Bare Naked App

As you probably now know, Carson Systems is just two full-time employees right now: Gill and I. For the past two years, we’ve partnered up with amazing freelancers to help us out with everything from design to Linux system administration.

Well, after two years, I was getting kind of sick of the hassle of finding freelancers. Let’s face it, it’s a ton of work and it’s a pain in the ass. Here’s the 5-step process:

1. Put together a shortlist of talented people
2. Ask those people if they’re interested
3. Ask those people if they’re available
4. Ask those people how much they’ll charge
5. Pick one

Argh… It would be so much easier if we just had a fulltime developer who could just get working. Why do I mention this? Well, we were really close to hiring a fulltime developer and then we had a conversation with Jason Fried of 37signals. He was explaining how it was a full year until Basecamp was bringing in enough cash to pay for a developer.

Then it hit me. DropSend is doing great, but it hasn’t reached a level where we can justify hiring a fulltime developer to look after it. We were looking to hire a developer just because Carson Workshops was profitable enough to enable us to afford one, not because we really needed one. Bad Ryan! Thankfully Gill realised this and challenged me on it. We wanted a fulltime developer because it would be convenient, not because we needed one. When you’re a small company like us, you can’t justify spending large amounts of cash unless it’s vital to your success.

So … we decided to go the freelance route.

As we only have one web app (DropSend), we don’t need a full-time designer at Carson Systems (yet).

So when we decided to build a second web app, we knew we would be hiring freelancers to do the design. We also knew that the design is one of the most important parts of the project. Why? Well, if the user interface is poorly designed, no one will use it.

We decided to create a shortlist of designers that we wanted to work with. We came up with three names. These are people that wouldn’t be absurdly expensive, but they really know their stuff. They are talented in design, usability, accessibility and standards - all vital things in a good web designer.

Once we put together the list of names, I sent out the following email:

Hey XXXX, XXXX and XXXX,

It was wonderful to see you all at SXSW. It was our first time and Gill and I had a blast.

I mentioned to you all that we’re about to begin our second web app. We’re looking for great designers and we’ve narrowed it down to you all.

Before we get too far down the road though, we need to talk about money, so we know if we can afford you or not.

Here’s what we need doing:

1. Branding (logo, colours, tone, but no letterhead/cards)
2. Site design (public facing site, about 5 pages plus a WordPress skinning)
3. Web app interface (about 30 - 40 screens)

For DropSend, all of this took about 4 weeks work, full time. However, it was a *much* larger app (about 100 screens). Based on this, we think all of the above will be about 3 weeks (15 days) fulltime work.

Can you give us an approx figure of what you’d charge for this? Please be aware that we’re not picking based purely on lowest price.

Once we hear back from you, and we know we can afford everyone, we’ll tell you the specifics about the app, and ask for a very basic pitch.

If you could get back to us right away, I’d totally appreciate it. We’d like to have the designer picked and ready to work start on March 31st.

Take care,
Ryan

One of the candiates got back with some questions, so I sent this clarification:

Hey guys,

XXXX had a couple great questions, so I thought I’d respond to all of you, just to clarify things:

By ‘Site design’ we mean Photoshop/whatever comps. We have our own CSS/XHTML ninja that will take your designs and execute them. He’s amazing (and probably cheaper than you guys!) :P

The WordPress skinning is in addition to the 5 page site. Again, we just need comps from you on this, not actually building it.

When I say 30 - 40 pages of UI for the app, I mean that they’ll all be fairly similar, with just slight differences (different forms, etc). They won’t be 30 - 40 totally different pages.

Regarding the ‘pitch’, we don’t expect you to do visuals for free, so we just want some ideas (text) in an email for what kind of look-and-feel you’d propose.

Again, please don’t feel like I’m going to hold you to the price you quote. It’s just an estimate so we can make sure we can afford you. Once we choose someone, we’ll give them the formal brief and then we can modify the price accordingly.

Hope that clarifies everything.

Best,
Ryan

After I sent that out, we received the quotes which were:

Designer #1

$5,000

Designer #2

$11,700

Designer #3

$8,000

It was really hard to choose between the three. They all have great experience, nice personalities and strong work ethics.

As with many of these type of decisions, it came down to a gut feeling. Who did we think was a good fit for us? We need someone who is both trustworthy and fun to work with.

Basically, we decided on Designer #3 (Jason Santa-Maria) for following reasons:

1. We spent a lot of time with him at SXSW and really felt like we could trust him.
2. He expressed a real interest (over and above the others) of working on the project, so we knew he’d do a great job
3. His quote was a little higher than our budget, but was workable
4. He’s a nice guy and will be fun to work with

It’s not an exact science, but hey, that’s how these things work!

(Thanks to Dave Shea for the photo, by the way)

Once we decided we had found a great idea (hopefully!) we had to figure out how much it was going to cost.

Thankfully, we have experience building a large web app because we finished DropSend in December 2005, which is great for learning how long things take and how much they actually cost. By the way, I’ve written in-depth articles about the costs of DropSend on Signal Vs Noise.

Gill and I were having breakfast at Magnolia Cafe in Austin, during SXSW and we started chatting about the new app (which doesn’t have name yet) - specifically the budget and timeframe. How soon can we crank this baby out and how much will it cost? Can we afford it?

So we divided the costs into these categories and put approximate figures on them:

Design

This involves branding, public website design, product blog (WordPress as always) and web app UI. This took about 4.5 weeks on DropSend, so we figured it would be less for the list app, as it was simpler. Gut feeling: 3 weeks at £250 per day = £3,750 ($6,500).

Coding

After DropSend, my gut feeling was that it should take around 4 weeks to code the new app. This is also based on my previous experience as a developer. If we got a decent developer (but not a rockstar) to code it, I reckoned it should cost us around £6,000 ($10,400). This is both the PHP and the XHTML/CSS. After the coder is finished, we would probably put them on a retainer of around £250 per month for updates and changes

Legal

Thankfully, we’ve already had general contracts written up for freelancers, for DropSend. However, we still need a Terms of Service drawn up for the new app, which should be around £1,000 ($1,740)

Miscellaneous

As always, there’s always miscellaneous crap that crops up. We reckon about £500 should cover it. This can be anything from extra phone calls, flights, meetings, hardware, etc

Hosting

Thankfully, all we’re going to need for the web app is a simple dedicated LAMP box. This should only cost around £100 ($170) per month

Total Cost

So all in, it looks like the app should cost us around £11,350 ($19,700). We can currently afford this cost (based on the ever-important Carson Systems cashflow Excel sheet), even if it fails and earns us no money.

The truth is, things always cost more money than you think. In order to allow for costs we haven’t thought of, it’d probably be safe to add another £2,000.

Total cost: £13,350 ($23,250)

Our problem is that if we don’t set out exactly when and how things are going to happen the likelyhood is that they won’t happen. Or at the very least take a lot longer than it should do. So we need to pin down exactly how long it will take us to build our new app and set a deadline for the finished application.

We decided to allocate time for each section of the building of the app based on what we had learned from our first app, DropSend. We started at the begining and allocated time for each thing we had to do - we didn’t build in any time for ‘things to go wrong’ but we did add some contingency at the end. So the following represents our ‘ideal’ timeframe for the building of the app.

Usability
Perfecting the wireframe - 3 days
XHTML for the wireframe - 1 week

Design
Designing the web app UI - 3 weeks
XHTML for the UI - 1 week
Coming up with a name - 1 day

Code
Development of the web app - 4 weeks

Website
Writing the copy - 1 week
Design of the pages - 1 week
XHTML for website - 3 days

Extras
Contingency and Launch - 3 weeks

Total = 15 weeks

Firstly we gave ourselves a deadline. We decided that we must have a designer and developer in place, with contracts signed by 31st March. This would enable the whole process to begin.

Once we had the timeframe set we could go ahead and set milestones in Basecamp so that we knew what should be done by which date. Because some of the work could overlap (ie the coder could be working on the development while the designer was working on the website) the whole process would not take 15 weeks from beginning to end but nearer to 11 weeks.

We set these deadlines:
31st March - wireframe perfection finished
6th April - XHTML of wireframe finished
10th April - site copy written
13th April - changes to XHTML of wireframe finished
13th April - developer and designer start
27th April - UI design finished
27th April - XHTML of UI starts
27th April - website design starts
4th May - XHTML of UI finished
4th May - website design finished
10th May - XHTML of website finished
1st June - Code for app finished
22nd June - App launched

Sounds simple doesn’t it?

Once we had come up with the idea, we wanted to get right into deciding how the app would work. We’re big fans of 37signals’ Getting Real process, which worked great on DropSend, our first app.

Gill and started sketching the user interface in my sketch book on a plane to NYC. It was funny because we were paranoid about people overhearing us, so we had to whisper. The above is a photo of our real sketches.

We started with the signup page and hammered out each page. Thankfully, the app was very simple (about 20 screens). We didn’t bother to sketch the admin that Carson Systems would see - just the customer user interface.

We didn’t go into any sort of detail - just the basics. Once we had the pages all sketched out, I refined them in PowerPoint. I know, it’s weird., but it works great for me. I find it quicker than Fireworks, Flash or Photoshop. You can quickly add pages and drawing boxes and writing text is super simple.

After I created the PowerPoint document, we re-visited it about 1 week later to see if we had forgotten anything. Gill went over it in detail and we changed a few things.

Now that we’ve got the wireframe all done, we’re going to have it build in unstyled XHTML to get a real feel of how it will work.