But I agree with what Ben had to say. After using various products for a while, I’ve concluded that the really “interesting” visualizations can’t be bought. They have to be built. A simple bar, line, or pie chart can now be done easily out of the box with something like fusion charts. But if you want to get into what I consider “premium visualizations…” you gotta build it. The good stuff is specific to your data – no generic tool will do.

Fortunately things like Flex/Flex Charting are finally giving us the tools to do just that.

As long as there is interesting data there will be interesting, custom, views that need to be built.

Much like other emerging technologies, I feel data viz is ahead of the wave and clear defined business benefits/measures are yet to be available and understood across many business domains.

Without measurable benefits, and I’m not saying they don’t exist, just that they are not clearly articulated to the business audience yet, service firms will require a “leap of faith” during the sales process.

If you look at Stamen as an example, they have a great niche but the companies are still other visionary startup style companies.

That said, the skills involved in handling terabytes of data, data mining that set and representing it for consumption and interaction using data viz are rare and demand of them is only going to increase over the next couple of years.

I do agree that FusionCharts aren’t especially elegant, but it is successful, which says something.

There are at least three hurdles that we typically see:
1) As noted by Peltier, implementation by someone who doesn’t understand infoviz fundamentals can end up eye-gougingly ugly and useless.
2) Data integration is non-trivial. If you are connecting to a public and well structured API, creating cool visualization mashups can be easy. Pulling data from proprietary enterprise systems is a whole different ball-o-wax.
3. None of this is valuable unless there is a lot of thought put into how the advanced visualization can connect to existing work processes and the decisions users need to make.

ILOG Elixir (http://www.ilog.com/products/ilogelixir/) is another example of a set of visualization components. At Juice, we’ve got a similar toolkit of Flex-based components which are a bit more, um, Tufte-compliant.

At one end of “premium” is the once-off “artisan” job. An example of this would be if I commissioned Nicholas Felton to build something beautiful and unique using my own personal data. At this end, you have to be good enough to charge $$$ for each job, or else your $ per hour of effort is low. So I don’t think this is a market for a software product.

At the other end of “premium” is the software approach. It scales better as you sell licenses so your $ per hour of effort can be high. However, at this point you have to compete with free software and commercial software like you showed (and stuff like dotnetcharting.com) However, it is still hard for the average person to make something like your walmart map using those tools. If you could make it easier then you might be on a winner, but it’s a tough ask.

For example, I could just plug my personal stats into many eyes, but it just doesn’t excite me because it is all to vanilla. So applying the above idea would require something that let me put my stats in (assume I am not a programmer) and produced something nicer than many-eyes, but it still wouldn’t be as good as a custom feltron original.

Funny about that… and here’s a little scoop for flowing data:

Look at: http://daytum.com

(@Scott – your wish came true?!?!)

Another way to look at it is from the horizontal vs. vertical perspective. The software you listed as examples are basically horizontal. That is, tools that you can apply to various specific viz tasks. I wouldn’t try to compete with free tools or BigSoftwareCompanies in that space.

There’s probably a market for vertical visualization however. Pick a specific subject or business and make the best visualizations for that space, say car dealership for example (http://www.dealerdiagnostics.com). Take all the tech-pain out of it. Sell that as software. You can probably find a well paid niche to avoid competition, but you’ll need the expertise and experience in that vertical to be able to imagineer something great.

That’s really what daytum is doing. A “personal stats” vertical using horizontal software (in this case google charts api).

Fun conversation here Nathan!

Another possible market can be found in Nicholas Felton’s Personal Annual Reports –

http://www.feltron.com/index.p.....al_report/

If users could easily plug in personal attention data each week and at the end of the year have a professional and attractive “annual report” produced, I could see such a service be quite successful.

The real question, is who can put these lame 3D effects aside and introduce a package for the masses that enables effective data presentation. I’m thinking of a broader implementation of something like BonaVista MicroCharts, not a 4D version of FusionCharts.