BA 850
Sustainability-Driven Innovation

 

Iteration

PrintPrint

A Parable of Turkey

graph of turkey temperatures
Via Andy James Thanksgiving

In the personal realm of "offering development," I am somewhat renowned for my Thanksgiving cooking, and my planning and discipline in doing so. I'd like to share a story that might frame iteration a bit.

I believe that some of this began in my formative years under watching my grandfather who was a retired High Colonel of the Air Force, a rose gardener, and a cook. I feel I extend the artform through my "maps." Where he would use a post-it note, I would use Excel, plotting each food item of the day on the Y-axis and time in 10 minute increments on the X-axis. Notes on tasks would go in each cell. This would be my ongoing salute of anal-retentivity to The Colonel.

Fast forward to 2014, my most ambitious Thanksgiving to date, hosting and feeding eleven with everything from cannoli shipped from Boston to exotic cheese pairings. The initial food preparation would begin early the morning before for the breads, casseroles, prosciuttos and sides, as well as the honey brining of the star of the show: a 24lb organic, free range, heritage breed turkey.

In the USDA turkey sizing scale, this would fall somewhere between "Extra large" and "Hedonistic." This specimen would be prepared in the smoky convection of my Weber Summit grill, starting at 500 degrees and tapering to 325, as all its gobbling forebears. Delicious, moist, honey brined turkey, followed by a mild tryptophan coma for all.

Thanksgiving this year would be terribly cold, to the point that I would have to place a warming jacket on the propane bottle to keep it flowing. As the grill hit 500, the turkey was placed with care, the digital thermometer probe placed in the turkey, and the display facing the nearest window. The lid was closed.

As I would watch the trusty grill temperature, it was climbing as expected after adding the turkey, but never passed 350. I thought that the sheer mass of the cold turkey paired with the frigid conditions and the cold propane had tapped the BTU output, as all burners were maxed. Nonetheless, I had planned for such occurrences, and had 30 minutes of "coast" I could add to account for the turkey progressing too slowly or quickly.

After the first hour, I went outside to check progress, and noticed that I could feel the grill's heat as I approached. Something was not right.

Due to the massive turkey, the temperature probe inside the grill cover had ever so gently rested on the cotton twine binding the "feet," conducting heat away from the probe and leading to the temperature showing as about 250 degrees too low.

My turkey had spent its first hour at a blistering 600 degrees, not 350.

Despite the foundry-like conditions, the turkey still looked quite good, but was progressing far too quickly. I immediately cut the temperature to 250, but we must remember that any turkey, let alone this heathen, responds slowly to temperature.

Knowing that it could be too late before I could adapt, I got out a ruler and a pencil and drew what you see in the image above: an impromptu time-temperature graph.

Because a parameter had changed out of my control, my only action would be to understand the new condition, measure, and respond appropriately.

My goal would be to adapt to ease the turkey to a 163 degree internal temperature at exactly 5:00, at which time I would remove it from the grill, bring it in, and allow it to naturally stabilize to 165. It would do exactly this, as you can see from the original graph.

Applying the Parable to Offering Iteration

I offer this story as an example for what you will likely encounter in your microtesting, and the actions we can take to test and iterate. You will see that the offering is not progressing as planned, that the relationship of your inputs does not seem to be bearing fruit in the measurable outputs.

You will find that while there is movement in the offering, it seems to be "decoupled" or "distorted" from what you are doing. Answers will not be clear. The "thermometer" may not progress as you anticipated.

In these cases, consider doing something I think of as "isolate and iterate," which is simply an application of the Scientific Method as applied to iteration.

Don't attempt to solve everything at once, but simply attempt to isolate one factor in the offering, iterate it, and test that variable. Proceed slowly and with purpose. Give the results time to develop. Test methodically, as the worst thing you can do is begin to drastically change multiple variables in the offering: you will simply mute any "signal strength" you had before from the problem, making it that much harder to improve things.

As we have covered quite a few times in our time together, this is just another example where innovators, typically seen by others as the "mad scientists" of the organization, are extremely measured in their strategies and responses.

Mike Cassidy on Product Iteration

Mike Cassidy, a VP in innovation and product management at Google and a serial tech entrepreneur himself, has some excellent ideas on the importance of iteration over "masterful knowledge."

The following video (set to play from 1:04 to 4:38) summarizes his seemingly simple approach to iteration, as well as a cooking question he asks when interviewing. Please watch the selected approximately 3 minute-long section of the following 11:42 video.

Transcript of Mike Cassidy of Google on Product Iteration

NINA CURLEY (WAMDA): We're here with Mike Cassidy who is the director of product management at Google and the founder and CEO of four startups prior to that. Mike I just wanted to ask you about, um, you typically advise that speed is the most crucial element in a startup success and I want to know is having deep experience in a market really necessary for implementing speed in the development your company? You say that you know hiring known talent quickly is important, and knowing what kinds innovation will capture market share is really important, but how can an entrepreneur implement speedy, iterative development in the absence of years of experience?

MIKE CASSIDY (DIRECTOR OF PRODUCT MANAGEMENT, GOOGLE USA): So, interestingly enough all the four companies I did were in quite different areas. One was in computer telephony linking telephones and databases together, one was an internet search engine, the third one was an instant messenger for online PC video gamers, and forth one was a recommendation site based on recommendations from your friends. So, all four were quite different areas and I believe it's possible to go into different areas and learn about that area quickly and come up with ideas once you get in the area.

NINA CURLEY: So, you didn't have special prior knowledge prior to getting into these regions you did just a quick study?

MIKE CASSIDY: Right, I didn't know anything about any of those four areas before I joined them. I always joking it's frustrating for me because in the beginning when I start my company's nobody will return my phone calls. Nobody knows me at all. And eventually after a year or so the company is doing well and then people start calling me back, but it's always a fresh start in every industry I go into.

NINA CURLEY: But so what special techniques it to imply any special techniques in quickly learning landscape did you develop over time techniques for assessing what you needed to learn when?

MIKE CASSIDY: Yes, so I believe in launching your products about three or three and a half months after you start the company, and then just iterating quickly with sort of improvements to the product overtime. And I believe by launching quickly and then by iterating you can adjusts to the market and find what people really want. For example my third company Xfire, we launched the product three months after we sort of it came up with that idea and then every two weeks for the next year we're coming out with a new version of Xfire. And the first version was pretty simple didn't do a lot of the things the final version did, but we kept listening to the customers and kept into iterating and it's really hard for competitors to stay at a pace you're going, so eventually you keep up. One of the analogies I like to use is any chess player can be a Grandmaster chess player if I can move twice every time the Grandmaster can only move once.

NINA CURLEY: Ah, so you didn't have, you didn't come up with the innovation right at the beginning you really just, it was the pace of the iteration the developed these innovations, and how can you sustain that pace? Does it have to do with scaling your business? Does it have to do with the quality of the people you hire? What is the crucial element for having, for sustaining this level of iteration?

MIKE CASSIDY: That's a great question. I often amass to give advice to people who have a company and that's struggling a little bit and they'll say Mike I've got this problem. The team is is not working as hard, its kind of a little bit slower getting stuff out. How can I get them to work faster? And I always say you're asking the wrong question at the wrong time. What you got to get is people who are similar minded at the very beginning. When I do my interviews with people I ask them questions in the interview that I try to get across this sort of intensity of pace. Like I ask them how do you cook dinner? And some of them will say, oh well, I don't know I put something in the oven and I wait an hour later it's ready. Other people will say oh I time everything. I want to eat at exactly at six o'clock, so I have a schedule three minutes before 6 I put the broccoli in, and at ten minutes before 6 I have the water start to boil so it's ready when I put the broccoli in, and at twenty two minutes before I start the salad, and that's the kind people I want in my start-up company.

NINA CURLEY: So you're really selecting for precision, basically. You want to psychologically pre-select your team this is the most important thing to psychologically pre-select your team for um..

MIKE CASSIDY: Absolutely you're selecting for a competitive spirit. You want people who want to win. You're selecting for ownership. Everyone who joins my company's takes a pay cut, but they get equity in the company and then in so far it's paid of for everyone. We made 22 millionaires at my second company. We made seven millionaires my third company. So yes, your select for people who take ownership for the product. I don't like people who say well that's Charlie's responsibility. I want everyone to say I see something wrong, I'll fix it.

NINA CURLEY: Okay, and how much of that comes just intact in the people that you select and how much do you incentivize ownership?

MIKE CASSIDY: So, I'm kind of a cynic about some of these things. I think the best predictor of future behavior is your is past behavior, so I don't believe I can go get people and get them to be, have, feel more ownership and more passion. You got to find people who have that in them to start with and I find them everywhere. At one the most important guys on a second company, I had never worked with before, but I played ultimate frisbee with, this game where you throw the frisbee back and forth and you run up and down the field. He's an amazing ultimate frisbee player he would dive, lay his body out full across the field and I said I bet he's good to work with and he was awesome. He was fantastic.

NINA CURLEY: Really? Amazing and what kind of numbers are we talking? Can you throw out any numbers in terms that you know when you do get this iteration going at this pace and you have your team, I mean what levels like?

MIKE CASSIDY: How successful our company is?

NINA CURLEY: Yeah, Yeah, how does that translate into monetary value?

MIKE CASSIDY: So I've, I've have been very lucky. The first company we only raised, I put in five hundred dollars and each of my two friends put in five hundred dollars. We had fifteen hundred dollars. We didn't raise any venture capital. We sold that one for 13 million dollars. The second company we raised 1.4 million dollars in venture capital in the first round, and 500 days later we sold it for 500 million dollars. So, we joked around we were making a million dollars a day. The third company we sold about two years after we launched it for for 110 million dollars to MTV. So, yes with this sort of speed I think you can generate significant amounts of you know market value and also speaking on the usership the second company the search engine we were serving 50 million people within a little over a year who were using our search engine. And the third company Xfire, I think we have over 15 million people using the product now. We had a couple million people by the end of the second year using it.

NINA CURLEY: Amazing. So, what you're saying is it's really not the seed funding that's the issue, it's really the team that you hire that's going to be the critical factor in your success?

MIKE CASSIDY: The people are everything, totally everything. And, everyone says that, but when you when you really live it you know. Yes the amount of money you have to start with I don't think is is actually that critical.

NINA CURLEY: Okay and to switch tax for a moment I just want to pick your mind about the future the web. I don't know if you read recent article in Wired called the "The Web is Dead Long Live the Internet" about how we're shifting as consumers we're shifting from browsing on the open web to paying for highly curated experiences in apps things like the iPad the iPhone. How is Google responding to that or more generally what sort of product innovation would you advise given this trend? Do you believe this trend?

MIKE CASSIDY: Yeah, so one of the reasons I'm totally excited to be at Google is we're huge proponents of the open web. For example, Android is our, is our phone operating system. We're turning on over 200,000 times every day someone's activating an Android phone. In the search world we are doing at any given time between fifty and a hundred experiments for improving the quality of search, so we're always sort of making it faster finding new content. We have over a billion people a week searching on our sites so we're using that information from the searches they do to come with with new things but.. we believe in open web. We believe that over time the open systems are the one that will eventually win. You can look at throughout history you know them the various things. Certainly there are advantages to in the short term sort of closed systems having less inoperable parts where things can go wrong, but over, eventually the open systems always win.

NINA CURLEY: Okay, that makes sense. I just have one more question for you. Given that approach that Google is always solidly going to be in that camp, what are innovations in, in the Middle East in the Arab world that a startup company could develop that Google would be interested in?

MIKE CASSIDY: Interesting question. I'm really excited about the Arab world. As you know I was here a few months ago in Jordan and I met a lot of exciting entrepreneurs. I'm meeting a lot of exciting entrepreneurs today. I think some of the things that are most exciting to me are location-based technology. I think there's a lot of things you can do with location-based technology in the Arab world where sort of through the phone systems or other technologies you can locate people and services and products. Whether it's you know, machine to machine connections or a location inside your phones. You know lots of GPS devices inside phones and there's all sorts of the things you can expose to people different services they might want or different ways of finding your friends. I mean right here this conference there's this technology we can find certain people the conference by using location. So Google, as you know Google Maps is a very popular feature and we're always interested in location-based things and things with maps and I I think there's lots of opportunity there.

NINA CURLEY: So, in terms of building Arabic content into that I mean what would make it unique in the region?

MIKE CASSIDY: I'm sorry? Could you say that again?

NINA CURLEY: What would make it unique in the region and building Arabic content into location-based services? Or how important is Arabic content?

MIKE CASSIDY: You know today at Google more than half of all of our revenue comes outside the US. So the Arab region in with three hundred million you know consumers is a tremendous opportunity for us. Whether it's Arabic content or even innovative ideas and as you know companies like Maktuo or Gira a are coming with really cool technologies that's very interesting to to Western companies and Western internet companies. So, I think there's tons opportunity.

NINA CURLEY: Okay, cool, thank you.

MIKE CASSIDY: Thank you so much for having me.


I would argue that the quality Mr. Cassidy probes for in his cooking interview question is indeed precision, more so than actual iterative approach. I have a feeling he would enjoy the turkey parable.

As an aside, if you would like a perfect example of exactly what NOT to do in an in-depth interview (or even a web video interview), start the video at about :50 and note the long, leading, multifaceted questions asked by the interviewer. This is a common symptom of "overinterviewing" when seen in research.

Five word summary - Scientific method for purposeful iteration