Hi all, as good as place as any to ask i think and likely better than many others.

I am going to be looking to have some SW written and wondered if anyone would be able to offer any constructive advice.
It is going to be a database with a user friendly front end. I have all the fields identified, all the algorithms i need performed in a XLS (it works, just very delicate) and am going to mock up how I would like it to work and look. Is there anything else I should do before looking for someone to turn this into reality?
Thanks for your time.

Some basic architectural statements would be useful.

Does it have to integrate with any other system?
How many people need access to it simultaneously?
Should it be written in a particular language?
Do you want to pay any Licensing costs or should be it based on freeware?
Does it need to conform to any standards?
Where’s it to be hosted?

as SwedishChef says

Know exactly what it needs to do, why, what if and all that.
How does the data link, can you draw that out
How much data?
Is anything time critical
Is there a response time needed
Backups
Security

Be SURE that the people who write it for you are going to be available for updates/bug fixes etc AND you get the source code. Met a lot of people who didn’t get source code and the original company either went under or the one-man-band went and got a job. Get someone to test the source code actually IS the source code (i.e. if it’s to build for a computer, does it compile or if it’s for the web, can it be installed by someone else). Finally, don’t pay everything up front. Pay them a fair percentage but under no circumstances cough up all the money beforehand (I’d withhold 20-30% unless using an escrow service).

Does this need to written from scratch? Sounds like something could be handled pretty easily in Filemaker.
Just where/how is the sw going to be used?

With anything like this it’s easy for feature creep to quickly raise it’s head.

Write down what is your MVP (minimum viable product) to get it released. Then pay someone to write you a clearly detailed technical spec based on your requirements etc. Unless you’re a techy I’d advise not to do this yourself as you’re likely to miss stuff. Once the tech spec is done get some quotes and rough timescales. Use the tech spec as a bible – it gives you a document that both parties have agreed to deliver and pay for.

Anything outside of that is creep unless agreed by both sides to integrate into the cost, or deliver at a later date. Once you have your MVP delivered – make lots of fast, incremental changes until you get it to a point you’re happy with it.

Lots of good advice. Another question to ask – where do people need to access it from? Only in the office, or while out and about, via laptop, mobile?

This might help decide on the hosting – I’d possibly consider one of the big public cloud providers (Microsoft Azure, AWS)s o a whole lot of admin is taken care of (back-ups, security, scalability, patching etc.).

I would start simple. Starting with the concept of MVP above. Summarise the business problem(s) the solution needs to solve. Document the [i]essential[/i][/u] features it needs to have. Identify the needs it must fulfil.

Before I looked to create any solution, I would look at the market first. Despite what we often think our needs are most often common and not unique. Most often better in terms of initial cost and ongoing support to buy rather than build.

Bespoke software is expensive; what’s your budget?

I run a small software business; and I’d be happy to look at your specs and give you a quote – I’d agree with all that’s already been said above. Work out your MVP; and go for that. Extra features can be added later (at a cost of course) – but generally I find that the extra features people thought they wanted are not the actual features they need once they’ve used the MVP for a few weeks.

That is all god stuff, Thank you.

Some answers to the questions i hope:

Only needs to be one user at a time.
No need for web/online unless of course it would help.

MVP: imagine an asset management database. The user will enter a set of assets and various (read numerous individual fields. Once that particular asset type has been entered subsequent ones ought to auto fill with the option to change the fields if needed. Once the assets for a particular site has been inputted it should be able to perform some algorithms (already written and working in an XLS) and output a report.

Each site should be available as a separate data set that can be worked on, saved and updated.

I was thinking I could make a mock up in power point with using hyperlinks to show how I envisaged it working. Bad idea?

Licenses for pretty much the whole Microsoft catalog are already held as well lots of others that i don’t have a clue about.

I was thinking it was just a case of a pretty front end on an Access database but as you can tell I am out of my area here.

As for budget, I have to get some quotes for Feb to go cap in hand to the board. Who can say how much they might throw at the idea. It has got the green light so far and looks like it will continue to do so.

Sounds to me like you’ve just described Microsoft Access.

inputted it should be able to perform some algorithms (already written and working in an XLS) and output a report.

I was thinking it was just a case of a pretty front end on an Access database but as you can tell I am out of my area here.

Sounds like a good application for VBA….

I was thinking I could make a mock up in power point with using hyperlinks to show how I envisaged it working. Bad idea?

It would probably help; always better to have something to talk through !

Doesn’t sound like it would be too tricky; but you need to be careful with specifications that are open to interpretation..

as an example – once you’ve specified an audit item of a certain type; then you want it to auto-populate for other similar items. What if you’ve got two with different entries and you add a third ? Is the template fixed from the first; or do you go with the most recent ? Or do you want the app to prompt the user to choose an existing item to pull the values from ?

Just bear that kind of thing in mind !

Sounds like you have already done quite a good job on the spec and it doesn’t sound too bad to do. Make sure to specify which field are required and which can be empty. It’s difficult for an external developer to guess that but it’s probably obvious to you.

Doesn’t look like a big job so it may be an ideal thing to ask for on elancer

UHC – download a free trial of Filemaker. It will do all you need plus a lot more, is simple to use, cross platform, can be web and mobile enabled and, if you get the developer version, you can save databases as runtime versions with no software to install.
You could do it all yourself.

I think that from the posts that I am talking about Access. So who do I go and see for a chat and rough quote that I can use for my case?
Is it something an enthusiastic amateur might be able to do? I have someone in mind if you think it might not be too big an ask.

Unless you are in the software business your board likely won’t understand the MVP approach and will perceive uncontrolled costs and a never ending project.

It does sound like you could do this in access or filemaker. I see them as tools that let you DIY a solution rather than having to formally specify and build a “product”.

access can work across multiple sites (presumably so can filemaker) but the last time I tried was ten yrs ago and it was flakey, and had issues with multi person access (if you don’t need this at some point then I’m sceptical you actually need a database!).

What you described sounds like the sort of thing a competent person could easily nock up and test in a week. Don’t underestimate test time. I’d expect to pay 2-3000k for that. Most boards will take some convincing that they will see an roi for that (in addition to maintenance and improvements) when your accounting software probably already has 90% of the functionality for no extra cost.

I’m always sceptical when anyone says their database doesn’t need to connect to anything else, that either they don’t understand the problem/application or it holds little valuable information.