Ask HN: Excel experts – Formula IDE in Excel (MVP screenshot) – is it useful?
Excel add-in and would love to hear your opinion.
The aim is to let Excel users write formulas in simple IDE inside Excel. Instead of editing each of them in a formula bar you could just edit all formulas in one file (or for more advanced projects in multiple files).
What are the advantages? You could save your formulas to a separate file and import them through the add-in. This lets you easily share your formulas with others and add version control and code review to your project.
Checkout this awesome MVP I've just built in Paint: http://i.imgur.com/yBaFArn.png
Questions I was inspired by:
Ask HN: Language compiled to Excel spreadsheets? https://news.ycombinator.com/item?id=9941086
Ask HN: What are the biggest problems with spreadsheets? https://news.ycombinator.com/item?id=9935357
14 comments
[ 1.7 ms ] story [ 39.5 ms ] thread1. Anyone who can get an add-in pasted corporate software security can also use a better tool outside Excel. In other words: think carefully about the customer. Users don't buy plugins in larger businesses, etc.
2. It looks pretty useful and would help users create a bit more structure in their sheets and calculations.
3. My idea of why terrible Excelsheets seem to be the norm, is that 'build once-use once' does not exist in business. Too often your boss wants you to redo your calculation and after a few iterations, time not spent documenting / cleaning catches up with you in Excel.
Often you also build one workflow and apply it to many not-quite the same situations. Tweak a few cells here and there. Hence custom workflows per worksheet and again, broken (re-)usability.
Your tool would help to start with a more long term perspective, and help in iterating on a sheet. Might even be faster to type, than to click-point and type.
But my baseline response would be near number 1. Plugins from Excel to R and Python exist too, but still all the sheets I see internally, and code even professional consultants come pitching is in VBA or some proprietary format. The Venn-diagram between programming and Excel has no overlap imho.
What other tools are available? A lot of technical people don't choose Excel, but come into situations where Excel has been relied upon historically and the non-technical people do not want to move away from it. So I could see such a plugin having value in these situations (non-technical people continue to use the spreadsheet as-is, while the programmer improves it via the plugin), which I would imagine are quite common (I have been in this situation before).
> Users don't buy plugins in larger businesses, etc.
Fact or opinion?
> What other tools are available? A lot of technical people don't choose Excel, but come into situations where Excel has been relied upon historically and the non-technical people do not want to move away from it. So I could see such a plugin having value in these situations (non-technical people continue to use the spreadsheet as-is, while the programmer improves it via the plugin), which I would imagine are quite common (I have been in this situation before).
In my enviroment (slow finance, seen a few firms, second line hence the reviewing of many a workhorse). The IT-department is pretty Microsoft-captured. The quants are pretty much do-as-they-want (Java, .NET, VBA). Whatever they know and can get their hands on. Then the financials in accounting. Excel is king there and is pretty much the workhorse surrounding the customer, financial and management accounting systems. My main point here is that I have yet to find a firm where the use of tools is not historic, and where those totally different skillsets / user sets work together on tooling.
The use case you describe (non-technical people continue to use the spreadsheet as-is, while the programmer improves it via the plugin) fails for me (or rather: is an exceptional situation), since the big value of Excel is that the non-technical persons builds his own model. I know of one bank that has started working with teams with explicitly different skillsets (so IT, business and finance always in one team). In that case, tooling like this would rock.
>> Users don't buy plugins in larger businesses, etc.
>Fact or opinion?
I would say this is a fact. It's the theme that's pretty common that B2B-sales is a different park than B2C-sales. B2B-sales in larger firms goes via a sourcing department, that possibly has preferred suppliers, etc. So just being a cool tool, still gets you installed on 0 corporate pc's that are basicly in lockdown.
>Fact or opinion?
Users definitely buy plugins, but these bring significant functionality and often tie-in to additional services. Financial data being the most familiar to me. These plugins, while wrapped in VBA, are often in C++ (again, finance world) for robustness and performance, and try to be as agnostic to whatever nutty layout or logic a user chooses to have for their spreadsheet.
So, agree, with the pragmatic assertion that technical people love Excel as a UI, but they're probably not necessarily developer-technical (could be, given the richness of Excel plugins can have on some platforms), but technical in another field and Excel is an easily hackable UI for what they really want to do, which is mainly, as the OP suggests, using Excel to change stuff and do what-ifs, which spreadsheets were designed to do (as if, to close the circle).
I have a gut feeling that there is a lot of analysts who use Excel for difficult and heavy computations and who don't want to learn programming but would love to structure their formulas in a more decent manner.
> Users don't buy plugins in larger businesses, etc.
I don't agree - take a look at https://www.think-cell.com/en/ they are a company providing add-ins used by big four. They have 350 000 users in 7500+ companies with a revenue of ~50-70 millions of euro per year.
> The Venn-diagram between programming and Excel has no overlap imho
I believe the overlap exists. IMO it's possible to teach Excel users some basic practices from programming (like code review, testing etc). We (I have a company with my friends) actually have created a project that allowed a company write formulas in a simple file which generated a fully functional webapp. They used version control and code review together with unit tests. This allowed analysts build a solution using their knowledge without the need to have programming skills. This is a different approach programming Formulas -> Web application, but I believe an approach programming Formulas -> Excel is also possible.
My point (while I wrote "don't buy" and "no overlap") of course being somewhat less strong. Even with "hardly ever buy" and "hardly any overlap", the sales of even a strong product is hard and needs thought up front. You can build this tool and build it well and then struggle hard on the B2B-sales.
I agree it's tough to sell such product to a big company (actually selling any product to enterprise without credibility - a lot of users or big users - is very tough). IMO if you are small and you want to succeed in selling such product you need to make smaller companies love your product first and then talk to the big guys.
On the other hand I have spoken to Marcus (CEO and cofounder of Think-cell) about how they started and their product went viral quite fast and was used by one of the big four companies even at the early stage. I believe that if you hit the spot and build something that has a huge urgency in the market then even big fence in front of enterprise is not going to be your problem.
Having said this, I agree that selling this product could be tough. It's not just adding value but it changes the way people work - and enterprise hates changes. The more time big company has to sacrifice to change employees habits, the harder it is for them to decide to buy the product even if it adds value long term.
I thought it would be cool if formulas would pop into little syntax highlighted text edit areas when double clicked (or some other meta-click)
I think having a look at some more complex sheets would be good. (Ask for your company's accounting department I'm sure they have some Excel monsters lurking on desktops).
Maybe an editor that would extract all formulas from the sheet and give you different view after double click would be a good solution. On one hand you would still have your normal spreadsheet but you can go to a different level if you want to do refactoring etc.
Is this plugin meant for programmers or non-programmers? The distinction is critical because...
>Inspired by two discussions [...]
You have to be careful with criticisms about Excel made by programmers. Discussion on HN will skew toward the programmer skillset.
The mindset of the non-programmers that use Excel is very different from programmers. To many programmers, the use of Excel looks illogical. It's a spaghetti of dependencies when formulas refer to other formulas. A complex spreadsheet becomes a ball of mud.
Those criticisms are true but they completely miss the reason why non-programmers find Excel compelling: the UI starts with a visual grid of the "output".
On the other hand, a programmer's UI starts with a blank text file (ready for the programmer to type in "do while x<10" etc.
If one studies how all these disparate groups of non-programmers use Excel, the grid's visual immediacy is the common feature that they all prioritize. Diverse groups of non-programmers such as:
- the Wall Street quants playing with rows & columns in financial models,
- the project manager using Excel as a grid for itemizing tasks instead of using MS Project,
- the facilities manager using Excel as an adhoc inventory system instead of learning a database product like MS Access, FileMaker, or MySQL
It's true that Excel is spaghetti but non-programmers prioritize the grid over the resultant spaghetti. In contrast, programmers prioritize code organization (formulas) over the visual grid. These 2 groups are not even on the same wavelength.
Your proposed tool is helpful but it doesn't really address a burning need for non-programmers. For example, your mockup shows the formulas off to the side and disconnected spatially from the cell. That separation is fine for programmers (who can mentally connect things like CSS rules in a .css file to elements in the .html) but not user-friendly for non-programmers.
As a programmer I haven't actually thought much about how important this link between a cell in a grid and its formula may be.
I have seen huge (i mean HUGE) Excel files full of complex formulas being built by non-programmers. Some even took whole night to recalculate. There was a programmers mindset emerging from those files - extracting common formulas to separate cells (DRY), writing 'unit' tests by using formatting (if result for sample input is less than 0 then make the cell red - which means someone broke the formula).
The question arises - is there a point where a non-programmer starts thinking like a programmer, on a different level of abstraction? Is it possible to teach non-programmers some basic programming practices to make their work easier?
The people who are organized enough and use Excel enough to find value in saving formulas in a library are already extreme edge cases...and then you are selling into that tiny market against the idea of using full blown tools like Visual Studio and C#.
And it seems to me that really sound previously snakebit engineers aren't going to go around changing working software that isn't backed up by a suite of regression tests. The problem isn't Excel, it's bad code. The cause of the bad code is bad programming practice, and the reason for that is that the people programming are not programmers and even if they were they're under just give me a kludge time pressure.
Spreadsheets are successful because they're really easy to use. Excel dominates because making one that's userful for important work is really really hard.
Good luck.