Intuit recently held a training session on the nature of Disruption, and I wanted to note two disruptions that I believe are happening today.

The first is fairly well known: solid state disks. Companies like SanDisk are making larger capacity memory cards every year. There are even 256gig solid state drives available now.

This disruption is following the classic pattern - coming into the market by providing memory for devices that cannot use hard drives - e.g. keychains, cell phones, cameras. I have an old Sony camera that uses a 3.5″ floppy drive of all things. Solid state memory is clearly the better solution.

But it isn’t good enough for bigger computers. Not good enough even for laptops - too expensive. However, for the newer, small Netbooks, it IS good enough.

Big drive makers are chasing the larger capacity market, but they’ll eventually be niche players for the very large capacity systems, while the mainstream will be switching to solid state. Seagate was smart to invest in SanDisk early on.

The second disruption is just beginning, and I haven’t seen anything about it - although I doubt I’m the only person who recognizes it. This disruption is Cloud Computing, more specifically, it’s offshoot Utility Computing.

Utility computing isn’t going to disrupt the personal computer any time soon, but what it will disrupt is the traditional corporate data center - big air-conditioned rooms with racks and racks of servers delivering one or more applications. Why hire a big technical staff, pay for facilities, add and maintain hardware when you can just outsource the entire thing with a single click? Amazon AWS is a big player in this space right now. You can get one-click deployment of your application, use practically unlimited storage, and even speed up the delivery of your app through Amazon’s Cloudfront - putting the data geographically closer to your customers.

Why would anyone attempt to implement such an infrastructure for themselves? If services like Amazon AWS can manage to guarantee the security, privacy and ownership of the applications and data that exists on their systems, they will quickly eliminate the corporate data center.

I’ve recently been in some meetings… and have been invited to meetings… that lack some of the characteristics of making a meeting successful, and have even “violated” espoused tenants.

So as I was talking with Jonathan in the iLab, we came up with what the Successful Meeting (of the Future) will look like.

First, let’s start with the Invitation. I’m never added to the “required” list unless I’m really required. I’m usually disciplined enough to decline meetings that I don’t think I need to attend. If I’m listed as “Required” and I decline to attend… would you reschedule the meeting? If not, then I’m not really required, am I?

Outlook is a tool, and it’s really easy to use, right? But it’s too easy to just add everyone to that one field. Don’t. Use the Optional field. Ask yourself, what happens if this person can’t attend? If the answer is, “Nothing, they’ll just have to be informed later”, then DON’T add them to the Required list - put them in the Optional list!

Don’t assume everyone knows what the meeting is for. Give it a good subject line. I recently had a meeting invite with an “out of the blue” subject line - I had no idea what it really meant. Turns out most people receiving the invite would know what it was about, but I wasn’t in the, well, the other meeting where they discussed what they’d be talking about in that meeting… er… the meeting I received the invite for… I was in another meeting.

A good invite should be detailed enough for me to know if the meeting was a success. That means it needs to have an agenda and a desired outcome - even if that outcome is just “share information”. Is there a decision to make? Then state it in the invite.

The Meeting of the Future… or should I say the Meeting Tool of the Future, will let me define what kind of invites I will accept. For instance, Outlook will currently ask if you really want to send a message without a Subject. The invitation you create will see that it’s going to “Matt” or “Jonathan” and will inform you that they won’t accept invitations that don’t contain the words “Agenda” and “Outcome”, so please add those to the invite. They will also automatically mark “tentative” any meetings with more than 10 people in the Required list.

Second, we have the meeting itself, usually accompanied by a PowerPoint presentation, so let’s talk about Successful Presentations.

There are two types of PowerPoints: the kind that are meant to be “read”, sort of a replacement for a document or a wiki or whatever. They can be emailed and generally absorbed without a person standing up and telling you what they are about. They can be filled with graphs and text and appendices. DON’T present these kinds of PowerPoints. DON’T put snappy animations in them. If they can stand on their own, then use them that way.

For the Successful Meeting, you need to make a whole new PowerPoint. Do not use the one that is meant to be read, and then just READ the stupid thing in the meeting. The people in your meeting probably know how to read. If you stand up there and read the fifteen bullet points on the slide, you are going to BORE us to tears. There must be some other reason we met in person than to just read a slide deck. PowerPoints should be used to get across one, maybe two points. Three if you are a great presenter. If you’ve been reading slides lately, stick with one or two points. That’s all people will remember no matter how many sales figures you read off the slide. Every slide you show should target your point… or points. Don’t read them. Show them and then talk - they should somehow emphasize or otherwise support what you are saying - they should NOT “say” what you are saying!

Meetings cost money. They cost productivity. They can drain energy. They can also lead to making money, increasing productivity, create energy. Always, always think about the pitfalls and steer the meeting around them.

Take the first one: meetings cost money. Take the number of people in the meeting, the average salary in your company, and come up with a cost for the meeting. We were recently invited to a 34-person meeting, for 1 hour. This meeting will be quite costly. That’s not necessarily a bad thing, but it should result in somehow making money… or at LEAST people should realize that this meeting is costing money.

What if the presentation had a ticker going, showing the cost of the meeting as time went on? It would be great if we could have immediate feedback regarding the cost of the meeting.

I’ll take the next two together: productivity and energy. Did any of you see that Simpsons episode where the kids where asked to indicate their emotions as they watched an Itchy and Scratchy cartoon? We want those devices for meetings, so the presenters can receive immediate feedback on audience engagement.

How would that go over? As children, we learn to pay attention. A big whack on a table with a yardstick will probably wake you up. But as adults, we only pay attention when we are interested. We have a lot going on, we have distractions, and we need something interesting before we can engage.

I don’t know if I would like it, or how I would react in the middle of a presentation. It would be difficult to “switch gears”, I think, if I saw audience engagement and energy going down. That’s why I personally present “stories”, not facts or figures. People can come to their own conclusions, hopefully matching the one you want to tell them at the end, if they are carried along the same storyline that you have experienced. People are pretty good at knowing what the next page will likely say.

If people aren’t getting something good out of the meeting, then their productivity for that time has been wasted. Their energy, potentially for the rest of the day or at least a period of time following the meeting, is reduced.

In the Successful Meeting of the Future, I will be truly required if the meeting cannot go on without me, I will be properly optional if it can, the subject and body of the meeting will set my expectations with a proper description and agenda. The meeting itself will have a minimal, or no, presentation slide deck, and will be productive, help me in my daily job, and will increase my energy.

Otherwise, please mark me as optional and just tell me what happened in the meeting. I have work to do.

Intuit’s PR department tells me that MyCorporation is giving away free incorporations, a $150 value, on Nov 11th, 2008, from 6am to 6pm Pacific Time. Use the coupon code FREE149 and tell them Intuit Labs sent you!

Amazon.com is rolling out an innovation called “Frustration-free packaging”. It greatly reduces packaging costs and materials while eliminating the need to find a pair of industrial bolt cutters to get your package.

I think one of my first real encounters with package frustration was a 512meg SD Card I bought a few years ago. It came in a box… that contained a heat-sealed clamshell… that was probably 48 square inches in size. All to contain a 1-inch square by 5 millimeter piece of hardware. I thought to myself how absolutely ridiculous it was to put so much packaging around a tiny SD Card.

Then, of course, there are all those Christmas toys, trapped in their containers and hard-wired to little plastic washer-things and attached to their packaging. Opening a gift was always followed by 10 minutes of unwrapping the wires to free the toy. For those really special gifts, I would unwrap it in advance so that my kids could immediately enjoy the present instead of waiting for it to break free of its packaging prison.

Amazon is seeking to eliminate this by putting gifts in simple, recycleable cardboard boxes - no wires, no plastic.

Compare the before packaging needs of a small toy with the frustration-free packaging.

And Amazon’s commitment to user-contribution is alive and well with this effort. You can share your own stories with Amazon.

Otherwise known as, “If you can read this, then you’re on the right server.”

ACT 1
Scene 1: Building exterior, Waltham, Ma. The night was sultry. A bucket truck is pulling into the parking lot. Overhead high voltage lines buzz in the night.

Driver 1: “Gotta love these Dunkin Donuts coffees!”
Passenger 1: “Yep. Say, how tall is this truck again?”

ZAP! The top of the truck hits the power lines. A droning sound fades down as the lights cascade off across the city.

Scene 2: Building interior, Intuit data room. Emergency lights glare as a red-shirted man peers into a rack of computers, pushing buttons as he speaks into a cell phone. The room is eerily quiet, air conditioning systems turned off. A bell dings and a voice is heard.

Computer Voice: Warning. Battery systems failure in 6 minutes. Warning. Battery systems failure in 6 minutes.
Scott(ie), brogue voice: “I’m workin’ as fast as I can, Captain! These laddies don’t turn themselves off, dontcha know.”

The last computer powers down as the voice sounds the 3 minute warning.

Scottie: “That’s the last one. I warned ya we didn’t have enough power in the data core. Ya better call Frankie the Host and get us moved quicker or we’ll be stuck in this time loop forever!”

ACT 2
Scene 1: Building exterior. Another bucket truck is pulling into the parking lot.
Title: “1 month later”

Driver 2: “I sure am glad Dunkin has pumpkin flavors now. Great stuff!”
Passenger 1: “Deja vu… you know, the last time I was here I was talking about coffee with…”

ZAP! The truck hits the same power lines. Sparks shower down as power goes out in the neighborhood.

Scene 2: Building interior, Intuit data room. A voice calls out its warning.

Computer Voice: Warning. Battery systems failure in 4 minutes. Warning. Battery systems failure in 4 minutes.
Scottie: “She canna take much more ‘o this, Captain! Admiral Cook and Secretary Smith are trying to get to us but we’re gonna have to boot ‘em out!”

Scene 3: Office interior. Large PostIt notes are on the walls. A frazzled man in a yellow shirt and dark pants speaks into a handheld communicator in a halting, intense voice.

Captain J: “We need this in 48 hours, Frankie. Don’t make me send over a box of Troubles.”
Frankie (voice on communicator): “Now why would you want to do that, Captain? We’ve got your hosting, no problem. I got my best guys on it right now. We’ll have your parts shipped over and setup, no problem.”

Scene 4: Office interior. A blue-shirted officer and the red shirted Scottie are tapping away at computers.

Scottie: “It’s no use Mr. H, I canna get my messages to go through.”
Mr H: “Try adjusting the domain compensator and restarting the system. That should align the header information and prevent proactive disintegration of the message containment prior to reaching its target.”
Scottie: “I’ll try, Mr. H, but the domain compensator canna’ be fully functional until the data system transfer, I dunno if she can ‘old up under the strain!”

Scene 5: Data room, deafening noise from the air conditioning system. Scottie is holding onto a computer rack as the wind tries to blow him down the hallway.

Scottie, yelling on handheld communicator: “I’ve got it, Captain! She’s ready to go! Test the system so we take this one down!”
Captain J (voice on communicator): “Exc..ell..ent.”

Narrator: “This is the third chapter in Intuit Labs hosting. It started out in a small DMZ inside the data room at the Waltham, Ma offices of Intuit. It had no power backup, no Internet redundancy, little in the way of computer system redundancy.

Then the intrepid crew of Intuit Labs moved to new offices and a new data room, with backup power and a redundant cooling system.

We were executing on a plan to move to a high availability system, but a few truck-hitting-power-infrastructure incidents (one was a pole, the other was the actual power lines…I took a bit of liberty with the facts) moved our plans to the fast lane.

We connected with Intuit data center folks in San Diego, and they brokered a deal with ComplexDrive to host Intuit Labs in a power-redundant internet-redundant facility, using two computer systems - one active and one ready to boot up as the backup, with sometime like a 10 minute SLA for reboots upon failure.

Our computer systems have never actually failed (HP DL360 workhorses) - it’s always been the infrastructure. Now we’re better shielded against those kinds of failures.

Check out a few picture of Intuit Labs at ComplexDrive. As you can see, we have plenty of room to grow!

P.S. Dunkin Donuts is a New England and Intuit Labs crew mainstay. Better than Romulan Ale!

There have been some problems with using Keyword Search on Vista, sort of crashing QuickBooks or somesuch.

I recently installed a fresh copy of Vista, QB 2009, and Keyword Search and it worked great.

I suspect that the problem is browser permissions. Try loading Internet Explorer, accessing Tools/Internet Options, click on the Security tab, select Local intranet and using the Medium-low setting (the default for Windows Vista).

Let me know if that works for you!

When it comes to “process”, the iLab’s modus operandi is “the less, the better”. It’s not that we’re process averse, but we want to know why we are doing it, and if there’s a better way, we’ll try to find it.

So take goal setting. You’ve all done it. How did you feel when you opened that official “performance management form”. Who is your dotted line manager? What is your True North statement? How best are you delivering for the three stakeholders? Have you applied the Intuit leadership model?

“Yargh..,” I thought as I looked at the form, echoing Jenny from a few days earlier. Let me look up my “employee number”, envisioning 7,000 employees trudging down into the depths of the machine a la Metropolis.

We came up with something else - stories! Why not just use stories to tell your goals? “The Adventures of Matt in the 22nd.09th Century” Relive the exciting echominder decisions. Glory in the release of simple hosting for unstructured time teams. Marvel at the shared visions. And so on… fiction, fantasy… it’s all just predicting the future anyway, so tell it like that.

I know, you might be thinking, “But how will we measure goals and expectations against actual accomplishments?”

What, is it already a quantitative measurement? No way, no how - such comparisons are inevitably QUALITATIVE. I submit that using stories to tell your goals will actually make it EASIER to compare to how you actually did.

What do you think about this approach?

I’ve been with Intuit since 2001, joining as part of a buyout from my previous employer. It was a year before I learned of anything “innovative” happening at Intuit. That was when my boss at the time told me about an opportunity to work in the newly-forming Innovation Lab (iLab).

I’ve always been an explorer when it comes to computing, writing applications so that I could learn new technologies, writing magazine articles or visiting customers during the times I consulted. I wrote some shareware applications that were well received.

So the iLab was a natural fit - fast-paced customer-focused investigations. The original intent was that we would transition our projects into a traditional business unit to be fully built and resourced: we didn’t have the capability to develop and support applications.

What we discovered was that the business units didn’t know how to integrate our projects. The first one, Customer Manager, was built without the focus on the most important lever that we identified for success: synchronization - that piece just didn’t work very well, so the product didn’t work well. The business wasn’t ready to build a version 1 that was a “learning” focus rather than a revenue focus.

Another project, Estimator, had its resources pulled to work on something else. It finally saw the light of day a few years later, only to be sunsetted after its first release. It has the same problem as v1 of Customer Manager - a focus on revenue rather than on learning.

Two other promising transitions failed for lack of resources to work on v1 offerings.

So the iLab changed its charter: we became responsible for figuring out how to actually build these things we were designing. It’s a hard task when you don’t have anywhere to even show, market, etc… your offerings, but there’s a way to get there.

The first step is to:

Create some quick wins

I took a couple of things out of projects, or things I’d heard from customers… or just my own wacky ideas… and built them. Keyword Search for QuickBooks was the first. It was initially published on the QuickBooks forums as a beta test feature of QuickBooks. It was definitely a win and is still going pretty strong today.

So now, we had a story: from customer pain to WOW feature. We also knew why it couldn’t be done internally by the QB team, so the story for needing an alternate solution for innovations was also known. And we told it as often as we could in order to get the resources we needed.

And that leads us to a more formal way to show innovation. Internally, the iLab developed wiki and blog hosting, and that led to our own “data center”. It was a minimum-viable implementation, with no redundancy and no 24/7 on-call support staff, no built-in backup system… It was, at least, isolated from the corporate intranet in a DMZ and supported by top directors at Intuit.

So step number two:

Get support from Tops rather than Ops

Our Ops people were peripherally aware of our playground, but they didn’t really have much control over what was going on in there. They could turn us off or on, but that was about it. However, we had the full support of the Intuit CTO (Bill Ihrie), Intuit founder Scott Cook, QuickBook’s VP of Product Development and others. It was the “quick wins” stories that got these people excited about supporting our efforts.

But it wasn’t all fun and games from there. Although we didn’t have what I would call “direct” support from Intuit’s operations teams, we did need them to get where we wanted in terms of a customer-facing playground.

So that’s step number three:

Work within your Ops group processes as much as you must because you really can’t do it without them

For us, that meant using the standard request ticketing system to provision our DMZ with static IP addresses, firewalls, etc… We put in the request, waited for months until it finally reached the Tops who supported us, then received the resources we needed.

We also spent long hours on conference calls with Legal and Privacy and Security folks at Intuit to make sure we were doing the right thing by our customers and not exposing our systems to hacking or legal issues. It took about 8 months, that’s right eight months, to put in place our EULAs and site terms agreements, evolved versions of which are still here on Intuit Labs.

That’s step number four:

Be persistent: talk to all the people you need to for as long as you need to

Your company really does want to innovate, they just might not know how to innovate with “web speed”. Intuit was a desktop software company, and is still in the transition period between desktop and web. Yearly release cycles and predictable capacity needs (e.g. tax season) are very important to such a company, and throwing something wild into this mix must be done cautiously.

So, now we had a new innovation website, innovation.intuit.com (now Intuitlabs.com), with some applications on it. COME ONE, COME ALL, PUT YOUR STUFF ON OUR WEBSITE!

Can you hear the crickets? We had a push from the business units to release our website quick! We have dozens if not hundreds of people wanting to put up their apps!

Pleep. Pleep.

We thought that these 4 steps would be enough, but it wasn’t.

We needed to emphasize innovation inside the company, not just “enable” a few mavericks

And that turns out to be the key lever to creating an innovation culture. Yes, you need an outlet for that innovation, and you probably need to be building that first. Yes, you need a few quick wins to get upper management on board with you. Yes, a couple of mavericks is probably all you need to get started.

But without real process change internally, you won’t see the culture of innovation. What Intuit, or rather those Mavericks, did to enable this was actually two more steps:

Give people “free” time to work on whatever they want

Intuit gives everyone 10% of their time to work on anything they want: could be an innovative offering, could be just learning a new technology - doesn’t matter, and you don’t have to report on it or get approval to work on it. That was the first thing we did, and it seemed to work… a little bit. More people were turning into “Mavericks”, but it wasn’t enough to really kick-start the culture revolution we really wanted. But we had an answer for that, too.

Innovation days, Idea Jam sessions, whatever you want to call it: dedicate a whole day every 3 or so months

Get people together and let them compete for prizes. This is the second big lever and it creates excitement. People win money (immediate delighter) and they can win time to work on their ideas. Talk about it, market it, report on it, video it, blog post (internally) as it’s happening.

We’re still figuring out what it means to “get your idea out to the public” - what we really want is to see idea iterations that lead to big new businesses, and we’re evolving our website and our process to make this happen.

But that’s the Intuit story: Some quick wins that led to senior management support, which meant operations helped us get resources while legal issues were worked out. Once we had the path to publishing in place, the final trick was to give people time and incentives. The innovation culture is now going strong at Intuit and we’re well prepared to create some new offerings that might evolve into big businesses.

If you’ve been reading my posts, you know that lately I’ve been obsessing over backup. I’ve lost hard drives, and so I’ve been pretty dissatisfied with backup solutions. The only ones I’d found that worked well have been dedicated hard drives - plug in a USB drive and run its backup software, like WD Backup on an eBook. But when you have lots of computers, plus you don’t really want your backup to run on a drop-able/crashable/loseable monster hard drive.

I finally heard of Carbonite.com, and it’s a service that I’ve been wanting to develop myself, off and on, for the past few years. Backup your files, right click a file or folder and back it up, simple file-based recoveries.

None of that wacky incremental backups where it’s hard to recover a file. None of that “it doesn’t work when the wireless network disconnects and reconnects” stuff I wrote about earlier. It simply tracks new and changed files, and it backs them up.

Wonderful - it works great, it’s priced right ($49.95 annually per computer), and I’m planning on rolling it out soon to all of my managed systems: about 10 of them. $500 per year for peace of mind sounds like a great price to me.

Let’s face it… there are some things that humans just do better than machines or software. Despite our years of effort, software still has a hard time understanding what you are asking it to do.

So we have filters, sorts, exclusion fields… I can’t, for instance, tell anything what kind of news I want to see… or ignore.

I’ve thought about the idea of “anti-tags” and even mocked up some applications. After all, it’s much easier to tell some search engine or application about the things you absolutely do NOT ever want to see. If you don’t drive, why return automobile results or ads?

So in the realm of understanding, I think computers might be able to do a pretty good job of being an information filter. Over time, you could anti-tag stuff, and it wouldn’t show up.

That is, it won’t show up if the human cloud does it’s thing.

You see, only the human cloud can actually define what “IT” is well enough for the computer to filter or find or whatever… Yes, I’m talking about the semantic web… but I’ll also get to other human cloudy things, so skip the next few paragraphs if you are sick of reading semantic web blog posts.

The semantic web is basically people trying to define what a particular piece of information on the net means, e.g. “This page is about…. something“. I can point you to some recipes I wrote on grouprecipes and can definitively tell you what the page is about. Anyone reading that page would probably agree: Simply New England Clam Chowder is about, well, a simple recipe for New England-style clam chowder. Or at least let’s all pretend they would agree on that, even without the title of the page.

There’s a bunch of humans on a bunch of websites trying to tell the web what it’s serving up. del.icio.us is the best example, imho. I often find myself searching for some obscure technical thing on Google, only to come up empty-handed before mentally slapping myself on the head and going over to del.icio.us. Sure enough, I enter a few keywords (aka ‘tags’) and get exactly what I need.

That’s because somebody on del.icio.us, somewhere, thought about a particular web page in the same way I think about it when I see it. So… a human cloud - people on the net - doing what a machine cannot.

But the human cloud is far more than that - it’s also a real cloud of humans you can point to. What makes Wikipedia work so well? The thousands of people contributing to articles? Yes. But that system would fail without a small cloud of humans doing what a computer cannot: ensuring the integrity of the system. There are dedicated folks monitoring every update and submission, restoring vandalized pages, adding “needs a reference” marks, and otherwise learning about a particular topic to the point where they can spot even complex mistakes.

The human cloud can also be bought, like a time share. Purchase a slice of human computing time. Amazon’s Mechanical Turk is a process (yes, and an application… but really it’s a process) whereby you can order up a human to accomplish a task, and then pay that human. Anything from complex tasks to simple ones, you can throw your problem into the human cloud, and get a human response. Now, humans make mistakes too, and you can weed out the troublesome raindrops in the cloud, but for certain complex tasks, humans are, for now, indispensable.

Would you want your final essay reviewed and corrected only by a computer for grammar and spelling mistakes? Microsoft Word does a pretty good job at catching spelling and grammar mistakes, but there’s no way it’s going to catch all of them, for every essay, all of the time.

If you call 9-1-1, do you want a computer to transcribe your call and automatically dispatch the appropriate response team? I doubt it - although there are applications for general transcription. Humans simply do the task much better.

So, here’s why I’m thinking about this problem: computing time in the human cloud is for sale, and it can be monitored, quality controlled, iterated upon, and maybe even commoditized. I think that’s what Amazon’s Turk is trying to do. It’s the human cloud platform, looking for developers to find human problems to solve (we actually thought about using Turk for echominder transcriptions).

Also, a presentation from our CEO got me to thinking about this problem in a new way. He showed the evolution of the “platform” from DOS to Windows to Desktop to Web to Mobile. I started thinking… so what’s after mobile? I landed quickly on a topic I’ve often written internally about, Agents - smart applications that take actions on my behalf. There aren’t many of them out there, and they don’t do very important things. But… why not? Can’t I tell an app to go buy a new music stand for me? I bought one today - but an agent should be able to know that: A) I already have a stand (flimsy fold up), B) I use really nice ones in the band at church, C) I must be wanting one of those kinds, D) The range in which I spend - usually as cheap as I can get it, but not always, E) I check the reviews and usually buy based on them.

I think the human cloud could probably be used to create some incredible, almost magical software agents. And the more that your software could learn from the human cloud, the more it can figure out for itself.

What else can the human cloud do? And can it be done cost-effectively? How else can you tap into the human computing cloud? For what tough, worthwhile problems can we engage the human computing cloud?