how do you charge for the design of a database?

[Cowboy_BeBa]

Hi

Appologies if this isnt the right place to post this sort of question, i wasnt quite sure where to post this.

What would be the going rate for designing and building a database for a small business?
I reckon it will take about a month to build (a week to speak to all the end users and get to know the way the company operates another week/week and a half to build it and the rest of the time to implement and train staff, as well as deal with any bugs that may come up) and i might also offer them a few months of ongoing support (will also need to figure out a fair rate for this as well, depending on how much work i actually do).

I've had very limited experience building databases for other companies, however have spent the last 6 years building a database for my dads factory (well that only took a couple weeks to build initially) and updating/expanding it. I work there more or less full time and receive a regular wage. So i do have plenty of experience (as the database at work has evolved far beyond its original scope, its essentially a fully functioning erp2 program and i've learned alot working on it)and im more than confident i can build the db (they basically want something similar to the one i built at work, but tailored to work for them) but dont know quite how much to charge. What would be fair? I was thinking about 10k for the initial month and then an hourly rate of $20 to $30 for ongoing work if required.
Is this too much or too little? Is there a standard that can be applied?

The person i spoke to about this said theyre aware of professional ERP2 systems that can cost between 60 and 100k, which is far more than this small business can afford. So i'm not too sure if im undervaluing or overvaluing my work

Does anyone here build access db's professionally? if so i'd love to hear from you (especially because i would love nothing more than to do that myself as a full time job, am hoping this job will eventually lead to that) and any other advice you can offer would be greatly appreciated as well

 

[JHB]

...
I reckon it will take about a month to build (a week to speak to all the end users and get to know the way the company operates another week/week and a half to build it and the rest of the time to implement and train staff, as well as deal with any bugs that may come up)...

Depending on how advanced the database would be, you are very optimistic.
Only for a small database I would deal with 3 times more as you here mention.
Testing and get the database to run smooth, the ½ or more of the time is used.
And many times you have to change it because it is not exactly what the users did expect.

 

[spikepl]

I'm with JHB on this:

The first 90 percent of the code accounts for the first 90 percent of the development time. The remaining 10 percent of the code accounts for the other 90 percent of the development time.

— Tom Cargill, Bell Labs

But the pricing issue is not related what it costs to make the thing, but what it is worth to the client. Remember that the client will not sign-off on thew deal unless they feel it is a worthwhile one. The price can be fixed (customers love that) but the amount of effort required is indeterminate prior to having all the requirements.!And requirements have this tendency to emerge as you go along.

The cure for this is to follow the agile thinking: offer fixed-price and fixed-time delivery of working and tested but functionally well-defined but limited subsets.

 

[Minty]

We were going to have Great Plains customised for a part of our business, and were quoted for initial install and configuration. Once they had been on site for the initial period they said that they were going to need extra time. (I think we had already agreed to £8k worth at this point)

Hmm ... Okay what's the rate - £1000 / day ...
! Okay and how much longer is this likely to take ...
Them - Well probably a month, maybe longer....!!

Okay we'll can that idea then and develop in house... !

 

[gemma-the-husky]

it all depends on a clients requirements

generally, an off the shelf pack will be the most economical solution. However there may not be a suitable off the shelf solution, or it may require compromises they can't make.

in which case they may decide to seek a bespoke solution. They will pay more, and probably get a system with fewer bells and whistles, but hopefully it will do exactly what they went.

If they go the bespoke way, they need to specify exactly what they want, which isn't easy. An estimate of a week or so to build it sounds wildly optimistic. If I had a project that would take that little time, I wouldn't need to do much research with the client, as I would have a good idea what they wanted without.

If you need a week to talk to the staff, I reckon you will be needing about 3 months to build the database. For a big project, don't expect to deliver a complete solution. You ought to prototype, and review the prototype with the client.

 

[The_Doc_Man]

You know the customer understands running a business. You have to understand that, even if you are a company of one, you are running a business too.

I might offer the service you describe as a two-part investment for them. Tell them up front that the big pitfall is that you cannot accept a true fixed-price contract for the whole schemer. Before you start, figure out your labor rate. When you guesstimate your hours, ALWAYS round up to the nearest even day. (Some authorities suggest rounding up to the nearest even millennium, but that might be considered price-gouging.)

Part 1.a - Talk to the manager requesting this service. Determine the number of people who have to interact with the app INCLUDING those who enter data AND those who only see reports. This portion of the discussion (i.e. pre-discovery) should be free.

Part 1.b - Make an estimate based on the number of people you need to contact and how long you think you will need to talk to them. Detail it out as: Interview x persons for approx. 1 hour (or 2 hours or 1/2 hour or whatever) . Put a disclaimer in the estimate regarding your time charges when people become unavailable while you are on their site. Obviously, in this phase you won't charge for days where they tell you not to come. Do your discovery. Get the requirements. As part of the labor description, include that you will be taking detailed notes during discovery and thus will need a small increment of time to fold in each set of comments.

Part 1.c - Be prepared to use something like OneNote or Paint or Word or PowerPoint that can do graphics (or bring a good pencil and an artist's note pad). Sketch out the rough appearance of the app showing where each text, combo, or list box will go. Where you have button-type controls, include them. Allow some amount of time in the detailed estimate for phase I to include your appearance design time and at least one visit (but optionally more than one) to get design approval.

Now, part 2 - Estimate how long this will take to implement as designed. Don't forget to consider the VBA portion that CAN'T be done via code wizards. Figure out each displayable item and its Class module and round up per item. I.e. don't charge 6.5 hours for a form. Charge 8 hours. (You pick the round-up target, e.g. nearest half-day or whatever.) Where you have modules that must interact or where you need to have common code among multiple displayable presentations, count those separately.

Add it up, multiply it to get the dollar rate for that number of hours, and present a fixed price. DON'T forget to include a clause that says that changes to the design as originally approved will have to be estimated separately if they exceed a certain level of effort. I.e. don't gig them for an extra day just to change color schemes on a single form - but if they want to change the scheme on 30 forms, that becomes tedious enough to show up on the radar as an extra charge.

Don't be afraid to remind them that you are approaching this as a business, too. But here is where you have to be very careful. If the contract includes a specific delivery time, verify that there are enough hours in the day to make that deadline AND include a comment that design changes big enough to incur an additional charge DO affect the delivery date.

 

[spikepl]

Just to illustrate (when the precise scope is seldom known at the outset), check out the figures that show the factor (and the ensuing uncertainty) to multiply the final true budget, depending on how far you are.

I can highly recommend Construx (am not associated with them in any way) - they have plenty of very useful things on their site and the main force behind Construx , Steve Mcconnell, has written very interesting and very practical books about software project management.

 

[Cowboy_BeBa]

wow, thanks everyone for all the advice, sorry for the late reply, it was about 4 am here when you guys started writing.

i guess im going to need to do a bit more research and prepare myself a little more before i get started, you guys have introduced factors i never considered so will give it a little more thought before i start (i dont think theyll need me till mid next month, so that should be enough time to prepare)

 

[Pyro]

One thing i would add if you go down the road of providing a fixed price is to provide a detailed proposal/quotation. I itemise every data entry point (form, automated import etc), as well as every output (report, data export, automated email etc) and make it very obvious that the quoted price is for the listed items (and those only).

I put it back on the client to provide a thorough scope, and have documents to guide them through the process. I offer to assist in scoping for a standard hourly fee if required. Once the quote is signed/deposit paid etc if an item comes up that was not included in the original scope and not accounted for in the quote, i submit a variation with a seperate cost.

One further recommendation i have would be to use a model that sees you paid an amount that you are comfortable with prior to handing over anything to the client (especially for new clients). In my case that is usually around 75% of the total paid in installments over the course of development. the final 25% is invoiced the moment the system goes into live use.