Hacker News Reader

Comment by revorad | Hacker News Reader

Of course, Excel is the default. However, I'm aiming to solve use cases which are either really hard or not possible to do in Excel or can't be done very well in Excel:

Making multiple panel graphs, boxplots, plots with error bars, heatmaps etc. Very useful for quickly summarising large amounts of data.

In research, a lot of people spend a lot of time adjusting figure margins and title sizes etc. I want to save them time by giving them something that just works.

Applying the same graph theme or formatting to a batch of data files. I'm trying to bring the power of R scripting to the masses who do not have any interest in learning programming.

A lot of people, such as PR and marketing firms, specifically don't want to use Excel to make charts. They use nicer looking alternatives (mostly Flash based solutions). I am making a simple web-version for them.

Those are some examples, there are more.

View on HN

Our experience (at Timetric) is that selling visualization as a direct benefit is a real challenge. The people who know what a boxplot is are comfortable with Origin or R or whatever: everything else either doesn't care, or doesn't know they should care, or doesn't have the budget to care. What'll be critical for you is communicating a value proposition which gets past that. How does using PrettyGraph make you money?

The Verifiable deadpool post is an eye-opener, too, as to the dynamics of this market: http://stuartroseman.com/post/619953720/out-with-the-old-bus....

Thanks Andrew. In my limited experience in research, a lot of the most brilliant people were not interested in looking at anything to do with code (such as R). Others who didn't mind would rather spend their time doing their core research than programming, but they had no choice.

I am trying to build something which can save such people time.

Outside research, I have been contacted by a few different business people, who said they would love to use something like my product. There are some alternatives like Tableau Public out there (which is admittedly more than just a graph maker) but costs a lot.

I did read the Verifiable post but the key point there (as the name suggests) is that data quality killed them. They were trying to be a databank. Pretty Graph is not. It's just a website, where you put in your data and get graphs out.

Anyway, I'm sure I could learn more from your experience at Timetric. Are you going to the HN meetup this week? I'd love to have a chat.

In that case you're in a head-on battle with the Google Visualization API and a couple of other players - iCharts (http://www.icharts.net/), Wondergraphs etc.

There's definitely something in this space, though.

Google visualization API is not for end users, but I suppose there could be services built on top of that (although I haven't seen any yet). Thanks for pointing to iCharts and Wondergraphs. I hadn't seen them before.

Google Spreadsheets (and the stuff coming down the pike like Google Fusion Tables) is, though. Really, you're competing with Excel (plus custom styles/plugins) and with Google Docs, I reckon.

Only few of your examples show off these features. I can replicate the rainfall and sales graphs in Excel in less time than it takes me to log in to your service. The heat map on your blog on the other hand looks great - why not replace the boring graphs with this one?

Yes you are absolutely right. This is where I get to use my "I'm a poor overworked single founder" trump card :-P

Jokes apart, I agree I have a ton of work to do to make the messaging on my site clear. I just put it out today to get some kicks in my butt.

Reply on news.ycombinator.com