicon) that might need to be used along with the varieties of grammar that we see online, but it would certainly be interesting.

However, my main point is that many different data sources are online now that can be incorporated into analysing the customer. Because other tools make that data available you should be looking to take advantage of that somehow. Over the coming weeks I will be attempting to pull together a list of the online data sources I would consider to be most useful to analysing customer behaviour or at least comparing trends.

Remember, unless you can map the data from these tools directly onto your own customers you are always going to be speculating about their causality. Once I get this "ideal" list of data sources, I'll then look at ways of joining the data in that list to your customer data.