Tips and Tricks for a new Freelance Developer

[Harsh Gupta]

Hello,

I am planning to take up freelancing on much coaxing from my friend. Though it's only for this one project that he is taking up, and we may start working from next week on. (I believe it doesn't matter how many people are involved!)

My problem is more with the kind of experience that I have gathered in 2-3 years of my career. It had been more of a coerced developer, pressurized by management to learn their ways of getting things done. Understanding what the end client is asking for, or looking for, was never my cup of coffee (umm tea, but well I have a coffee fetish )

So I am taking up this project as a challenge and planning to take up the task of foreman and interact with the client directly. But I am not sure what comes after that.

I discussed this with my friend and he told me to create a documentation or a synopsis of proposed system and get it verified with the client.

I don't know where to start from but based on conversation with 2-3 people, I planned a following to-do list:

1. Creating a point-by-point list of jobs that the client want the system to get done.
2. Creating a flowchart (probably in Visio) explaining the flow of the system.
3. Draft of a probable DB design.
4. Making designs of the proposed system and attaching its screenshots, explaining the flow of windows/Webforms and the expected data entry or what data can user expect at this point etc.
5. Take it to the client, discuss it with them and getting their nod.

Am I on the right track? Am I missing something here or maybe doing too much? What if the user in the end denies that this is what they were actually looking for?

Could anybody please share some tips and tricks addressing these issues? I shall be obliged.

Thanking you.

[Hack]

If you are going to free lance, you are missing a very important part: your fee.

[szlamany]

You have to get sign-off from the customer that you both agree on what will be delivered.

I just got a new job yesterday - big job.

The customer was nice enough to create a spec sheet of business process for us. If they had not done this we would have needed to visit with them in advance of everything else to gather that info.

You need to know what you are automating. You need to see paper forms they might be using now. You need to understand how data comes in and how it goes through the office and what is expected on the output side of that data flow.

This customer also gave us sample screen shots that they created - also rare. If they hadn't done this we would not have prepared any screen shots in advance of starting the project.

At any rate - we worked with them for about 2 months reviewing specs from them - discussing process - watching them complete more of the spec they developed for us.

We created a task list from all this - with milestones.

And yesterday we agreed on a price - signed an agreement saying we would deliver what was in the spec. We also agreed that the spec was a living document and could change - and we would not change the price as long as the amendments were in line with the original scope of the project (careful about this one!).

We got a down payment - are going to start on steps 1,2,3 of the task list. At the first milestone we show them what we have so far. They either accept it or ask for modifications. If it's accepted we get another payment and move along on the task list.

We are big believers in the unavoidable nature of iterative development - nothing can be done on first attempt that meets the actual needs...

[kevchadders]

We also agreed that the spec was a living document and could change - and we would not change the price as long as the amendments were in line with the original scope of the project (careful about this one!)

Very important part this one. make sure you fully understand what is expected from you. If you agree to make the odd 'minor' change here and there, then make sure you nail down on what consistutes a minor change.

Also, if for some reason they wanted a big change mid way through the project, somthing might have to give. e.g. The projects slips by a few weeks due to the extra agreed change (plus your additional costs), or someting else could give in the project so that you could fit in the new requirement in, which would have to be agreed by both parties.

As its your first trip into freelance development, you might be eagar to please so just be careful what you get yourself into!

There are some business methodologies out there which can help you manage your project so it might be worth taking a look.

ps. szlamany you sumed it up nicely there.

[Harsh Gupta]

@Hack,

That indeed, was a good thing that I missed out. Gosh, how can I forget money!!

@szlamany, thanks for such a super insight on the subject.

I talked on the subject with my friend (the fore-moron of us all), and the typical flow he described matches closely to what you explained, specs from them and us speculating, but trying to be definite as good as we can be, task list - with milestones etc. I doubt they will give us some screenshots (rarely as you mentioned) and all that.

And yesterday we agreed on a price - signed an agreement saying we would deliver what was in the spec. We also agreed that the spec was a living document and could change - and we would not change the price as long as the amendments were in line with the original scope of the project (careful about this one!).

This is something I wanted to know in depth, and will eventually understand with some experience that will come by. I am not really sure how do you guys calculate price per se milestones, guess I can understand while watching other members addressing the issue.

Anyways, that was a great help, and I appreciate it. Sorry, I can not rep you for this one. Need to spread some more reps.

Guess, I wouldn't be doing this for all the time to come. This looks like a monster to me.

As its your first trip into freelance development, you might be eagar to please so just be careful what you get yourself into!

Thanks for the advice. I am, in fact, really curious if I could ever be good at freelancing after all these goose bump. Just playing a gamble. Thanks for the suggestion though.

Is there anything else I forgot or is worth mentioning, or should be in my to-do list?

Thank you all.

[szlamany]

Once you have the spec the price actually comes kind of easily...

The specification needs to be turned into tasks - "Maintain lookup tables", "Print PO's", "Massive tree-view on the left", "List-view on top", "Drag-and-drop capability", "E-mail from VB".

That's just a sample of a few from the app we just started.

We created this task list in a group meeting with 3 people - all reading the specs and throwing ideas out - all getting onto a whiteboard.

Missing a major task would be a big deal for us - so we spent hours on this.

Then we all stepped back and started "costing" each task. Since it's a real 3-tier app we put hours against each task for UI / BL / DL. Some only had UI - some only had DL - whatever.

And we were generous with those hours - everything was in 4 or 8 or 16 or 24 hour increments. Nothing gets done in one hour - you have to think - code - test - show others - come back and re-work.

There are 6 information tables in a single task on that list - they ended up being UI:12, BL:6, DL:6. That's 3 man-days of work. And we kept challenging each other - could you really develop 6 pop-up screens with a couple of dozen fields and checkboxes in 3 days - you have to be honest with yourself otherwise you lose your shirt.

At any rate - each task got popped into an excel spreadsheet - with those 3 columns of figures. Easy to grand total that.

Let's say it came to 500 man hours - what would you need to earn where you live to make 500 man hours a reasonable job. That's basically 3 months of time. If a well paid consultant earns $200,000 a year then that's a $50,000 project - this is easy math to finish up.

Then that excel spreadsheet is organized so that the tasks are put in an order that we feel is best to approach the project. We wanted our milestones to be every two weeks or so - so we simply "bracketed" them in a way that we could fit x-hours into a two week timeframe. Some items start in an earlier milestone period - so we had columns to the right for each milestone where we spread the hours of the task across.

In the end we present to the customer the part of the spreadsheet that only shows the tasks and the "milestone" completed aspect. All the hours/days info is not presented to them.

[kevchadders]

Something like a GANTT chart (Using Mircosoft Project) might be useful to set up when you have all your tasks defined.

It can be pain to initially set up, but it can also be useful in keeping track with timescales, where you are up to, effect of project slippage, which task you should be doing, milestones etc.

The link below taks about using a Mircosoft Project Gantt Chart, and there is a diagram near the bottom give you an idea on what it looks like.
http://www.me.umn.edu/courses/me4054...nts/gantt.html

Also, there is a PDF link in there to show you how you could do a Gantt chart using Excel
http://www.me.umn.edu/courses/me4054...ents/gantt.pdf

[szlamany]

...Some items start in an earlier milestone period - so we had columns to the right for each milestone where we spread the hours of the task across.

Yup - that PDF looks pretty much like what we have been using for about 20 years now!

Used to do it in a notepad like "editor" on mainframes back in the 80's! With "asterisk" characters along the 132 character line!