In July 2009 an Atlanta rental car company asked me for a second opinion on some logos somebody else had drawn. I gave it, and then I took the job over, and thirty two months later I had designed the brand, built the website, written the tagline, specified the airport sign, wrapped the van and shipped a car rental booking engine inside a Facebook page.
The company was a rebrand. There was an operator at the airport already, and the new owners wanted a name that did not sound like a local one. What they got was iTravel.
I was not hired to design the mark
The first concept is stamped July 24, 2009 at 12:23pm, one day after the first email. I drew four directions and none of them was a tweak of what I had been shown.
The name and the direction were settled by the client in a single written instruction on August 13, and I had it executed four days later. The tagline came three weeks after that. I sent it at 6:54pm on September 7 with no preamble:
What do you guys think about this for a tag line (and I would include an icon for each): Point. Click. Drive.
LOVE it!
I like it.
That is the whole approval. Three lines in half an hour for a tagline that then went on the building.
A competitor could not work out who we were
Five days later the client lead wrote to tell me the banner had gone up at the airport while the counter was still being built, and what had happened next.
Apparently, the logo in banner form has been put up awaiting the CONRAC opening and the actual backwall being installed - - and it has created a buzz. An executive from Hertz corporate shared at an airport meeting that they could not figure out who iTravel was, and assumed it was a European company.
I have never had a better piece of feedback on a logo. A competitor looked at a brand new company with one banner and no counter, and could not place it. That is the entire job of a rebrand for an operator nobody has heard of.
The empty grey box
The first wireframe I drew has a rectangle in the top left with the words Book Engine in it. It is a placeholder for the one part of a rental website that actually takes money, and the reason it is a placeholder is that it was never going to be mine.
The launch date moved three times. It was set for September 15, then September 23, then a Monday in October. On October 16 the client made the call that decided the next two years: take the booking engine off the home page and go without it.
The site went live on October 22, 2009 as a rental car website you could not rent a car on. That was the right decision. A brand with a counter and no website is invisible; a website with a broken booking form is worse than one that sends you to a phone number.
Three vendors in eighteen months
The booking engine was never one thing. It was a hosted widget from whichever reservation company the business was using that quarter, dropped into my page through a script I did not control, and it changed three times between 2009 and 2011.
Each swap was a real engineering event and none of them was my decision. In January 2011 I was given a day’s notice on the last one:
Will you be able to switch out the book engine after 12AM this evening?
I can make the switch, however, we will be without the booking engine on the home page. Instead, I’ll have to create a button to take it’s place, which will link directly to the [vendor] booking engine.
I have left my own typo in that quote because it is what I sent. The point of it is the sentence before the however: I said yes, and then I said exactly what saying yes would cost, in writing, before it cost it. What it cost turned out to be the best thing on the site.
The booking engine inside Facebook
In April 2010 the client wrote down an ambition rather than a requirement:
I would really like iTravel to be the first car rental company with a book engine on their Facebook page.
I registered as a Facebook application developer in May, and by August there was a working car rental booking engine running inside a Facebook page. You picked up a location, set two dates and two times, chose a class and booked, without leaving the tab.
We did not get there first. I have never established by how little, and I am not going to claim a margin I cannot evidence. What I can evidence is that it worked, that it ran for about five months, and that the January 2011 vendor swap is what ended it, because the engine it wrapped stopped existing.
The lesson I actually took from it is not about Facebook. It is that the most interesting thing I built on that project was a shell around somebody else’s booking system, and when the business changed booking systems the interesting thing evaporated in a night. I have been suspicious of building on a widget ever since.
What I would not build this way again
The site had a free loyalty scheme called iTravel Preferred. The signup form asked for a driver’s license number, a date of birth, the issuing state, an employer and two phone numbers, and the table stored the password next to all of it in clear text.
There was no points ledger and no member number field anywhere in the database. The page told members to enter their membership number and promised they could view their points. Neither of those things existed in code. Numbers were issued by hand and points were a promise.
Every one of the three signup forms was protected by the same thing: a pre-rendered image of a word, compared against that word hard coded in the handler. One word per form, never regenerated. It stops nobody who submits twice. And every database call on the site, including the logins, built its SQL by pasting strings together.
I am not going to dress that up. It was normal for a small classic ASP build in 2009 and it was wrong then too. The reason the exhibit below runs on invented rows is that the real table held all of that about real people.
What the client said at the end
In January 2012 I asked the client lead whether the numbers she had sent me were any good, because I genuinely did not know what good looked like in that industry.
Ancillary sales (Misc, RCP and RH) represented an additional 31% revenue on top of t&m.
Repeat Customers - 130 customers rented two or more times in 2011. That’s so impressive.
That was a fleet averaging nineteen cars a month. A hundred and thirty people who came back to a rental brand that had existed for two years, with no third party booking channel sending them.
A month later she sent a written recap of the whole engagement. It opens:
We came to Atlanta with a goal to build a new model for car rental that could also be adopted by other travel related providers - - a direct relationship bypassing the more costly online and offline distribution channels. We accomplished that and more.
The site moved to the owner’s own server in March 2012 and the engagement ended there.
A footnote about the name
In April 2010, eight months after the brand launched, Apple filed a patent application for a travel product called iTravel. The client forwarded me the trade coverage the day it broke. We tracked it as a brand event for another two years, and in 2012 she was still explaining a traffic spike with the line that some of those visitors were probably looking for Apple.
It still runs
The whole site is back up and you can use it. One hundred and twenty three content files, the deepest structure I ever built on my own front controller engine, measured against every other project in the archive. Not one corrupted file in the tree, which is rarer than it sounds.
Dates here come from the files, not from memory. The logo concepts carry Illustrator’s own XMP timestamps, the layouts carry Fireworks timestamps, and one comp is named as a Dropbox conflicted copy from June 27, 2010, which dates itself. The engagement’s own emails supply the rest. The design sources are layered Fireworks documents and stay unpublished; what is shown above is flattened from them.