Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
Jim Beam
Posting Yak Master
137 Posts |
 Posted - 2011-09-05 : 13:36:05
|
| Hi all, been a stranger round these parts for a while, and I've decided to go it on my own starting small and working up.Got a meeting in a couple of days with a fellow who runs a small family-staffed wholesalers, pretty much a blank sheet site, processes about 2,000 orders a year with data strewn everywhere across umpteen Excel files and other proprietory systems. He's hoping to have a database put together containing all product ranges, button-touch remote reporting on profits/sales etc, stock levels, Billing, etc, none of which are currently automated/interconnected.I'm thinking that it's not going to be much more complex than Northwind, but as this is my first foray into 1-1 consultancy and proj-managing the entire SDLC, I wanted to get some idea of what questions I'll need to ask, what kind of timelines I should quote for R+D.Anyway, I think an Access db should suffice for the above purposes, again not dissimilar in structure to Northwind.My questions are;1) Anyone done this and have checklist/roadmap?2) What kind of timelines I should be quoting? (assuming uninterrupted work, free access to all data etc.)3) What kind of questions should I be asking him at our meeting?4) He feels he'd benefit from BI as do most businesses, but what kinds of KPI's/benefits should be agreed here for a small firm like his?Thanks,Jim |
|
|
Kristen
Test
22859 Posts |
Posted - 2011-09-05 : 14:01:40
|
1) Have you got a budget and time-to-read some books? Microsoft Press Steve McConnel - e.g. "Software Project Survival Guide" and possibly even "Code Complete"2) Time and Materials. Project-creep is inevitable. Your "boss" might agree to monthly rate for as long as it takes. You don't have to spend X% of you time trying to find the next/other clients, so can reduce your rate. You also (and this may be a selling point) have to spend Y% of the project, upfront, working out how to sell it to him ... money that you would otherwise have to re-coup). Just a thought ...3) Nah, you won't know that until you've done a few, and we could only tell you in response to the Q&A that will come up in the meetings ... its largely down to "experience" I'm afraid, someone here might have a URL to a nicely thought out page of questions though ...4) All those Buzz words worry me ... Sorting the problem properly, and thoroughly, will bring all sorts of benefits - some hard to quantify.Some 30+ years ago when I was doing much as you are - given an opportunity by a family friend - and whilst still at the discussion stage (and given that, way-back-then, no small business had computerised-anything) I heard the Client's support staff on the phone saying "... and anyway, we are computerised now so we couldn't possibly change that". That one, bold, remark probably saved two of his staff a day's work - bit hard to quantify that! or to even worry about trying to quantify it up front.I come at "What's the benefit" from a different angle these days. My metric is: "Is that 'task' just 'feeding the beast'?" Anything where human-time is spent collating and keying-in data for analytical purposes that cannot EASILY be justified is "feeding the beast". Staff become disillusioned at having to do it (assuming that management don't reap HUGE benefit from it), data is poor-quality, and management just print it out and WEIGH it to gauge its benefit, rather than either interpreting it or just plain finding-it-useful.Hope it goes well  |
 |
|
|
Jim Beam
Posting Yak Master
137 Posts |
Posted - 2011-09-05 : 14:54:11
|
| Just reading through "Survival Guide"...were you one of the authors? It's really your style!!:) |
|
|
|
Jim Beam
Posting Yak Master
137 Posts |
Posted - 2011-09-05 : 17:50:30
|
| Just had another thought, typically where's the best place to locate the main CRM/ERM DB for a company that can't afford a dedicated DB server? With the web host? |
|
|
|
Lumbago
Norsk Yak Master
3271 Posts |
Posted - 2011-09-06 : 02:38:18
|
| What you could probably do is to create an access front-end with a sql server (express or azure) backend. Azure is really cheap for small scale databases and by setting up an ODBC-connection using the sql server native client I think you should be able to connect Access to it. With Access you can do really rapid application development and by using a sql srever backend you'll make sure that it scales if the business grows.1) No suggestions here..sorry2) Try not to make any guaranteed deliveries in x amount of time. Requirements are constantly changing and taking this in to account for such a small company will just give you headaches. Try instead to set an hourly/daily/monthly rate and say that you will work there for x months3) What are you expecting to get out of this? How much are you willing to pay for it? How do regard the prospects / growth possibilities of the company? 4) For small companies (all companies really) it is important to measure profitability of all aspects of the business and surprisingly this is often neglected. Quantify EVERYTHING. Small businesses tend to calculate profit as a factor of sale price minus manufacturing cost, while they don't take in to account customer support, management overtime, delays in the supply chain, delays in customer payments, staff coffee breaks, etc- LumbagoMy blog-> http://thefirstsql.com/2011/07/08/how-to-find-gaps-in-identity-columns-at-the-speed-of-light/ |
|
|
|
Jim Beam
Posting Yak Master
137 Posts |
Posted - 2011-09-06 : 04:27:25
|
| On timelines, I'm almost certain he'll be looking for a ballpark amount of time (I would). I'm considering quoting Christmas Eve to have everything up and running that we discuss, any extra work (whether related to the project or not) could entail overrun. |
|
|
|
Kristen
Test
22859 Posts |
Posted - 2011-09-06 : 08:29:23
|
| You could have a look at Fogbugz. Once you set up the Trial there is a setting in CONFIG where you can set it to "max 2 users" licence, and that is free.Put all "tickets" in, give everything a time estimate, be sure to use it to record time (its not important that you accurately record time, you just need to be consistent. So if you leave the clock running (on whatever ticket it was on) when you have a meeting with the client then do that consistently.You'll then have Burn Down charts to predict completion dates. You can put things in separate phases, and have delivery dates for each. As you overrun, or client adds more Tickets, then the delivery date will change :)There is a "timesheet" module somewhere (i.e. might be 3rd party) which will give you a decent enough report for billing purposes. But Fogbugz tries not to be time recording, per se, as they think that makes Bosses sit on Developers, and then time recording becomes whatever the bosses want it to be ... which then wrecks the ability to predict project deadlines of course! |
|
|
|
|
|
|
|
|
New Project, advice wanted!