http://gigaom.com/cloud/cloud-computing-fiascos-lessons-from...
The key thing here is really to address two things that seem to come up a lot:
1) Cost items that can vary considerably and where a mistake -- code bug, forgetfulness, sloppiness, whatever -- can leave you with serious sticker shock. (At one company we were at, an engineer spun up a couple hundred EC2 instances for a map/reduce job and forgot about them, racking up an extra $100,000 that month.)
2) Cost items that don't vary much but which have such a low friction to signing up that you wind up with bunches of accounts across an organization without realizing it.
Finance departments already tend to have the software purchase cycle fairly well managed, but SaaS/PaaS/IaaS doesn't fit nicely into those procedures. Because of that, we're building our tools to play nice with others, so whether your finance group is managing things in Excel or you have a fancy SAP based system we're working to make it easier to get this data where it needs to be.
This kind of tool would be aimed at precisely the market that would never pay for it.
If employees need to circumvent company policy to use Basecamp, the only way in which such a tool would be used is to identify culprits and get them back in line.
edit: on the other hand I would love a startup to prove me wrong :)
Scraping the company mainframe for billing data: Increased rev by $1 million fixing accounts incorrectly billed.
Compiling the scrapped data into MySQL: $3 million in lost inventory found.
Waiting 6 months for a site visit so I could tell the COO directly that any employee could dump all the credit cards in our database: Priceless.
If you're circumventing finance, then we may not be a particularly helpful tool for you.
Eventually, the whole thing came to light and the company was forced to admit how stupid they were being.
See, they had paid for a year's service of the bad tool, and they were adamant that they get their money's worth. They were absolutely blind to how much money they were wasting daily. The sunk cost was sunk. There was no way to recover from that, and the tool just couldn't be salvaged.
Subverting the bureaucratic channel at a large corporation to "get on the cloud" must drive the CPO and CSO nuts. There is a time and place for bureaucracy.
CIO's job is to manage massive resources and organizations - first and foremost to keep the critical line-of-business systems running. For most of them cloud right now is neither the top priority nor a viable substitute for core infrastructure. You are right on with your strategy to get foot in the door to circumvent their slow-moving processes. However you should be thinking of how to work yourself into their higher-level planning process instead of positioning yourself as adversary of their corporate policy by calling it stupid.
Just remember, the CIOs have the power to ban technologies that sneak out through the back door. Alienate them at your own peril!
EDIT: Whoever down voted this, please post an actual substantive rebuttal.
I agree about the general idea of not throwing stones into the sky. They have a tendency to fall back down on one's nose.
However banning tools that were 'snuck in through the back door' because of their efficiency and because nobody had the heart to follow the actual procedure to get them through the front door strikes me as profoundly childish. Especially if it is done as some sort of vengeance, or 'just because we can'. In fact about just as childish as throwing stones into the sky.
I'm being a bit idealistic here. There will always be turf wars in BigCos. But hey if some people switched from trying to look like the best and brightest to getting stuff done instead, i'd be all for it.
There is no place for idealism in selling to BigCos. You have to work through the process and build your case or someone else will steal your deal, possibly with an inferior product.
Salesperson: This software is so easy and ... CIO: Won't it be a big learning curve Saleperson: Well, actually, your company is already using our software
What you pitch them is on buying larger license for the whole enterprise, increasing the usage and integrating deeper into their systems.
Getting from a handful of users to thousands is easier to do when you have both initial end user champions and the support of IT.
http://everythingsysadmin.com/2011/06/avoid-using-the-term-c...
