Monday, June 16, 2014

Twitter Should Be Embarrassed With Their World Cup Performance Thus Far

What has Twitter offered so far for the World Cup?

Pretty much nothing besides a generic non-homepage "homepage":



It is a shame considering Twitter has such great potential for the World Cup, and they should be embarrassed with the effort they have put forth. Here are some of my suggestions:

  • Country Flags Next to Usernames

There needs to be a way for a user to represent their country of choice. One such way of doing this is to put country flags next to usernames so that whenever someone tweets, next to their name will be their country's flag.

Next to this users name should be the Iranian flag so I know what county this individual is associated from (assuming she is rooting for Iran). Don't automatically put a flag next to users obviously, but ask them if they'd like to "represent their team" with the flags next to their names.

While I'm on the subject of what should be next to a users username, the "@NegarMortazavi" should NOT be there. I never understood why Twitter put the @username in a tweet. Use this space for something such as my flag suggestion; it makes things prettier. No need for two names.

  • Photoshop Material


There are bound to be plenty of photoshop material pictures, so there should be an account where one can see all these photoshops. One of the good things about Twitter is that the audience will do the work for you (the photoshops), all you need to do is create an account where one can see any/all photoshops.

  • Flops
Whether one likes it or not, flops are prevalent in soccer and will be for the entirety of the tournament. A strategy should be formed around flopping. Maybe a flop count (form a hashtag so users can Tweet to this hashtag whenever there's a flop?)? How about faces of the players who are flopping when they flop? @FlopFace? Anything is better than nothing.

  • Fans


There is nowhere I can go to see fan pictures. Make a central location where one can filter by country and see fan photos from each country's fans. I want to photos of those who are in Brazil, both in and out of the stadiums, as well as those who are in their home country celebrating. Additionally, funny fan moments is never a bad idea. 

Twitter is also great for detecting subtle moments such as the picture above where this man was only on camera for a second or two but long enough for some people to see. Make a hashtag where users can submit content.

  • "Best of"
    • Celebrations
    • Handshakes
    • Dances
    • Goals
    • Fans
    • Moments
    • etc.


A goal celebration is a fantastic moment where teams not only celebrate the goal just scored, but it offers a chance for the players to represent their country in a unique way. Twitter should encourage a "competition" amongst teams and players to see who are the most creative/funny/whatever when it comes to celebrating.

- - - - - - - - - - - - - - - -

There are just a few. As time goes on and I watch/experience more World Cup, I'll come up with more. 

Twitter undervalues the fact that they can make and promote accounts for ANYTHING. Why promote boring team accounts when you can find much more interesting/relevant content that connects with Twitters younger demographic? I can think of so many different accounts that would be appropriate for this event. One of Twitters goals should be to make accounts that are genuinely interesting enough that users want them on their timelines. I'm not sure how viable it is to make something where I'd be able to share a Tweet/photo to one of my friends on Facebook, but that should definitely be a goal as Facebook users will be much more willing to come over to Twitter if Twitter offers unique content.

Thursday, May 22, 2014

To Mike Hudack Regarding News in The Digital Age

The largest missing ingredient to current news publications is that these brands are not leveraging their audience properly. The way I see it, covering the news in the Digital Age means taking advantage of the fact that not only can the audience engage with the author, but they can also engage with each other. Meaning, when a story is published, rather than it being a means to an end (as it currently is), it should be viewed as the beginning of a conversation. Users can subscribe to any story or any parts of a story they find compelling and even potentially be called upon in certain stories where their experience/knowledge is valued. This publication will be a one stop shop for any news story which provides information above and beyond the standard.

Features:

  • Wikipedia-like layout?
  • Ability to comment on any word/sentence/paragraph
    • Quartz has this right, but misses the mark on design of such a feature. I think what makes most sense is putting an underline under anything that is commented on. The more it gets commented on, the bolder the underline.
  • Ability to comment on any and all pictures/videos
  • Highlight news information when returning to a story
  • Polls
    • On certain stories, only specific audiences can answer certain polls.
  • Cliffs notes & "… For Dummies"
    • Some stories are long/foreign so a summary would be needed.
  • Subscribe to a story (as a whole or individual parts)
    • "Level of Subscription"
      •  3 options: passive, moderate, & active.
        • Passive: only receive message when big events occur
        • Moderate: receive messages when moderate level updates happen
        • Active: receive message when any and all information is updated
  • Titles
    • examples: Teacher, CEO, Product Manager @ Facebook, Mother of Autistic Child, Freshman @ USC, etc.
    • This is important. I believe the audience is the most important part of the equation, and being able to leverage your audience is vital. If there's a story about autism & the author wants input from mothers of autistic children, what better to ask your audience by sending a push notification to those mothers?
  • Real names & pseudonyms
    • example: ebay cyber attack. Ebay silent? Maybe an insider wants to give some insight on the story without revealing who (s)he is
    • Pictures of faces associated with a name mandatory. When I read comments, I need to associate faces with names.
      • Photoshop tool that cuts out background and only shows faces
  • "Jump to…" at top
    • When you jump to x, back arrow to return to top
  • Comments/Suggestions/Concerns/Criticisms Box
    • input from users on content of article
  • Reputation?
    • Reddit style "karma" points?
    • ability to take away or give certain features to users who are 
  • Live discussion
    • maybe there is an event going on (TechCrunch disrupt) and people want to talk about it in real time. 
  • Ending the article with asking the audience a question to facilitate discussion
    • either yes/no or open ended, but a user should be able to choose a "side" which will appear with their comments if they comment 
      • also, the "sides" are split into two categories: on one side is the "yes" audience with their reasoning and on the other side is "no" audience with their reasoning
    • example: your talk about the state of news (which gave me incentive to write this post), begs the question: Is it fair to blame Facebook for all of the various click bait offenses committed by media outlets? (Taken from a paragraph in GigaOM)
      • 2 Polls: One is for the general public and the other is for journalists/media types
  • Conversation layout
    • Basically, the more engagement a comment gets, the higher it is
      • comments to a comment more valuable than a "like"
    • any response author makes bumps to top
    • ability to quote person/persons
  • User submitted content
    • This is a big one. Sometimes users are the story. There needs to be a place where someone knows they can submit content relevant to a story and know they will be heard. The problem with Twitter on this front is that there is no specific place/handle to post your content to that you know will be heard. Being known as the place to "send your photo/video to" can offer an incredible competitive advantage.
As I said, this news publication should be viewed as a "one stop shop" for a news story. Lets take a look at the Donald Sterling case. There are still a lot of questions that remain answered and that no news site has bothered to look into.

The Tape
- The tapes contents
- Was it legal to record? (Ask the audience (Privacy Lawyers)?)
- Written dialogue of the tape
     - Ability to comment on any/all aspects of the dialogue
Reactions
- Public reactions
- Reactions from sports figures 
People
Donald Sterling
Shelly Sterling
V. Stiviano
   - History/Behaviors/Live Updates of all individuals
Legal
- Call on the (sports) lawyers to have a discussion regarding this story. Sometimes it's best to know when to shut up and listen.

Friday, May 9, 2014

Creating a Better NFL Draft Experience

The NFL Draft is gaining popularity, at least on television, but it is still in the early stages of taking advantage of the mobile environment. Here are some suggestions to connect further with the audience:

  • NFL Prospects For Dummies
Many viewers aren't hardcore college football analysts, and thus don't know every thing about every player in the draft. There needs to be some type of "overview" for each prospect where one can get a good idea of a player in a multi-senstence synopsis. Whoever does this needs to be honest about the player (give both the good and the bad) instead of just hyping someone up. Here is one example:

- - - - - - - - - - - - 
Derek Carr, QB, Fresno St.

So this Derek Carr fellow has received quite a bit of love. There have been rumors that he's the #1 or #2 QB on a lot of teams' boards, and that he may very well be a top 10 pick. His arm is no doubt megaelite, and is probably the strongest in the draft. His deep ball accuracy is great and he runs quite well. However, at Fresno St. he ran the spread offense and rarely needed to make more than one read. He almost never took snaps under center and essentially looked at one receiver each play. He threw a ton of screen passes and faced very little pressure. When he was under pressure, he looked fairly SHOOK and his mechanics tended to break down. He was also somewhat erratic in a couple games, namely in the bowl game vs. USC. So yes he does have an elite arm, but as far as his ability to go through his progressions and feel pressure, he is as raw as they come.

Player comparison: Matthew Stafford/Jay Cutler
- - - - - - - - - - - - - - 
(source: http://forumserver.twoplustwo.com/showpost.php?p=43172167&postcount=1)

  • Reactions
One of the most exciting parts of the NFL Draft are the reactions to a draft pick; especially the early ones. Reactions are everywhere: at physical locations and on the internet (included but not limited to social media) . These reactions come from everybody: fans, pundits, and fellow professionals. BUT, there is no place that compiles the reactions to a draft pick. I would love to see a service that compiles a complete list of reactions, in which includes a general polling of fans to see whether they liked a pick or not. Reactions can include tweets, videos, pictures, camerawork, etc.

Say the Vikings made a horrible pick (in the viewers minds), and users were polled on whether they liked a pick or not. (NOTE: When logging onto this theoretical service, one could pick their team of choice.) America is then polled about the pick and you could see not only the reaction from Viking fans, but NFC North fans and all other conference fans as well.

  • Last Longer
Before the draft starts, users are encouraged to submit their entry as to which team will pick what player and when. Separate the pundits from the rest of the field to see how they fare. As each pick goes by, obviously a % of the total pool would get it wrong and be eliminated. The only goal of this is to see if anyone can correctly guess the 1st round of the draft.

  • Camerawork
There are obviously a ton of cameras in whatever venue the draft is held in. It would be phenomenal to be able to choose your own camera and theoretically switch between cameras to see whatever you'd like. I bet people would even pay for this option.

Thursday, April 3, 2014

Snickers Social Media Strategy

This one is easy.


Snickers has their "You're Not You When You're Angry" motto, but they don't convert this slogan to the real world.



The most obvious cast of figures who get (understandably) emotional are sports figures.


More specifically, coaches.

But really, anyone who has an emotional outburst on a televised event has room for Snickers' social media promotion strategy. Here's how it goes:

1. Set up some type of catchy hashtag catchphrase that resonates well with social media users. This way, your eyes and ears will be your social media users which eliminates a major hurdle in finding good "You're Not You When You're Hungry" content. These users will also tweet in real time, which gives you the ability to take full advantage of the opportunity.
2. Sit back and wait until users find good content. The idea is to ultimately be associated with any and all televised outbursts, and people immediately recognize the situation as a "Snickers moment".*
3. Think of some way in which this can be translated into sales. Going off the basketball coach examples, imagine having the crowd being "prepared" with a Snickers in hand, and if a coach has an outburst, the floor gets flooded with Snickers bars.

*Being associated with a "negative moment" is not a bad thing. I'd rather form a connection with say Twitter users, who generally enjoy outburst moments. 

Wednesday, April 2, 2014

Why Doesn't Facebook Allow Polling?

This Status by Senator Ted Cruz (~600k page likes) currently has over 56,000 comments.

Facebook is the perfect platform for polls. Users want to engage with current events but there is no medium, such as a poll, to easily compile opinions. Why is this? Polls not only foster engagement, but also provide quality real-time content that users find interesting (and can lead to reading time if implemented correctly).

My idea for polls:

Using the picture above as an example, Senator Ted Cruz offers a poll for the question: "Are you better off with Obamacare?" with options of:

- Yes
- No

  • Obviously, once someone votes, the percentage and number of votes for each option will show. 
  • Comments should be separated into two categories: those who are better off with Obamacare, and those who aren't better off with Obamacare. This way, viewers can see both sides of an argument. Of course, "Yes"users can respond to "no" comments and vice versa for discussion/argument purposes.
  • Anyone who votes will have their answer included so it is easily known what side they are on. In the above picture for example, all users will have their opinion answer in the right corner of their individual chat box.
-        -       -        -        -

There are a lot of benefits that I won't get into because I think they are pretty obvious. Maybe most importantly, polls offer another form of engagement; one where an individual doesn't feel inclined to "share" something (read: an endorsement) or "like" something (read: doesn't provide context). 


Tuesday, April 1, 2014

Why Can't Anybody Get Conversation Layout Right?

It's pretty amazing to me that no social network or publication can get a (seemingly) simple thing such as conversation layout right. Ignoring this problem degrades the user experience and perhaps more importantly hinders discussion. Lots of publications think of publishing a story as a one way street, where the story is published and that is it. In other words, publishing a story is seen as the end of a story, but it rather should be thought of as a start to a conversation. Those who want to gain a loyal following will consider this and start being serious about facilitating discussion.

I'll be explaining exactly what these platforms are doing incorrectly and what they can do to improve the user experience.

Twitter:



Twitter is by far the worst and hopefully they are working on it because I would argue that their current layout actually hinders discussion, which is pathetic. The picture above more or less exemplifies the entire spectrum of what is wrong with the conversation layout. 

1. Too much spacing between users. 

There is so much blank space; not only between users, but also in each individual tweet. It can/should be condensed.

2. The @ symbol. 

I don't understand the reasoning behind including the OPs (Original Poster's) Twitter handle in commenting tweets. It gives the person less characters for replying. It is also obvious that Jeffrey Carter's 1st tweet is replying to Cuban's tweet so what is the point?

3. Not knowing who is talking to who.

Andy responds to Jeff's 1st tweet but it is hard to follow since:

- Cuban's handle is included. Why? He is talking to Jeff, not Mr. Cuban.
- There is no left-to-right spacing. If I'm responding to Jeff, it would make sense that my tweet isn't included in the "first line", but rather is spaced to the right, under Jeff's tweet, to show I'm replying to him. But that makes too much sense. Facebook does this (correctly). Actually, most others do this as I'll show you, but not Twitter!

4. Quoting/Retweeting

Stephanie I believe "quotes" Mark's tweet, and for some reason Mark's entire tweet is copied again. I presume that it is mainly for her own timeline so readers can see the OP and her response. Even if that's the case, it makes no sense to show a reply FIRST before the OP. The layout needs to be different.

As far as retweets go, there's no reason why someone retweeting should be included in the comments section. It adds to the noise and confuses readers.

-        -         -          -          - 

This all ignores that "1st" replies are always at the top. This is a mistake and should be reevaluated. There are a lot of directions this could go (such as sorting by followers or by favorites) but I'll leave that to the guys who made the current layout. I'd at least like to see maybe different colors for each individual in a conversation. That way you can figure out who's who relatively easy.

For a better example, please try to follow this conversation between multiple users who are all trying to talk to each other: https://twitter.com/pmarca/status/448526165889015808

Instagram:


- Like Twitter, Instagram doesn't do the left-to-right spacing (response to a response) AND responses aren't directly under the response (if that doesn't make sense; crevm's response to cole1world should be directly under it, and a bit to the right to show he is responding directly to him). 

- This may be a bit nitty but spacing between users I think is a little too much.

- The OP and all his responses should be "highlighted" or whatever is appropriate to know that the OPs responses stick out to let viewers know the OP is responding. Maybe make his username a different color? 

- All comments that the OP responds to should automatically go to the top of the conversation.

- Say the OP tagged jusroman in his post. If jusroman responds, that comment will be bumped to the top. This is important, especially for the more popular Instagram accounts, so the person OP is tagging/talking to and their reply is not lost in the noise. It adds a lot of value to the user experience and encourages those being tagged to converse (if they know their response will be the 1st thing seen in conversation) more.

Disquis:



Note: not copied to scale. Two different pictures.

- Disquis does the left-to-right spacing, but they overdo it. Look at the real estate GeorgeWire gets compared the other responders. I would put CashMcCall's response icon directly under the halfway point of FlashCrashNoMore's icon. That way, you know he is responding, but it leaves plenty more room down the road if people keep responding.

- No avatars. 

This may seem trivial, but it is important for discussion to associate faces/avatars with words, especially on topical stories. It adds to the experience. Also, it is much easier than trying to match usernames to responses. 

Now, I am not saying for Disquis to start connecting FB/Twitter accounts for avatar purposes, especially since some/most of these users want to remain anonymous, but there needs to be some type of encouragement to put a picture up.

- Order of comments. 

Disquis allows you to sort by Newest/Oldest/Popular, which is great, but it is still lacking. Insightful comments that are late to the party may never see the day of light. I don't know how to fix this, but there is definitely room for improvement.

BuzzFeed/FiveThirtyEight

 

- Requiring FB login.

Just lol at this. I don't see any real reason why people have to sign in to verified FB accounts to comment on these articles. I guess they think it adds to the "realness" of the discussion. I'd argue that it causes people to not speak their mind & act cordially, but I can see the arguments for not letting "anonymous" trolls into discussions. If you have good content I don't think it would be much of a problem to let anonymous users comment. 

- Left-to-Right spacing.

Again, maybe nitty, but this is too much. Once people keep replying to a replier, after the 4th or 5th, reply, it looks ridiculous. As I said with Disquis, space a repliers icon to start under the midpoint of the users icon above them.

Slate:



I don't know wtf discussion software Slate uses, and I know I've seen it on other publications, but this one is god awful. Anonymous profiles, weird usernames, left-to-right spacing bad, no avatars, people "listening" sounds good in theory but doesn't really do anything, sign in too complicated. 

Quartz:


The thing I like about Quartz and which most other publications don't do is that you can comment on individual paragraphs. So a large discussion is possible before getting halfway into the story if, say, someone disagrees with a certain statement. Still, I'd like to see it go one step further with the ability to comment on a single sentence.

I don't think Quartz does an end-of-story commenting system, or maybe I missed it. But if that is the case, that is weird and dumb.

Facebook:


Facebook at least does it better than most. Simple is best and Facebook understands this. Unfortunately most comments get drowned out by the sheer volume. One thing that I'd like to see on FB & other publications: separating comments by positive/negative (like/dislike), but that would never happen on a platform as big as Facebook.

-        -         -         -          - 

To Summarize:
  • Left-To-Right Spacing: don't overdo it, but don't not do it
  • If someone is tagged in a post, give them top priority in the sorting algorithms
  • Color schemes to separate users?
  • Don't make signing up too hard/annoying
  • Give OP and his/her responses higher importance in algorithms
  • Give me the ability to comment on lone paragraphs/sentences
  • Give users more options than just FB/Twitter login. Some people don't want their profiles attached
  • Don't do what Twitter (currently) does. 

Monday, March 31, 2014

Jos A Bank Digitalized

Suit retailers such as Jos A Bank have a profound upside going into the future. Not only can their product translate well in The Digital Age, but the market will only keep growing when men as a whole start dressing nicer because:

- Shopping for said product will be easier, and
- The value of dressing nicer will be more and more prevalent in our social media world.

Both of these combinations point towards a good future for those who can capitalize on digital by providing personalized experiences and hopefully gaining customers for life.

-         -          -         -        -        -

I want to try on my outfits/suits virtually because I want to see what I look like in them. Such an experience enables me to see how I look in certain outfits instead of the current model where one searches for outfits by seeing how they look on models or outfits as a standalone. This sets up the possibility of virtually trying on millions of combinations of outfits.

The Entire Experience

I really want this product to put an emphasis on lifestyle. This isn't limited to dressing nicely. Correct haircut/hairstyle is important (as well as other face products), as is putting an emphasis on getting and staying in shape. Consider these elements outside of the actual outfit itself and there are opportunities to capitalize beyond the core products.

It starts with the B&M store. Most important, the customer will get a digital scan that is saved and uploaded to the Jos A Bank app. My picture will be used to virtually try on all types of combinations of outfits. An individual will also go in to get (at the very least) necessary measurements which will be saved in that customers digital profile so that when they order product, they won't ever be bothered to input measurements and the like. Going into stores also helps because the professionals that work in the B&M stores can give their advice on what styles may or may not work based on the customers physical appearance. They can even input this advice onto the customers profile. B&M locations will have less emphasis on product and more emphasis on services, such as taking measurements, providing advice/input, setting up customer profiles, etc.

Once the boring stuff is taken care of at the B&M stores, the real fun starts.

Most importantly, I will be able to mix and match a combination of any outfit I want; my digital picture will be center stage for all of Jos A Bank's products. These products include:

- Ties
- Dress Shirts
- Sport Coats
- Pants
- Socks
- Shoes
- etc.

I can virtually look at any style, color, and combination of product of my choosing. I can zoom in to see the minute details of a product, see a 360 degree view of the outfit, and/or have different "poses" that highlight/emphasize different parts of a product (i.e. sitting down and crossing legs show off more of a sock).

As opposed to current shopping for a product in which a model is wearing an outfit, seeing myself in my choice of outfit connects more with me from an emotional standpoint. If I can't see how I look in a certain outfit, how can I truly judge whether or not I like the outfit?

Social

Users will be encouraged to upload their picture once they receive and try on their new outfits. Other users will be able to see and make comments on said outfit, i.e. the outfit is too big. It will be a productive feature of the app that ultimately encourages discussion and leads to long term satisfaction of outfits.

This also allows people who are unsure of what types of outfits they want to choose can look from a library of other users and possibly buy the same outfits (and maybe if someone buys your exact outfit, you are rewarded in some way?)

Other Products

As I said, we are encouraging a lifestyle, not just a product. Hair and facial products are an important consideration when thinking about more than the outfit, and I see no reason why these shouldn't be included.

Promotions/Rewards To Think About

Points

A points system can and should be implemented. Points can be used down the road to purchase smaller or larger purchases. Points are acquired if:

- An item is bought (the larger the purchase, the more points you receive)
- Someone buys your outfit
- "Weighted" points (see below)
- More?

Weight Loss

It is not debatable that an individual who is in better shape looks better in an outfit. Thus, overweight/out of shape individuals will be encouraged to get into better shape. They will have incentive to do this because they will be rewarded in doing so.

Overweight individuals, say a 300 lb man, will get free trade-ins if they lose weight and their outfits are too big (or they are discounted). But maybe not as I can see that being costly. Or, for every percentage of body weight they lose, they will get that amount in discount on all future purchases. The 300 lb man who gets down to 200 lbs will receive a 33% discount on purchases. Alternatively, they will receive "weighted" points. The 300 lb man who gets down to 200 lbs receives 1.33 points for every normal point they would have received if they had not lost weight. All the details can be hammered out later, but encouraging users to lose weight and look better in outfits will build loyal customers and a great sense of community.

If overweight persons are potentially rewarded, there also needs to be some type of rewards system for those who aren't overweight to make it fair.

Color Schemes

There are countless color schemes one can do with a suit. Maybe a discount can be implemented if one buys "matching" parts of an outfit such as a yellow tie with yellow socks.

Other color schemes that can be implemented are school/team colors. Since I am a Chicago Bears fan, and I like the orange/blue color scheme, the app should have a "Chicago Bears" section that gives all possible combinations of Bears colors.

There is a lot of creativity that can be implemented with color schemes, so I am only scratching the surface and trying to give you and idea of a few possibilities.

Discounts

Current Jos A Bank discounts are incredibly sporadic and confusing. Discounts are also a big part of the Jos A Bank brand so this is an important feature of the app. The biggest thing to think about is providing personalized discounts, which will lead to more business.

Buying Competitors Product on Jos A Bank App

Maybe someone is strolling down a sidewalk one day and sees someone wearing a great dress shirt, but finds out it is from a competing brand. Jos A Bank can implement some type of system where if someone encounters this problem, they can still buy a similar product on Jos A Bank's app. Something to think about.

Education

There are a TON of things to think about when choosing an outfit (and even the other products such as hair & facial products), and users need to be educated. Educative tools should be provided; everything from the basics & vocabulary of wearing a suit, to the more advanced and minute details of outfit attire.

Mobile Strategy

Tablets are the optimal platform to use this product. Ones digital rendering will be large enough that they will have no problem seeing how they look when exploring outfits. In addition, a lot more features can fit onto a page, adding to the value proposition of the user experience. This product can also be used on smartphones, but there is a lot less real estate which will have to be considered.

-         -          -          -          -          -

This is all about customer retention. We want Jos A Bank to be a lifestyle brand that extends beyond the actual products and emphasizes bettering oneself. There has never been a better time to have a serious digital strategy that not only puts you miles ahead of the competition, but gains a loyal following. The use of digital allows for personalized promotions, rewards for being loyal, and most importantly, the ability to provide a simple yet considerably rewarding experience that will make users wanting to come back for more.

Tuesday, February 11, 2014

Flappy Bird Analytics & Monetization



Analytics

When it comes to Flappy Bird, most people are wondering why it was deleted from app stores, why it became viral in the first place, or why such a simple game is so frustrating. I am only interested in the game analytics.

Even a simple game such as Flappy Bird can provide so many statistics that many would never think of. It's a good thing I love numbers; but unfortunately I don't have access to any Flappy Bird data. Here are a few things that come to mind in my limited time of playing the game:

General App:

  • Total Downloads
    • What % have the newest update?
    • Current # of users?
    • Peak # of users?
  • Download Chart Since Inception (May 24, 2013 for iOS; January 22, 2014 for Android)
    • Growth
  • iOS/Android Segmentation
    • % on Tablets?
  • How many times was Flappy Birds deleted (% of total downloads)?*
    • What % of those deletions have re-downloaded?
    • Did this vary among iOS/Android users?
  • How many Flappy Bird queries have there been since it got taken down vs. before it was taken down?
  • Ad Data
    • How many times has a display ad been displayed?
    • % & # of display ads clicked on

Game Specific:

  • Score
    • Highest score?
    • % of users that get >10? 50? 100? 200? 500?
    • With sample size of 100 or greater….
      • What % have gotten to 10? 50? 100? 1000?
  • Total # of (gross) plays by all users?
    • Avg # of plays per user?
  • Gross time spent playing the game
    • Avg total time per user
    • Avg time spent on Flappy Bird in one session
      • Longest Flappy Bird Session?
  • # of Taps
    • Gross
    • Per User
    • Per Game
  • Biggest Winner/Loser?
    • Most plays w/ lowest high scores
    • Least plays w/ highest scores
  • Correlations?
    • Are users more likely to die as they get close to their high score? Right after high score?
      • Same question for milestone markers (50/100/200/etc.)
    • Do users perform better during certain times of the day?
    • Did less people delete Flappy Bird after Mr. Nyugen removed from the app store?
    • Do different colored birds perform differently?
    • Do users get better over time?
      • Average score from tries 1-100, 101-200, etc.
      • Average score on try 1 compared to 101/501/etc.
    • Do certain display ads do better than others?
      • Color Scheme?
      • Product?
    • Do larger screens perform better?
      • Tablets vs. Phones
    • Is there high correlation with incoming data and dying (i.e. receiving calls/texts)?
  • Medals 
    • # of medals given
    • Breakdown by Medal
      • gold/silver/bronze
      • how many of each medal users have
  • What % of all plays die before/on Tube 1?
  • Geographic Data
    • Which country/continent had highest/lowest avg score?
    • USA
      • Breakdown by state
        • Avg score
        • Volume
        • Time spent
If only I could access Flappy Birds' data to answer these questions for you…

*Flappy Bird is considered to be one of the most frustrating games out there, which I'm sure has led to tons of app deletions. Would be interesting to see the amount of people that deleted the app out of frustration (only to re-download it, of course).

Monetization

You've probably heard that Dong Nguyen makes an estimated $50,000 per day on display ad revenue, which is probably more or less accurate. If I understand correctly, Mr. Nguyen receives compensation every time a user clicks on a banner display ad. While this has obviously paid off for him, there is without a doubt a way to make more money (without hurting the user experience): in-game purchasing.

When a user inevitably dies, he/she has the ability to start over OR spend 25 cents to continue on. Simple as that. The user could have $5 worth of "Flappy Credits" before the game is started so they can use those credits in the game 25 cents at a time. Additionally, there should be a leaderboard which highlights the highest scores (and how much money, if any, was spent) sorted by categories such as country, state, friends, and/or all users. Users who don't spend money with highest scores will also have their own, separate leaderboard for best "organic" score. This strategy gives users a "crutch" to lean on but at the same time does not affect the actual game experience itself. Since users are so addicted to getting high scores among themselves or friends, this would be a winner. This is in ADDITION to advertising revenue that would be made.

Some quick math (assuming 5M downloads):

Conservative:
$25,000 x 365 =  $9,125,000
$0.10/user x 5M x 365 = $182,500,000
= $191,625,000

Average:
$35,000 x 365 = $12,775,000
$0.25/user x 5M x 365 = $456,250,000
= $469,025,000

Optimistic:
$50,000 x 365 = $18,250,000
$1.00/user x 5M x 365 = $1,825,000,000
= $1,843,250,000

I personally believe that every app is unique and should therefore possess its own unique strategy for monetization. Lots of freemium models use advertising as a crutch (Facebook), or give users a super basic version of an app to entice (read: scam) users to pay for the "normal" version (Spotify). Flappy Bird is a great example of someone who can do in app purchasing correctly. Unlike Candy Crush where after a certain amount of lives you have to wait x time to play again (or pay), this Flappy Bird idea wouldn't take away anything from the game or hold the game hostage to users. It is simply an addition that allows users to enhance their experience, but if they don't want to use it then it doesn't take away from the grace of the game.


Tuesday, February 4, 2014

Making Chipotle (Even) More Efficient

The Problem

Chipotle's throughput isn't fast enough. It's pretty fast, per their recent earnings statement highlighting their "four pillars of great throughput", but it still can get (much) faster in my opinion.

Theoretically, the throughput that Chipotle should be striving for is the time it takes to make an item and no more, which I believe is possible. If a burrito takes a minute to make, a minute should be the goal. There are currently factors adversely affecting this time such as payment, other ordering items (chips/guac/drinks), and customer inputs (specifics such as brown/white rice, black/pinto beans, etc.).

The Solution

I believe ALL of these adverse factors can be smushed into the time it takes to make an order. By the time my burrito is made, I have already paid for my burrito and any other things I want to order. In the article I listed above, CEO Steve Ells says he isn't against introducing a new POS system, which is essential for the solution I am about to offer.

Ells should start by pushing people to place their orders in mobile, especially when they are waiting in line. By the time a customer steps up to the employee, the employee already has their complete order and most importantly doesn't have to ask anything about the order to the customer. This frees up customer time so they can immediately go to the cashier and pay for the meal and get anything else they need to get. By the time they are done paying for everything, they are only waiting on the employee to finish making the order.

Instead of:
-place your order, name all ingredients, wait for order to be done, add chips/drinks/guac, pay

You:
-place order on mobile before getting to the front of line, pay while waiting for order to be complete, get order.

It replaces a few steps and allows throughput production to be limited to the actual order itself and nothing else. A key step of this is the customer not having to name all his/her ingredients in real time by instead putting their order on their phone. This order goes to a tablet in front of employees who will have all the info necessary to complete an order without asking a customer anything.

This also opens up to all types of possibilities on mobile, which is for another day. Most importantly, mobile allows people like me (who go to Chipotle a lot) to not have to place our same order over and over every time we go to a location. In essence, it allows Chipotle to "remember" their customers better.

If there was ever a time to push mobile, it is now.

The Numbers

Old Way: 1 Burrito (rice, beans, chicken, sauce, corn, cheese, lettuce)

Assuming
(1) You pay after your order is complete*
(2) Each individual ingredient ordered by customer takes 3 seconds**
(3) A burrito takes 1 minute to complete with no interruption***

1 min + 3 sec x (7 ingredients) + (Payment = 25 seconds) = 1 min 46 seconds

New Way: 1 Burrito (rice, beans, chicken, sauce, corn, cheese, lettuce)

= 1 min = 56.6 % faster than old way.


* Rough estimate. A customer may have to fumble through wallet/purse, hand card/cash over, wait for payment to be processed, receive card/cash back. I feel 25 seconds is a low estimate.

** Also rough estimate; maybe a little too high. Point being that an employee won't have to bother with asking customer for individual ingredients as they'll have the complete order in front of them, which eliminates (valuable) time.

***No interruption means the complete order is already known and the employee can complete the order as fast as possible.

Friday, January 31, 2014

Zombie Hockey Players


For Halloween last year (2013), nothing special was done by brands. It was the same old movies and same old Halloween themed corny shows that don't stick out. This year, lets try something different.

Imagine a viewer tuning into the nationally televised NHL event and seeing "zombie" hockey players playing their hearts out. The players have individualized makeup on as well as ripped jerseys that all play into the Halloween theme we are shooting for. Not only that, but when play is stopped and cameras pan out to the audiences, the crowd is decked out in zombie face paint and other Halloween themed clothing, giving off a "zombie hockey game" on and off the ice. Both hockey and non-hockey fans alike will tune in to see such an event (assuming good marketing pre game).

The NHL is a perfect candidate to do something halloween themed that:

1) Provides great content
2) Is centered around social (and mobile)
3) Stands out (and "wins") on Halloween

The Players

Assuming there is face paint that would not in any way interfere with a players ability to play hockey (i.e. sweating), this should absolutely be done. This is a chance to create a social product that is fun for both the players and the audience. IMO, there would be nothing cooler than tuning into a live sporting event only to see zombie hockey players flying across the ice and acting like it's just another day on the job. Hilarious.

Obviously players shouldn't be forced to participate if they don't want to, but I don't see any reason why such a fun idea wouldn't be celebrated by all. It is an opportunity for a select few teams to showcase a product that has never been attempted before.

The Crowd

If stadiums were to provide face painting before the game, and individuals were encouraged to show up early and get their face painted, there is potential for an entire crowd of people at a hockey game to be a "zombie crowd". That in itself is just amazing to think about. Just tell people to wear their hockey jerseys to the game, hire face painters, and sit back and watch as all their photos go viral online. Younger people are always looking for cool photos to share with their friends, so it should be popular with them and sellouts should be no problem.

The TV Audience

Lots of people are at home on Halloween night doing nothing besides waiting for the doorbell to ring for trick-or-treaters. Unfortunately, Halloween 2014 rests on a Friday night, so plenty of people will be out, but if you promote this event correctly most of those people out will still try to get a glimpse of the game wherever they go. Either way, tons of households will tune into the game because

1. It's a live sporting event
2. There is nothing else to watch on TV
3. It appeals to all types of people

Marketing

Each social media network has its unique advantages that should be leveraged.

Snapchat: Sending snaps to people that disappears in seconds will give people a "glimpse" of what is to come. This would most effectively be used pre game, when the hockey players are getting the face paint put on. Most of the younger crowd is on Snapchat so it would definitely capture attention/buzz on other social networks with these individuals.

Twitter/Facebook: Obviously strategies on each network should be different; how you do it is up to you. I would at least suggest posting compilation albums of the players/crowd and also starting a #hashtag promotion so you can aggregate all the pictures possible.

Sunday, January 19, 2014

Why McDonald's Should Fund Jamaica's Bobsled Team



The Jamaican bobsled team just qualified but needs $80,000 in funds to even go to the Sochi Games in a couple weeks. What brand better than McDonald's to make this dream come true? If my beautifully photoshopped picture above doesn't already convince you in itself, let me explain why McD's is a perfect match for Jamaica's problem.

The cost is only ~$80,000 which is peanuts from McDonalds perspective. At a very minimum, it provides great publicity for the brand and sheds them in a good light with all who hear the story. They will most certainly be praised from the Jamaican community which provides a good opportunity to connect again with this market? It could provide a perfect segue into expanding into the Jamaican market again, just saying. The publicity alone should be enough to convince you; and imagine if the Jamaican team were to actually do well in the event or even win! I think any brand would be elated with being part of such a story, even a corporate behemoth like McDonalds.

In our ever expanding Social Age, plenty of news outlets have gotten wind of this story and you can guarantee the costs will be covered by someone, but who will step up and take advantage of this enormous opportunity sitting in front of us?

Monday, January 13, 2014

Skittles: Taste The Beastmode

Fact: Marshawn Lynch of the Seattle Seahawks loves Skittles.
Fact: The Seattle Seahawks are currently still in the Super Bowl hunt.
Fact: This is a great marketing opportunity for Skittles to take advantage of.

Skittles has a fantastic opportunity this weekend (and possibly the Superbowl) to set up a promotion. I am disappointed they didn't do anything noteworthy this past weekend when the Seahawks played the Saints. Marshawn Lynch scored a touchdown after which followed a shower of Skittles from the stands:


I mean, how much more obvious should it be? Skittles dropped the ball. They at the very least should have known that "Beast Burgers" were being sold which included a bag of Skittles as a side dish. It especially hurts considering they are one of maybe a few brands that has a native (and real time) marketing opportunity in an NFL game; which is incredibly difficult to come by.

Obviously Skittles should want to run a promotion. Free exposure on one of the highest rated (live) televised events should warrant such an activity. Last years NFC Championship game garnered 42 million viewers, while the Super Bowl had 108.4 million viewers tuned in. I fully expect these two upcoming weekends to have an even larger audience, considering the storyline (old school [Brady/Manning] vs. new school [Wilson/Kaepernick], which is an absolutely phenomenal storyline for the NFL.)

Considering that this weekend will be the only opportunity for a game to be played in Seattle, Skittles needs to act fast.

Here is my proposal:

First off, let me say there are a ton of different ways for Skittles to spin this off, the following is only the first thing that came to my mind. Also, I am not sure how content the NFL will be with such a promotion, but that can hopefully be figured out.

-Get as many Skittles as you can into CenturyLink Field during game time.
-If Marshawn Lynch scores a touchdown, the fans will make it rain Skittles*
-Capture a hopefully amazing moment
-Have social media posts ready

"Real time marketing" is a phrase often thrown out a lot these days, and it is a very real possibility with your brand this Sunday. You will have no better chance at owning a moment in front of 40+ million people at home. I can't think of an "endgame" to this proposal (such as how to get the viewers at home involved and ultimately make your brand money), but I'm sure something is there. If the Seahawks end up winning Sunday's game, then you have another major opportunity at the Super Bowl to highlight the brand in a creative way.

THE CONTEST:

One idea I thought of was a "How many total Skittles will be on the field by the end of the game?" contest, with the assumption that Marshawn Lynch scores at least one touchdown. A crew will pick up all scattered Skittles throughout the event, and at the end of the game they will be counted. This gives Skittles an opportunity to divert attention to the brand AFTER the game as well.

In order to qualify for a guessing number, one will have to purchase one of your products (say a $10 product)  before this Sundays game online to get a code which is used to "unlock" the ability to have a guess at the end amount of Skittles on the field. Winner receives two Super Bowl tickets? that they can either use or sell.

I'm a big math guy, so I think the use of numbers/data/statistics should also be used. Since the contest revolves around guessing a number, I think it's only fair to show consumers other interesting numbers.

According to Ask.com, there are approximately 65 Skittles in each regular sized bag. CenturyLink Field has 67,000 non-club seats. If one bag was sold per customer (at $1 each?), there is a possible 4,355,000 Skittles in the stadium. Assuming people run out after Lynch scores a touchdown, and there is additional time left in the game where Lynch can score again, there should also be vendors walking up and down aisles so people can "refill" at a cheap price.

*                  *                 *                 *                *

There is an absurd amount of creativity that can go into this promotion, but the most important part is that you act fast. If you have any questions, please feel free to contact me. In my opinion, it is an exciting opportunity for the Skittles brand to showcase their creativity, and maybe most importantly get some great exposure and make the brand lots of $$$.

Tyler Yearian
@KizmitSolutions
847 650 5319

*While I am not advocating fans throwing Skittles and potentially causing injury to anyone, it can't really be prevented. This past Saturday's game is evidence. If Skittles wants to do a safer route, then there are definitely alternatives. Additionally, Marshawn wears a visor, so at the very least no Skittles would hit his eyes. Also, if the Seahawks (and Lynch) had advanced notice of this, there can be a protocol on what is to be done if Lynch scores.

Friday, January 10, 2014

LeBron James' Ripped Jersey

Last night, LeBron James ripped his jersey towards the end of the game against the New York Knicks.


What's my point? My point is that it was a perfect opportunity for a brand to take advantage of real time marketing. I'm still trying to figure out exactly what the rip looks like and what brand would be best to take advantage of the situation (looks like a hood to me, which I pointed out in my @IdeaPundit Twitter account), but the point is that no brand recognized the situation.

Another thing to point out is that when something like this occurs, it is not a free-for-all model. I can't see any way in which, say, Pizza Hut can take advantage of this situation. The uniqueness of certain situations just naturally benefit some brands compared to others.

This goes back to the Oreo Superbowl tweet. What made Oreo's tweet so great was that it was a spontaneous action by an brand in a non-scripted situation. (Hint: just like LeBron's ripped jersey!) No brand has come even close to replicating such an execution. The only "real time" marketing tactics that have been used by brands are premeditated tweets, such as Denny's tweet after the recent BCS championship:


or Red Bull's VMA tweet:



In order to duplicate Oreo's Superbowl tweet, brands need to recognize that real time marketing can occur at literally any time, and need to always be on the lookout for opportunities such as the one that occurred last night.

Until next time.