Apply HN: Turn your spreadsheets into a complete webapp in minutes

23 points by singold ↗ HN
PROBLEM

Most businesses have important information registered in spreadsheets, those spreadsheets generate a lot of problems like loss of files, versioning, loss of information, lack of accountability in changes, etc. Some of those spreadsheets start as something small but end up being pretty important for its users. To substitute those files with custom software is difficult because in medium or big enterprises IT departments work on the most important things and can’t develop software for “small things”. On the other hand outsourcing that development can be difficult and/or costly. For smaller business without IT departments the problem is the cost of developing custom applications.

SOLUTION

Let users upload a spreadsheet to our platform and transform that into a custom web app, adding useful functionality like search, filtering, reports, security, users, permissions, etc.

BUSINESS MODEL

Basically we will have two products the hosted app (recurrent revenue) and the downloadable app to host on your own server (as a premium product). As a business strategy we need to have affordable prices so that medium management can approve the purchase. Also our apps need to be functional, with good usability and good design (in that order).

VISION

In the long run we want to make great software for any kind of companiess. To do that we need to enter the market with smaller and less risky applications because “no one ever got fired for buying <big enterprise software company>”

MARKET

Every business that uses spreadsheets for long term work (need to work more on this)

TEAM

Right now it is only me, but I am looking for co-founders. I have a technical degree on software development and experience in various non-IT areas of a bank (first hand experience of what I call the “spreadsheet hell”).

PROGRESS

I have a sketch of the application and am developing the first working prototype

Any kind of questions, opinions, etc. are welcome.

Thanks in advance!

39 comments

[ 3.1 ms ] story [ 75.5 ms ] thread
We already have Google spreadsheets and many more such apps/web apps. Why would someone jump to this?

I think you need to clarify more about your USP. Becoming more specific about your functionalities such as what kind of security or what kind of user permissions, etc. might help you gain more attention.

People already have such apps/web apps available. So, you need that one wow factor to make the users jump from one service and come and use yours. Try to think more from an user's perspective and try to redesign the way your web app is going to work or help companies/individuals in handling spreadsheets with ease. Good luck! :)

Maybe I didn't explained myself well or maybe I didn't understand your comment (or both :).

In other words the idea is that you upload a spreadsheet and we make a CRUD-like app for you with the information that you already have in your spreadsheet.

The purpose is not making spreadsheet management better, but make spreadsheets "obsolete" for some kind of uses.

Have you used google spreadsheets on mobile?
This is a problem I have that I thought was near to impossible. I have a nightmare test case for you. I am a property manager of 1000 units. We have 40 properties each of which we track vacancies with a spreadsheet. They've got all relative property data on them as well. Let me know if you are interested.

My solution was to build a web app for this completely out of spreadsheet land. But I've got some other things to tackle first. Either way good luck!

Is this still an issue for you? It would be interesting to discuss a solution.
You will probably not be my first client :P but I love a nightmare test case.

I would love to see that spreadsheet (or a part of it with fake or no data obviously), my email is in profile, feel free to contact me.

Wow sounds like the kind of thing patio11 touched upon in a comment about where there is a spreadsheet being banded about there is a potential saas.
Not affiliated with them at all, but airtable (www.airtable.com) might get you there faster (basically simple database modeling using a spreadsheet metaphor).
Try Airtable.com, you might turn this into an "app" in an hour.
Hey there, I created an app called spreadsheet.io that did something similar (Google Refine + Fusion Tables). I have no bandwidth for dev, but am open to partnership.
I would like to see what your app is about, but spreadsheet.io returns a 503 error, can I see something about it on another place?
Yea I took it offline. I'll update this thread with a link to details.
This seems like it would be a great idea if it can be done because the market for it would probably be very large.

However it seems to me that doing this would be very difficult for at least 2 main reasons.

First spreadsheets can be organized in essentially arbitrary ways and there is no standard way of defining the semantics of data in them. There can also be all sorts of references between data items and unlike SQL there is nothing to explicitly define the nature of these references. How could a program take an arbitrary collection of spreadsheets and transform them into a webapp that has significantly more functionality than just loading them into Google docs ?

Second spreadsheets often use macros and other programming constructs so I think you also have the problem of building a general translator from macros/VB/Excel functions to javascript/whatever backend language you're using. This again seems like a fairly difficult problem to tackle in general.

I agree with you that it is a difficult problem to tackle and I've been thinking partly on what you say.

About your first point there are two important things, first that we are going to start with simple spreadsheets and then evolve to more complex ones. And second that we are going to have some user interaction, because there is a lot of information that is in the user's head that we can't infer only from the data.

About the second point, the first "simple" spreadsheets are going to be (probably) without macros, etc. (anyway at the start we can do ThingsThatDontScale(TM)). I think that making a transpiler from excel functions to a programming language shouldn't be that difficult, but the macros are a totally different beast.

Thanks for your reply!

The sketch of the user flow for the first prototype can be seen here https://app.moqups.com/santiago.ingold@gmail.com/KCTLq4WnAL/...

Again, comments are welcomed :)

I took a look at the mock up, and I think you're aiming for the same space as Karma Platform; converting spreadsheet-as-db to web app.
Yes, that is basically the idea for the first version, I will take a closer look to Karma Platform
This is already happening on a lot of fronts, I've built entire apps on zapier.

What are your thoughts on the following:

Flex.io zapier.com Blockspring Zoho Creator Domo

If you build this platform I think you have to start application first and not try to theorize it.

From the apps you mention, except for zoho creator, I feel like are all developer oriented (I can be wrong).

In a way, what I want to do is to make apps using spreadsheets as the requirements for the app.

What do you mean by 'start application first and not try to theorize it'? Isn't that I don't agree but I didn't understand it well and it looks like somenthing interesting to take into account.

I'm not a developer though..... I use these ones.
Ok, maybe 'Developer' wasn't the best term, what I really meant was that those products are for people that are somewhat technical oriented.

For example, I see from your submission history that your are probably a biochemist [1] and also that you know what is a regular expression [2] so I guess that you know how to program to some degree, and that if you wanted you probably could make your own crud app.

The majority of people that I imagine using my product dont't know nor want to know how to program, let alone what a regular expression is.

BTW, im still interested on what you meant by 'start application first and not try to theorize it' :)

[1] https://news.ycombinator.com/item?id=10863718 [2] https://news.ycombinator.com/item?id=11024116

Have you looked at spreadsheetweb.com and calcfusion.com? The former seems mature, with plenty of real paying customers. The latter is new, and I guess they're still working on getting traction.
Here's two more: karmaplatform.com and kdcalc.com. This whole area of converting end user produced Excel solutions to scalable systems does seem very promising. But no one has nailed it yet, and I suspect there are at least two reasons. 1) Over technical solutions that require a .xls to be converted into some other technical deployable eg kdcalc.com producing a .jar 2) The diversity of Excel based solutions eg spreadsheet as DB vs. spreadsheet as calc engine vs spreadsheet as reporting dashboard.

So when you say "convert a spreadsheet into a web app", it means many things to many people. Which is a real expectation management challenge!

The expectations management you mention is really important, I need to think how to communicate that clearly to customers.

In this case we are going to make it "cloud based" so there's no need to deploy anything. About the different kind of spreadsheet uses, the more use cases that we can found the better, also we are going to start with the more simple cases to reach a functioning product and then expand to the most complex ones.

Seems like my google-fu is a little rusty, because I didn't find this kind of companies that I assumed that existed.

This looks like real competition, what is good to validate that a market exists.

Thanks for the tip! I would look into them with more detail tonight

Hey lvp3, this sounds like a good case for Fieldbook (my startup): https://fieldbook.com. It's as flexible as a spreadsheet, but with the power of a relational database.

We've had a few real estate offices and property managers use us; you wouldn't be the first! If you send me your spreadsheet I can even get it migrated for you as a demo. jason@fieldbook.com

It feels disrespectful to hijack someone else's Apply HN application like this.

We detached this comment from https://news.ycombinator.com/item?id=11458280 and marked it off-topic.

Oops, my apologies. I only mentioned it because other people were mentioning alternate solutions on the same sub-thread.

Promoting an app when it's relevant seems common on other threads, but I can see how it would be rude on an Apply thread. Won't do it again.

If it can be solved beautifully, it would be cash cow. like the others mentioned, translating UDF (user defined function) would be a challenge. also in bank, it usually load quant library xll. It would be impossible to host quant library as service. nowaday i see bank just use spreadsheet as front end. a lot of logic are wrapped in code and fully tested somewhere else
Thanks for your reply, I didn't know about quant libraries for excel, I`d investigate that.
It's not impossible to host a quant library XLL as a service. I've done it with the QuantLib XLL. My HN profile has links that detail how it's done.
This is really cool. I tried a few c++ call as endpoint in a webservice. but i find it difficult to maintain state as you know market are in memory. I am also thinking to recreate dependency graph as futures in tornado but I gave up as it seems too difficult. eventually I just move xll out of excel process to make it s/s a bit light weighted and faster to run
Hi lvp3, that use case is quite close to what we're building our platform for. It's still early beta and probably not ready for production yet but we are looking for feedback: https://news.ycombinator.com/item?id=11491383.

Our focus is on the business logic of things so that the data input gets easier. Spreadsheets tend to become messy and require a lot of repetition of data and maintenance.

I am aware of a few "spreadsheet-on-line" that work wonderfully as well. With our platform (Simitless) we are trying to get away from the spreadsheet metaphor. We want to give niche businesses requiring custom apps but with little investment an opportunity to define their data app little by little. That way you can build it as it should be: focused on the input tool and the data exploitation.