Twitalyzer 2.0 – A Big Step Forward

As you probably know, there are quite a few tools out there that attempt to "score" Twitter users. I've looked at most of them, and I have yet to find one that does everything. But the one that's the most flexible, customizable and useful to me as a micro-blogger is Twitalyzer 2.0.

Twitalyzer is the brainchild of Eric T. Peterson (@erictpeterson), a noted web analytics expert and author of Web Analytics Demystified: A Marketer's Guide to Understanding How Your Web Site Affects Your Business. Eric brings a passion for analytics and an understanding of the need for actionable metrics and reports to the Twitter scoring arena, something I haven't seen in any other tool.

What's new in 2.0? Quite a bit. There are more metrics, a 51-page handbook, tools for segmentation of users, benchmarks, goals, sentiment analysis, and, of course, more of the flexible dashboards and reporting that set Twitalyzer 1.0 apart from the other Twitter scoring tools. I counted 15 separate reports, and I probably missed some. You can plot trends for 22 separate metrics over time.

The two things I liked the most about Twitalyzer 1.0 were:

1. All of the metrics were defined. You could see what was being counted and what those counts meant.
2. There were clear recommendations on how to improve your scores.

Twitalyzer 2.0 has kept that. There are many more metrics, but they are still all defined. And the recommendations are still there, along with a new "Goals" report that allows you to set goals and track your progress towards them.

But in my view, the most important new feature of Twitalyzer 2.0 is the Segmentation / Tagging functionality. I'm still learning how to use this, but the examples in the handbook are very well written, and it's clearly a vital part of any analytics tool set.

How does Twitalyzer compare with the other Twitter scoring tools? There are two others I've used in depth, TwitterGrader and Klout. TwitterGrader reports only a single score, and there is no definition of how that score is derived or what actions one should take to improve it. Klout has a few reports, a number of metrics and recommendations for how to improve them, but the Klout reports seem to be full of old data, and it can take hours for them to update your results. And I didn't see anything like Twitalyzer's segmentation capability.

There are a few things that could be improved.

1. Location: Twitalyzer maintains separate lists for all "spellings" of a locality. For example, there are separate lists for "Portland, OR", "Portland, Oregon" and "Portland, Oregon, USA". Twitalyzer isn't the only tool that suffers from this - TwitterGrader does too, and many tools don't do location-based analytics at all. But it would be fairly easy to combine most of the spellings and misspellings of a given metropolitan area like Portland / Vancouver into a single location, using a combination of Twitter Search and the Google Maps Geocoding API.
2. CSV export of metrics time series: Twitalyzer can export a single time series to CSV format now in the "Trends" menu. But there are 22 or so metrics; a combined CSV file of all of them would be very useful, especially for someone like me who wants to correlate Twitter metrics with other metrics, campaigns, events, and so on.
3. I'd like to be able to integrate Twitalyzer data with the Clicky web analytics tools. There is Google Analytics integration now, but I'm not sure I'm going to stay with Google Analytics, even though it's free and an "industry standard." Clicky is real-time; Google Analytics isn't.