Včeraj sem čistil disk in sem že skoraj zbrisal datoteko s temlem tekstom. Objavljam ga tukaj, mogoče bo komu uporabno...

Je pa že tako dolgo na mojem disku, da se ne spomnem kje sem dobil zadevo...

Why it's important, and how much was actually spent on DropSend

What's the big deal?

You don't have to be big any more. Remember boo.com? Raised 100 million, spent

5 million on a server, nothing to show for it. This isn't the case any more.

Why now?

* Broadband is widespread

* People are more comfortable with web apps - Webmail, online banking

* Hardware is dirt cheap - no need to spend 5 million

* Open Source software

What's enterprise?

Debatable, but mass-market and 1000+ users.

On a budget?

Under Ł30 grand. That's the bare minimum you can bring a web app to market on.

How to come up with ideas? Find a personal problem. "an itch"

DropSend is used for sending and storing large files - how do you email a 10MB

file? It also acts as an online storage repository. 9,500+ worldwide users in

2 months. 5 servers at 365 Main in San Francisco. Desktop apps which use an

(unreleased) API. PHP and MySQL.

The most important thing: Make sure you idea is financially viable. Not just

handy but worth paying money for. "Would you pay for it?"

Be cautious about your projections. Make an estimate, cut it by 45% and see if

you're still in business. Just an Excel worksheet for cash flow - how much do

you make a week, when do you run out of money?

Acquisition my ass! That's a stupid business model. "I just think it's lame".

It's not a sure thing - you're betting a lot on that.

Forget about the "Is this another bubble" effect. Betting on something so

vague is not often a good business model. If you have an application that's

financially viable, and good, then a bubble doesn't matter.

The budget:

Different budgets for different apps. DropSend is hardware intensive (big

files).

Branding and UI Design: Ł5,000 (Ryan Shelton, probably a friendly budget done

over one month, by the sound of it plus accomodation and food)

Development: Ł8,500 (Plum Digital Media - http://matchbox.plumdigitalmedia.com

based in Bath, plus a "small percentage" of equity)

Desktop Apps: Ł2,750 (Windows and Mac variants)

Build (XHTML/CSS): Ł1,600

Hardware: Ł500 (development server in-house)

Hosting/Maintainance: Ł800 per month (Bit Pusher)

The cost difference between 99% and 100% uptime is huge, often out of

range for smaller companies)

They deployed it and maintained it for that price a month.

Legal: Ł2,630

Accounting: Ł500

Linux specialist: Ł500 (development box config, apache virtual hosts,

subversion, etc.)

Misc: Ł1,950 - e.g you have to take an unexpected trip, your laptop craps out

Trademark: Ł250 (lawyer quadruples this fee, do this early as it will have an

impact on the branding)

Merchant account: Ł200 - a bank account designed for online transactions.

NatWest's merchant account is called StreamLine. They said that "taking

recurring payments is too risky" so Carson had to go with Halifax instead.

Payment Processor: Ł500 setup fee - someone who takes the transactions.

TOTAL: Ł25,680 (spread out over an extended period of time)

Think about something small that will fund your idea. Carson Workshops is a

side project that helps fund the web apps business.

Took one year to save up the cash. It's OK to take time.

How to build a team on a budget

* Don't go for rock stars - go for quiet talent.

* Offer a small percentage of equity (2% - 5%)

* Ask people for recommendations. If someone ends up being crap it will cost

you lots of money to fix their mess.

* Outsourcing is an option

- Carson tried India, didn't work out

- Might work for you - it'll save you a ton of cash if it does (Ł2,000 for

all development on DropSend)

Scalability on a budget

* Enterprise web apps are enterprise because they have to scale. Initial

intent was blade servers etc - way too pricey.

* Buy just enough hardware to launch (basecamp launched with 1 server)

* Set it up so you can just plug in more hardware when you run out of disk

space - think about it ahead of time.

* Plan, but don't obsess.

How to keep it cheap - some tips to stay on budget

* Don't spend money unless you're absolutely sure you have to

* We wasted Ł1,000 on stupid DropSend stationary - don't do that!

* No new shiny machines - you don't need that MacBook Pro!

* No luxuries - company car, fancy meals - wait until you have real cash

* No "froo-froo" features - "what if it did this; that might be useful"

The more ideas you have, the longer it will take to develop, the more

money you have to spend

Launch ASAP to get the money rolling in - and keep it simple and easy to

use as a side-effect

* Less features is more simple (probably)

* Anything over Ł25, double-check you need to spend it

* Make deals

Offer free advertising on personal sites, build small sites for service

providers

* IM, not phone calls.

* Do as much as you can yourself - wireframing, morketing, testing

Wireframing tip - Use Flash. Then watch your friends click on it. Give

them beer and pizza.

Bookkeeping and copy writing

* Get friends to help out

* Shop around - first hosting offer was 12 grand a month!

Pessimism has its place!

* You WILL go 10% over budget and 3 months over schedule.

Holp crap lawyers are expensive! You need:

* Terms of service: Ł1,000

* Freelance contracts: Ł800 (you need these, even when working with friends).

Get one made, then change the names.

* Privacy policy: Ł15

Clickdocs (very cheap legal documents) http://www.clickdocs.co.uk/

* Barter with your lawyers! Every lawyer needs a website.

* Get a free 1 hour consultation - from as many different lawyers as possible.

Cheap software is your friend!

* Project management: BaseCamp

* Bug tracker: Trac

* Meetings: Skype and AIM

* Version control: Subversion (plus Tortoise SVN

http://tortoisesvn.tigris.org/)

* LAMP

Get a Ł200 Linux dev server for your office - if you can't afford anything

better.

How to not spend money on marketing

* Dont spend money on marketing to start with

* Blogs, word of mouth, is your web app viral?

DropSend emails include an unobtrusive note explaining DropSend

* Writing for magazines in your customer's sphere of interest - a great way to

raise your profile

What about Venture Capital

* You might need it (if you need to scale or expand quickly)

* But these days, you need a really really good reason to go that route. If

you can build it for 30 grand, why give up 25% of your company?

If you don't remember anything else, remember this stuff:

1. Don't spend money inless you absolitely have to

2. Barter services to bring down costs

3.Cut cut cut features so you can build quickly

4. Be realistic, if not slighty pessimistic

5. Plan for scalability but dont obsess

Contributors:

Simon Willison - http://simon.incutio.com/

Gareth Rushgrove - http://morethanseven.net/

Drew McLellan - http://allinthehead.com/

Tim Beadle - http://www.timandkathy.co.uk/ & http://www.iop.org/