Thursday, July 24, 2014

Designing for Innovation Adoption

During my time at Quicken Loans, I've worked on a project called Bullet Time, which lets team members work on whatever they want every Monday after 1pm as long as they follow 3 simple rules:

1. Do something you normally wouldn't work on.
2. Your time must somehow help the company.
3. Don't break production.

Bullet Time is essentially innovation anarchy, where a bunch of creative minds will try and build something new. Most ideas don't get very far along due to lack of effort, feasibility, or missing skillsets, but every once in a while, something really powerful is created. I've noticed 3 basic tactics that help great ideas see the light of day, and want to talk about why they're important.

Tactic #1: Sell the Benefit, not the Innovation

Great ideas will often fail to be accepted, because the benefit is not explained properly. I can write up some brilliant programming fix that decreases computational time by a few milliseconds, or improve my prediction model's ROC curve from 0.54 to 0.83, but my CEO has no idea what this means or why he should care. If I take the same innovations and explain that the extra time my programming fix saved means we can do an extra $2million dollars worth of work each day, or that my prediction model can now pretty accurately predict when someone will buy a specific product over another, then he understands. Then he gets excited. Now my new idea is adopted.

Let's look at the digital music player as an example. When do you think the first digital music player was created? 2000s? 1990s? 1980s? The first working digital music player was actually created in 1979. It was called the IXI and looked like the picture below on the left:
Glorious, huh? The one on the right is circa 1999 and a more evolved version with a whopping 2GB of storage. Anyway, my point is that this technology was around for a while, but wasn't widely adopted by the masses because few people understood how they could use it. Enter Apple in 2000 with this brilliant marketing campaign:
Now people understand what the technology can do and why they want it. Apple could have gone the route of their competitors talking about storage space and hardware specs, but people don't usually respond to that. What is more appealing sounding? 1 GB of MP3 audio files (what's a gigabyte? what's an MP3?) or 1,000 songs in your pocket? It's the same thing, but words matter. Apple showed the benefit of the new technology to the consumer, and now everyone and their mother owns an MP3 player.

This graphic sums my point up nicely:

Don't waste time describing your idea's petals, stem, and flower. No one cares. Focus on what your innovation can do, not what it is. The benefit is what gets people excited. The new superpowers your innovation gives is what people will buy into. Here's another example, recognize this guy?
Subway wanted to show the world that they make healthy fast food. They could have shown that their sandwiches have 12.2 less grams of saturated fat or 36.5% less cholesterol compared to the other McSandwiches out there, but most people don't respond to this kind of message. That's why Subway went with Jared and the imagery above.

"Eat at Subway, lose weight (because it's healthy)."

By selling the benefit (losing weight, new pants), instead of the details (healthier food, less calories), Subway captured the attention and wallets of hungry customers across the USA. You can do this to! Focus on the benefit of your new idea. Show how it helps, instead of why it helps. Help people envision a better version of themselves because of your idea, and it will catch on.

Tactic #2: Roadmap for Implementation

Alrighty, so you have your brilliant idea, and you've shown people how it will make their lives better. But how do you actually make your idea happen? What steps need to be taken for it to become mainstream? Let's look at a few innovations that didn't think this through to show why this is important.
The electric car is a pretty good idea. You're not reliant on expensive gasoline, and a lot of the electric cars out there are pretty energy efficient, which make your green-hippie-friends happy. Pretty straightforward to see the benefit, so why isn't everyone driving these? Well, what do you need for an electric car to work? Electricity! And electricity from a very specific type of charging station. Here's a map of charging stations (Courtesy of around Michigan as of last week:

Notice anything? There aren't very many! If I'm running low on juice in the northern part of the state, I have maybe 7 options to choose from, all a couple hundred miles away. And this is showing stations that are around today! Imagine the early adopter nightmare driving an electric car when ZERO stations existed.
This is one of the reasons we see more hyrbrids than pure electric cars today. With a hyrbrid, if you run out of electricity, you can still use regular gas. For electric cars to be used, there needs to be more charging stations. For more charging stations to exist, more people need to use electric cars. This chicken and egg kind of problem is pretty common and often needs to be addressed for a new idea to be adopted.

Let's look at a different type of example, math. Americans have researched better ways to teach math since the 1800s. We have invented plenty of new techniques that improve learning and enjoyment of the material. But guess what? We don't use any of it. This NY Times article does a great job explaining the phenomenon.
Here are a few examples of why this happens. Some brilliant researcher will do a rigorous academic study around a new learning technique. They will publish their results in a journal that maybe 12 people read cover to cover, say "we should do this", and then move on to the next project. There is never a plan to implement the new solution. Never a roadmap drawing out how to make the change. Similar problems happen with legal red tape. School boards will be too busy forming committees to create petitions to write a report to recommend a decision to the board regarding whether the change should be considered. Without a clear course of action for how it can be implemented, a good, proven idea will waste away into obscurity.

Nintendo recently ran into, and solved, a similar problem with their new console the WiiU, and their upcoming fighting game Super Smash Brothers. Versions of this game have been extremely popular since the first release for the Nintendo 64, and a highly competitive pro gamer scene formed around playing it. One of my college roommates paid for his University of Michigan tuition playing high stakes Super Smash Brothers. This is serious business.
But Nintendo had a big problem. Their new controller was really awkward to use with the game. There are a lot of technical moves and button combinations to rapidly mash to make high level play happen. The new WiiU controller wasn't cutting it in the competitive scene, and made the play experience poor at best.
Nintendo did a good job listening to player feedback, recognized what was needed, and made a simple change to the hardware. Now players can use their old Gamecube controllers (best controller to play a Nintendo fighting game on) with the brand new console. Everyone is happy.
Whenever you have your next brilliant idea, think through all the steps that need to happen for everyone in the universe to use it. Don't turn your brainchild into an electric car with nowhere to recharge. Don't waste effort making a better system, if you can't also design a way to turn the system on. You can be like Nintendo and add a fix in after the fact, but it's always easier to design for everything up front when you have less constraints. Think about what your idea needs to grow and design a way for it to happen.

Tactic #3: Timing is Everything

A great idea, at the wrong time, isn't great. Some brilliant ideas will often be shelved due to technology constraints, world events, or some other random roadblock. Let's look at a few examples. Can anyone guess what this is?
Baby Dalek? Rejected Star Wars droid? Nope, this is Polaroid's first attempt at a digital camera using the most advanced technology available at the time. It could store a handful of pictures, and also had the added benefit of dramatically increasing the arm strength of every user (wasn't very light). Great idea, but needed to wait for the technology to improve to a place where the camera was actually useable. Check out their progress:
Looks a lot better, right? But now technology has advanced too far. Most people don't care about printing photos. I can store all my photographs in the cloud, on the internet, in my phone, and display them on my computer, digital picture frame, or tv screen with the push of a button. Polaroid missed their technology timing sweet spot and is now struggling to stay relevant after a 2001 bankruptcy.

Another great example of using proper timing is the social media giant Twitter. Jack Dorsey tried (and failed) a few times to launch Twitter before hitting success. Timing was the reason. Here's what the mobile phone landscape looked like during his first attempt:
Jack's first version of Twitter was designed to send and receive emails from lists using the elite of mobile phone technology, the Blackberry.
But he ran into a problem. He was the only one in his social circle of friends and families with a blackberry, so no one else could use his creation. He shelved the idea and bided his time. After mobile technology advanced into the iPhone and Android era, he had better luck since more people owned a Twitter-compatible phone. Twitter then grew into the popular social media site/app that it is today.

I ran into this same timing problem with one of my ideas. I designed a system that would help Quicken Loans clients submit their paperwork faster and more efficiently than before. I presented the idea, gained support and acceptance from leadership, and was all set to build it. But then we had one of our largest production months ever and our team members were overloaded with work. My system designed to feed them more work faster didn't seem like such a great idea. So like Twitter (hopefully anyway), I shelved it and am waiting until we have more breathing room to build the idea out.

Putting it all together

So the next time you come up with a million dollar idea, spend time thinking about how you will get people to use it. These three tactics are paramount for good ideas to become adopted, and will make sure that you're not spinning your wheels working on a brilliant idea that will never see the light of day.

1. Sell the Benefit, Not the Innovation - Focus on the quality of life change for the user.
2. Roadmap for Implementation - Have a battle plan for people to implement and use your idea.
3. Timing is Everything - Try to pick a time to implement when everything is working with you, instead of against you.

Design your idea with these three in mind and it will become reality. Happy building!

Wednesday, July 16, 2014

League of Legends: Spirit of the Ancient Golem and Effective Health

League of Legends is one of my favorite games. Never heard of it? Watch this or read this or sign up and play for free. All caught up? Cool. The game is ridiculously hard, and there's a lot of math behind decisions. This is the main reason I enjoy playing. I prefer tank champions, which means my job is to be a brick wall, soak up damage, keep my teammates safe, and cause utter mayhem for the other team.
Riot games recently released a new item that tank players can use called the spirit of the ancient golem. I'm going to walk through some of the math behind why this item is awesome, and how to use it effectively in the game. Feel free to stop reading if that's not interesting, and I have a TL DR; summary at the very end. Otherwise onward!

Effective Health And Why It Matters

So lets say you're playing the game and about to fight an enemy attack damage carry (ADC). You need to stay alive for as long as possible. When your health is gone, you die. You can choose to be one of the following two tanks:

Tank 1: 5000 health, 10 armor
Tank 2: 3000 health, 100 armor

So which one do you pick?

To decide, you'll have to look at how League calculates damage reduction from armor, and then think about how much extra health each point of armor gives you. This is called effective health, and can be calculated like so:

Effective Health = Health * (1+ (Armor/100))

If we run the numbers for tank 1 and tank 2, they look like this:

Tank 1: 5000 health * (1+ (10/100)  ) = 5500 effective health
Tank 2: 3000 health * (1+ (100/100)) = 6000 effective health

So Tank 2 would have 500 extra effective health thanks to their armor reducing the ADC's damage, and survive a little bit longer. Here's a look at different combinations of health and armor so you can see the pattern:
What's interesting about this effective health stuff, is that armor becomes more valuable (creates more effective health), the more health your character has. Notice as you go from left to right in the chart above how the armor gets stretched into MORE effective health. If you run the numbers and simplify a little, you can figure out how much extra health each point of armor will give you:

Extra effective health = (Health/10) * (Armor/10)

Still with me? The main takeaway is that you need to balance stacking health and resistances like armor on your tank to be the best tank you can be, and that there is an optimal combination of the two at any given time. You can figure out this optimal combo by looking at effective health.

Spirit of the Ancient Golem = Tank Steroids

So what's so special about this new tank item then? Take a look at what it does again. Specifically this:

Unique: 25% Bonus Health

What the heck does that mean? Every champion in League has base health. This is starter health and a little bit extra with each new level. You can also get Bonus health from your runes, masteries, and items that you buy in game. There are also some weird abilities that champions like Cho'gath, Nunu, and Sion have that give extra health when used. The Spirit of the Ancient Golem item, takes that extra bonus health, and multiplies it by 1.25 to give you extra. Take an item like Warmog's armor as an example:

If you didn't have any weird runes/masteries, buying Warmog's Armor would give you 1000 health + 250 extra health from the Spirit of the Ancient Golem passive. That's kind of a big deal, but you can optimize it further if you build a little smarter using runes and masteries. Specifically these:

If you max out your % health runes that means:
3 quintessences * 1.5% increased health + 9 seals * 0.5% Health = 9% Increased Health

Riot takes these into account separately, so you're multiplying your total health by 1.03x and then 1.09x respectively to get extra health. This counts as BONUS health, and also gets the 1.25x health multiplier from the Spirit of the Ancient Golem. If you run the numbers, this combo pretty much multiplies any extra health you buy by ~1.4x. This becomes huge since it changes around the optimal way to build a tank from an effective health per gold spent standpoint.

How Can We Use This?

Let's say we're running the above runes masteries setup, and have bought the Spirit of the Ancient Golem Item, but nothing else. You're level 6 and playing Sejuani, so you have ~1500 hp. You also have ~70 armor. You have 1000g which is enough to buy one of the following two items:

Here's your new effective health equation:

Buy Giant's Belt: (1500 health +(380 health*1.4))*(1+(70 armor / 100))  = 3454.4 Effective Health
Buy Chain Vest: (1500 health)*(1+((70 armor + 40 armor)/100))             = 3150.0 Effective Health

Giant's Belt wins out. Lets say you add in an extra cloth armor which costs 300g and gives 15 armor. That would be about the same amount of gold and looks like this:

Buy Chain Vest + Cloth Armor:
(1500 health)*(1+((70 armor + 40 armor + 15 armor)/100))                       = 3375.0 Effective Health

Stacking health still wins out in this situation (and most situations). This new item is strong because it changes the cost of health. Here's how much each stat normally is worth:

1 hp = 2.67 gold (ruby crystal cost / health given)
1 armor = 20 gold (cloth armor cost / armor given)

But now with this new build 1 hp = ~1.9 gold. That means every 20 gold I'm spending to get 1 armor, could be getting me ~10.5 health, when before that same amount of gold would be getting me ~7.5 health. This changes the break even point. Let's look and see the relationship between armor, health, and effective health:

The first chart on the left is the same as before and shows you a champion's total effective health at different armor/hp combinations. The second chart on the right shows how much gold each extra effective health point costs. Notice how the cost of effective health per armor bought changes based on the amount of health you have, and gets cheaper (more effective health per gold spent) as you gain more health. So how much health and armor should you buy in different situations? And how does the Spirit of the Ancient Golem change this?

Let's walk through the math to optimize effective health per gold spent. Here's what we know:

1 armor = 20 gold (cloth armor cost / armor given)
1 hp = 2.67 gold (ruby crystal cost / health given)
1 hp w/ Spirit of the Ancient Golem bonus = 1.9 gold (ruby crystal cost / health given with bonus)

We can rewrite our effective health equation from this:

Effective Health = Health * (1+ (Armor/100))

To this where we assume that we have H gold to spend on health and A gold to spend on armor.

Effective Health = 

(Current Health + H/Health Cost) * (1+ ((Current Armor+A/Armor Cost)/100))

This also assumes that A + H = Total Gold available.

Let's look at a level 6 Sejuani with 1500 health, 83.5 armor, Spirit of the Ancient Golem bought, and 3000 gold to spend as an example:

What we're looking at is the change in optimal gold spending with and without the Ancient Golem bonus to maximize effective health. With the bonus, you want to use ~600g extra buying health than you would without the bonus. That said, since your health and armor change as you buy more items and gain more levels, the optimal spending can also change. 

Let's look at the same Sejuani, but at level 11. 
Your items are Spirit of the Ancient Golem, Sunfire Cape, and Randuin's Omen.
Your health is at ~3375, and your armor is at ~230.
You have 3000 gold to spend again.

Here's what it looks like:

Notice that without the ancient golem bonus you would want to focus on armor (maybe a thornmail?). But now with the bonus, it's better to grab health and armor. Notice that the sweet spot numbers are larger with the golem bonus for the same amount of gold as well. Generally speaking, you want to have a ratio of health to resistances of ~14 (At least for tank mastery Sejuani). That means take your health, divide by armor, and get it as close to 14 as possible.

That means that tanks will perform better against mixed damage team compositions. Now if my bot lane feeds the enemy caitlyn, I have more effective health against AP damage if I choose to grab some extra armor. Before, the 0-5 enemy midlaner would still be able to chunk me down significantly more. Riot did a great job here, because now this means tanks should be grabbing more health and less resistances to get to that optimal effective health.

TLDR; Summary:

Running this setup essentially multiplies every additional point of health you add to your champ from items by 1.4x. This also decreases the cost of health from ~2.67 gold per 1 point in health to ~1.9 gold per 1 point in health! This allows you to get more effective health per gold spent building health than resistances compared to without the Ancient Golem passive.

Health Regeneration and Added Utility

Another thing to think about when stacking health is bonus health regeneration you can get from masteries and items. Check out this combo as an example:

Let's say you have 4,000 health with Warmog's Armor and Perseverance maxed. You are passively regenerating 1-2% of your maximum health every 5 seconds (that's at most 80 health per 5, or 16 health every second). Not to mention that you'll have your natural regeneration and regeneration from any items you bought as well. Imagine that on a Dr. Mundo with Spirit Visage passive, Mundo Passive, Second Wind Mastery, and Mundo Ultimate going. Late game, you're looking at a champ with 6k+ health regenerating more health than the enemy team can dish out. This means that you can facetank towers indefinitely during sieges, or safely tank baron with 1/3 of your health. Just something to think about since this mastery and passive become stronger the more health you have.

Now what about champs that benefit from stacking health? Let's use Sejuani as an example with her Flail of the Northern Winds ability.
The basics of this ability is that you're doing some flat amount of damage + a % of your ability power + a % of your bonus health in AOE damage every second around your champion. Using this build, a level 18 Sejuani running standard tank items is doing 240 base + ~400 from bonus health over 4 seconds to nearby enemies. Add in sunfire cape and you're looking at ~812 damage over 4 seconds to anyone standing near you (before resistances and stuff like that). Without the bonus health from Spirit of the Ancient Golem, you're only doing ~700 damage per second. That's ~110+ extra damage and a pretty big difference considering that you're not necessarily stacking health for offensive purposes. This means anyone going after your frail and fragile carries will have a ginormous wall of hurt to go through first.


Spirit of the Ancient Golem is a huge buff to tank junglers, and a lot of players don't know it yet.
  • The item decreases the cost of health from ~2.67g to ~1.9g, which changes when it's most effective to buy health versus resistances from an effective health per gold spent standpoint. More health over resistances early is generally better.
  • The item adds a lot of utility for additional health regeneration from tank items/masteries making champions like Dr. Mundo very scary during the lategame.
  • The item offers a considerable boost to abilities that scale with health like Sejuani's Flail of the Northern Winds.
I really like this change since it allows tanks to focus more on health than individual resistances, which allows you to be stronger against a mixed (AP and AD damage) enemy team comp!

Thursday, July 10, 2014

Rob Cantor's Brilliant Marketing Play

A guy named Rob recently made a music video to promote his new album "Not A Trampoline”. Instead of just singing his song normally, Rob decided to impersonate 29 celebrities. The impersonations are pretty amazing, and his video grabbed the #1 spot on youtube a few days ago. He currently has ~7 million views, <10 days after he published it.

Check it out, it’s pretty solid for a 3min low-budget music video:
(29 Celebrity Impressions, 1 Original Song – Rob Cantor)

So why did this go viral?

If you’re a Jonah Berger fan, you can go down his list of 6 virality ingredients and see that Rob hit almost every one of them:

1.       Social Currency – People share things that make them look good to others.
I’d argue that this video is pretty cool, and a lot of my friends were pretty happy that I shared it last week. A few million others agree with me.

2.       Triggers – Top of mind, Tip of tongue.
By using 29 well known celebrities, Rob included both audio and visual triggers that make people remember his music. People see/hear the celebrity, and then remember Rob’s video + his music.

Thanks to a well-executed Gilbert Gottfried impression, I am now going to be humming this song every time I hear an AFLAC duck commercial. Same idea for the other 28 celebrities he uses. Rob has effectively weaponized the fame of 29 established individuals to promote his own brand, and did so in a tasteful way.

3.       Emotion – When we care, we share.
Emotional content works best with emotions that get your blood pumping and neurons firing. Stories invoking high-arousal emotions like awe, excitement, anger, anxiety, and humor get shared more than stories that promote low-arousal emotions like sadness or contentment.

Rob picked two of the best emotions for getting people to share things. Awe, and humor.
Awe – He nailed every single impression, and did this in a single take. Oh and the song’s good too. Seriously impressive.
Humor – The video is hilarious, especially since some goofy white guy is perfectly singing like Christina Aguilera and Flipper.

People share more when they are in a high-arousal emotional state. This video executes perfectly here.

4.       Public – Built to show, built to grow.
Rob released the video on youtube, which is a platform built for viewing and sharing content. He also had a pre-determined plan to announce his creation to the world via high-profile social network personalities and leveraging connections from his old band Tally Hall. A lot of people were aware of his video minutes after release and had the tools to easily pass it along.

5.       Stories – Information travels under the guise of idle chatter.
This video falls a little short here. I mean, it’s kind of interesting to talk about a guy nailing 29 impressions while singing a song, but there isn’t really a dramatic plot there.

Or is there?

This video is actually 100% fake. Rob released a video last night once the initial round of sharing died down, showing his millions of viewers that they had, in fact, been punk’d by a 20 person team and a very well-produced video. 

Now everyone that watched the video originally, is re-watching to see if they can spot any slip ups or clues to the fakeness. A flood of comments on the original video appeared minutes after the hoax was revealed saying “I knew all along it was fake”. Now there’s a story to go along with the video, and everyone has something juicy to talk about. Oh, and the entire time Rob and his music are showcased. Very well executed.

Watch his "Making Of" video to see the reveal:

(The Making of "29 Celebrity Impressions, 1 Original Song")

The genius of this can be best summed up by a random youtube comment, of all things:

His story is now being picked up by sites like Gawker and the Today Show. The reveal has now poured gasoline on a smoldering marketing flame and the resulting inferno is still going strong. Check out his Google Trends:
(Can you spot where he launched his video?)

6.       Practical Value - News you can use.
So the video(s) accomplish two goals here.
First, I can now pass on some quality music from a relatively newer artist to my hipster friends.
Second, I can now use this brilliant strategic play as an example of viral marketing done right, and share with all my business pals. This also makes Rob look like a marketing genius (he might actually be one), which is pretty useful hype to have on your resume when it comes to getting shows, appearances, and attention in the music scene.

So there you have it. How to create and leverage a #1 Youtube video to achieve viral hype. This is still blowing up, so it will be fun to track and see how far he can stretch his 15 minutes.

At the very least, you have a quality soundtrack to listen to today:

EDIT 7/15/2014: Here's a quick before/after look at Rob's social media sites.

Youtube: 700 subscribers to 37,499 (+36,799)
Facebook: 1,250 likes to 4,775 (+3,525)
Twitter: 1,700 followers to 3,365 (+1,665)

Not too shabby!