Ask HN: Rate my Product : QueryCell
Hi,
I've saved up 6 months worth of money and am taking time off to try and start a software company.
I have just released a new version of my app QueryCell and would love any feedback people have.
QueryCell is an add-in for Microsoft Excel that allows you to query Excel data using SQL, Generate SQL Insert statements, generate test data and mark rows.
http://www.querycell.com/ http://www.querycell.com/welcome.html (welcome is going to be the landing page for ads)
Thanks
(
35 comments
| 45 points
| on HN
)
http://www.querycell.com/welcome.html
http://www.querycell.com
Since "standard" users are not going to learn SQL your market is narrowed to developers only.
You may add a GUI based Query-Builder such as can be found in MS Acess and MSSqlServer and other DB systems and by these you can gain more market share.
Build an easy to use wizard and let the end user choose between SQL mode and GUI mode.
Power users aren't _that_ narrow a user base. There are an enormous number of data miners (analysts, integrators, etc.) out there who work in Excel because they must, or because they're used to it, or because that's the format that their data arrives in, or because it's a nice visual tool that allows them to easily inspect intermediate results---but who ultimately need to run reports across multiple large spreadsheets. SQL is astonishingly easy for such people, and I've watched dozens of data experts learn it from scratch (with help).
I'd beg you not to add wizards to your product, and instead focus on treating power users rather than being all things to all people. (Your syntax highlighting is good, and little red underlines on unparseable SQL would be good.) Trying to train users to use (or avoid using) SQL with wizards is a dangerous business for a piece of software to be in.
This looks incredibly useful. I'll check it out more thoroughly on Monday.
I can see the arguments both for and against wizards and gui query tools.
On the one hand I personally agree with the jcdreads and viggity. These sorts of interfaces are very difficult to get right and dilute the SQL focus.
On the other hand a number of people have suggested going this way, so people do think it would be useful.
I have no immediate plans to go this way but I intend to be responsive to what customers request.
Cheers
Only thing to concern yourself with now is how long before Microsoft implements this in Excel?
P.S. Great to hear an Aussie accent in a demo video!
There is a Perl DBD for this. It doesn't run in Excel, but you can run SQL queries on spreadsheet files.
Does QueryCell work when excel is being run on a non-windows OS via wine? ( I assume not, but if it does, I know more people who might be interested ).
--
Also, the video cut off for me in mid-sentence. I think you are missing the end. For me, it ended at with "if you would like to try query cell for free, feel free to download".
@smicallef
Yes the idea of Microsoft proving a similar functionality is scary, but I feel a little bit like "if they were going to do it they would have done it", also they have access to protect. But maybe I am in for an unpleasant suprise, certainly a risk.
@tzury
Yes there is an interesting marketing proposition here in my opinion. I can either market to heavy Excel users who don't use SQL yet or market to SQL users who occasionally find themselves using Excel. It's an interesting one. At the moment I am leaning more towards SQL users, they can immediately see the possibilities. hmmm.
I could see a lot of one-off usage (for occasional Excel users) and a few requests for a refund under the 30-day guarantee.
That said, it looks like it'd make the job a lot easier. The first thing I do when I get a big excel 'database' is convert it to something else!
In other words: I build a spreadsheet with QueryCell, but I want people who don't have QueryCell to be able to use it.
So I hit "Compile", and the program generates standard formulas using hlookup/vlookup and/or VBA so that the spreadsheet works for those who don't have the Plugin. (Of course it adds a link to "Built with QueryCell" :)
I don't quite follow though. To use querycell you load a normal Excel file into Excel. You can then open the QueryCell pane to query the regions of data in the open worksheet/workbooks.
The result sets returned from queries are then added to a new 'resultset' worksheet.
So as it stands you can pass around spreadsheets you have used with QueryCell to people without QueryCell.
Let me know if I misunderstand.
Cheers
It sounds like what you're doing is generating a new worksheet and populating it with (static) data generated by running the user's SQL query.
This is fine for sharing static data but b/c you're apparently not using excel formulas in generating the 'resultset' worksheet (it's just raw data, again) the 'resultset' sheets aren't going to be auto-updating if the underlying data got changed (on the 'sourcedata' worksheet, if you follow).
If you can compile SQL queries into excel formulas that'll stay updated as data change this goes from an A to A+ produce.
You should start by looking at the documentation for Vlookup and Hlookup, as those are the built-in functions people use to (fake) selection.
I now understand what you is being suggested.
Yes it's true, that is not how QueryCell currently works.
I will have a think about this idea. My first impression is that it would be quite complicated to implement and get right but very powerful if done well.
Thanks Guys.
What do you think about this : http://www.querycell.com/captionMe.html
I need help with my Comic Ad.
Not sure how hard it is to create a Mac OS X version as well, but that would probably not be the initial target audience.
sorry, I prefer to like start up ideas but this one just doesn't seem to have any legs in my mind.
I would say in defense that:
- The return time doesn't scale with the number of rows returned. I have just run a query that queried 6 tables, invoked numerous string concatenations, included 2 MOD functions, 2 RAND functions and returned 22,000 rows, it took 11 seconds. This is not up to database performance of course.
- There is a static overhead to running a query, and that's much greater in QueryCell than in a database directly.
- QueryCell is not meant to replace databases, just make life easier for people working in Excel.
It was pretty much the 'countless hours trying to use formulas' that made me think about QueryCell.
Cheers
Congrats on the bold move. I am in the same boat -- just quit my job last Friday and will be working on my startup. I wish you the best of luck!
Alex
When you're generating the table names, I think naming the table after the sheet it's on might be more intuitive than a single letter, especially when there's only one data range per sheet.
The table names can be changed, you could click on the default letter name, 'A' for example, and rename it to SALES_NUMBERS for example.
I see the sense in using the worksheet name although there may be complications if there were multiple data regions on the same sheet or if the sheet name was long, as there isn't a huge amount of screen real-estate for the regions grid.
Thanks for the food for thought.
1. Perhaps you can let define the tables in the FROM clause like: FROM workbook1.xls~sheet2~A1:D4 AS table1
2. Add an option to autohide the not selected rows instead of coloring the selected rows, this will give a condensed result set that the user can maybe copy and paste to another sheet.
3. Don't worry about Microsoft competing - they might offer to buy you out.
keep focused! Good luck!
I don't work with Excel often enough to purchase but will keep in the back of my mind in case I have a future need.
Marketing advice: invest in a copy writer and voice actor and reproduce your video.
You might want to get a professional PSD designer to sharpen the overall look. It has a template like look for design, though for this concept design might not be as important as selling it's functionality.
Good luck!
Cheers