Uncategorized

Redesiging the Seafood Supply Chain

Warnock, Eleanor, “Startups UPend Japan’s Seafood Business,” WSJ, 5/16/2015

Questions

[1] What is the current fresh-fish supply chain in Japan?

[2] Where are there opportunities to redesign the service supply chain?

This is a great case study in service and supply-chain innovation.  In this case, technology enables the disruption.  More generally, however, it is a study in thinking creatively about the problem and about how to streamline the process.

Convergence of Design for Discovery (SEO) and Design for Distribution

Seetharaman, Deepa, “Facebook Pushes Speedier News Publishing,” WSJ, 5/14/2015.

Questions.

[1] Why is optimizing your design for delivery to users so important?

[2] What is the difference (if any) between design for delivery in a Web context versus the mobile context?

[3] Why is “distribution” (aka. download speed) a factor in organic discovery?

When Facebook’s new Instant Articles initiative was first announced, the buzz surrounding the offering focused on the threat to publishers and the dual threat of monopoly pricing and envelopment.  However, this article highlights a second implication of Instant Articles.  Namely, the convergence of Design for Discovery and Design for Distribution.  Recall that the thesis behind Design for Discovery is that Search Engine Optimization (SEO) is fundamentally about design.  In particular, design is all about understanding and optimizing for the user’s need.  Similarly, SEO is all about discerning the user’s need.  In a classic SERP, links are ranked according to the likelihood that the result will serve the user’s need.

However, we know that speed matters and is a dimension of “relevance.”  This article notes that, “[t]ech companies long ago drew a connection between slow loading speeds on the Internet and their bottom lines.  In 2006, Amazon.com Inc. said previously a delay of even one-tenth of a second in loading its website nipped about 1% from its sales.  In 2009, Google Inc. said slowing search results by 100 to 400 milliseconds — less than half a second — meant users conducted between 0.2% and 0.6% fewer searches.”  In other words, speed is an implicit element of the user’s need.  If a site is slow, users will bounce before they even know whether the link contained relevant content.  And so Instant Article is about faster load times.  How much faster?  The article suggests that Instant Articles “[w]ill allow users to access articles about 10 times faster than they can now. The average load time?  Just eight seconds, but for smartphone users, that’s a lifetime.”

As evidence of the link between faster load times, SERP and relevance, the article notes that Instant Article will rank articles based upon “[h]ow much ppl “like” comment, or otherwise interact and how much time users spend reading them. If users embrace the faster loading speeds, those articles are likely to find their way into more users’ newsfeeds, industry executives said.”

Interestingly enough, the relationship between download speed and relevance is exacerbated by the transition to mobile, where Facebook dominates.  “Speed matters, and doubly so in an impatient word where our attention is fragmented,’ said Bubba Murarka a partner at venture capital firm DFJ, who previously led Facebook’s product-development team for Android and who is an investor in a startup called Twin Prime, which focuses on speeding up native apps.”  A Pew Research Center Survey said that “nearly two-thirds of American adults own a smartphone, up from just over a third in the spring of 2011.  A separate Pew study found 39 out of 50 new sites have more traffic coming from mobile users than from desktop.”  Because of this trend, “in 2013, Google recommended to website operators that their mobile pages should load in less than one second.  At that time, just 16% of Fortune 100 websites came close to that threshold according to a report form the marketing firm The Search Agency.”  The newest Google search algorithm actually makes this explicit – mandating “mobile” and even creating a free service to rate the mobile-readiness of a website.  (Need to link to the relevant articles from late April or early May on the latest Google search algorithm shift).

h

MVPs: Not just minimal but easy – optimize your UX

Loten, Angus, “Meetup’s CEO Aims to ‘Get People Off the Internet'” WSJ, 5/14/2015

MVP
Simple UX
Metric that matters
Metrics that are easy to observe v. The metric(s) that matter.

“We can measure how many people are actually going to a Meetup.  The thing we cannot measure yet, but we hope to crack, is how many people are finding a job — or finding an investor, or finding a friend, or running a marathon — because a Meetup pushed them to do it.”

The number of people joining Meetups is spiking right now, because we made the process easier, taking it from three steps to one step.  Making something easy is hard.  But making it easier makes more people join.  That’s the story that plays over and over.

In 2005, cut out advertising
“We tried different business models, but eventually (in 2005) decided let’s just have a product that’s good enough that people will pay for it.  Even so, we lost 90% of our activity in one day (after adding fees for event organizers.). It was a scary moment.  We made (the fee) $12 a month to list a Meetup.  That’s roughly what it is today.

We want to devote 100% of our energy to making the platform better and better, and not be distracted by having to serve advertisers. It’s funny that people just presume that online companies should be in the media business.  We’re not a media business.  We’re not in the advertising business. We help people find other people.

Goal:  build. Website (service) to help “get people together”.  Founded after 9/11/2001

Scott Heiferman, co-founder
13 yrs old
2 yrs before Facebook
Charge coordination fee per event and size
Mobile tech challenge

Cash flow positive
25Mm users
20K events per day

What is your mission? Lessons from the military

Schultz, Howard, “The View From the Top: book review of Team of Teams by Stanley McChrystal,” WSJ, 5/13/2015.

Questions.
[1] How is the military shift from WWII command-and-control to early 21st century warfare in the age of Twitter, related to the business organizational shift from hierarchies to markets?
[2] In what other ways are lessons in leadership, management, and organizations from the military, transferable to modern, civilian, market-oriented organizations?

Yes, the reviewer is Howard Schultz of Starbucks fame. Howard Schultz also co-authored a book about leadership and the military, interviewing and recounting anecdotes of heroism and leadership from “non-stars” in the US military throughout history. The book was sold in Starbucks stores and all proceeds went to benefit veterans. This means that as a book reviewer, Schultz is biased to support the military. That said, the book review (I have not read the book itself), is particularly important for designers in at least one respect. Nail the mission. Virtually any product design process begins with the Mission. I use the text from Ulrich and Eppinger, but search online for the Amazon product development process, and they also start with “Mission” (for example). At Amazon, they call it “writing the press release.” Setting the mission is no less important for a product than it is for the military. It is a tool not only for aligning the team, but it is touchstone for critical decision-making.

Schultz writes:
Drawing parallels between war and business is not new, and while some of the task force’s “aha!” moments and the book’s themes are familiar, they remain relevant. That said, one salient point arrives late, though when it shows up the authors nail it: how leaders communicate and control their organization’s overriding mission and values. Today’s leader, the authors write, must transition from an ‘all-knowing puppet master’ to an ’empathetic crafter of culture.’ Defining and maintaining guiding principles — basically why a company exists, other than to create profit — is as important as any growth strategy. Concepts such as ’empathy’ and ‘values’ still don’t make it into enough boardrooms. But when a battle-worn general suggests that ‘nurturing an organization’ is more effective than trying to oversee its every move and insists that his own leadership is more akin to gardening than chess, it’s worth paying attention.

Surveys to learn about customers

Koh, Yoree, “Pinterest Study Suggests Site Helps Users Find Things to Buy,” WSJ, 5/13/2015.

The news story is actually about Pinterest and its potential as an advertising platform. However, this was a cue to me in thinking about pros and cons of surveys as a research tool for identifying customer needs.

The story reports on a a marketing survey that Pinterest ran with Millward Brown between Feb. 20 and March 7. The survey reached out to 2001 active and inactive Pinterest users by email. The response rate was not reported. However, 24% of respondents identified themselves as “inactive users” meaning someone who had not used the service at least once in the past six months. We know that 52% of roughly 1500 *active* Pinterest users “agreed” that the site helps them find items they want to buy.

Nearly 80% of active users responded that Pinterest’s food and recipe content helped them decide what to buy. 60% of respondents said that Pinterest content helped them pick home decor. The persona (or market segment) “Millenial Mom” led the way in all product categories but one (health and fitness purchase decisions). The other categories were: fashion, home decor, food and recipes, health and fitness and hari and beauty. The market results are meant to support the growing advertising business at Pinterest, where advertisers can buy “promoted pins” that look ike user-generated submissions targeted at particular user groups (perhaps not so different than the trend in publishing towards sponsored content).

In the design context, what is interesting to me is:
When is a survey tool a useful vehicle for discovering opportunities and/or customer needs? This survey highlights strengths and weaknesses of surveys for needs “discovery” versus needs “analysis” (meaning prioritization).

As with conjoint analysis, the problem with anything but a totally open-ended, free-response survey, is the question: what do you ask? If one already knows the relevant categories to ask (e.g. fashion v. home decor), it is easy to fashion a survey asking for responses. However, why did Pinterest choose those two categories to survey? Is it any surprise that women (millenial moms) ranked those categories higher for purchase decisions than men? How many men are making fashion or home decor decisions at all, let alone going to Pinterest to help with the decisions?

Second, although the response rate was not mentioned, who _did_ respond? We know that 24% of respondents were not (current) active users. We also know (colloquially) that Pinterest derives the majority of its viewership from women. So the response population is biased first in who is a user and who responded. Again, is it any suprise that women led the way in categories about the use of Pinterest?

Third, is the question of how you ask the question – note that the article carefully notes: “About 52% of roughly 1500 active Pinterest users surveyed agreed that the site helps them find items they want to buy …” So you imagine a question: rate from 1 to 10 with 1 is strongly disagree and 10 is strongly agree vs. “agree” or “disagree”. How you ask the questions and how you interpret the results colors the interpretation of the result.

Finally, this was about a two week study. In two weeks, they surveyed 2000 people (how many did they hear from)? A 2% response rate would mean about 40 people. A 20% response rate would be 200 people. Certainly interviewing 200 people by hand in two weeks would be cost and time prohibitive.

Design for discovery (SEO) and Platform Strategy

Alpert, Lukas, “Publishers Warily Embed With Powerful Facebook,” WSJ, 5/12/2015

[1] How do information-and-content based services like Gawker, Vice, and Mashable “design for discovery.”  In terms of the Business Model Canvas, what are the “channels” by which these different services meet their customers.

[2] What are the different type of strategic threats a content-firm faces from a gatekeeper/platform like Google or Facebook?

An entire industry is built-up around SEO and SEM – how to optimize your product/service for discovery.  Instead of trying to game the algorithm, the principle behind “design for discovery” is that, to paraphrase James Carville, “it is the design stupid” … good design is design built around the customer.  As a design heuristic, keep in mind that Google and Facebook are simply trying to design algorithms to better discern what the user is looking for (what are the user’s needs) and trying to find the best solution to that need.  In this context, this is why entrepeneurs can discover (and prioritize) user needs by using tools like Google AdWords, Google Trends, Twitter feeds, etc.  Therefore, it is the intention of the algorithm designers that, the “best” design (one that best identifies with and satisfies a particular user need) will naturally rise in SERP as the gatekeeper algorithms get better.

Note that publishers have long adjusted to this – editors write their stories and craft their headlines around what will capture the user walking by the newstand.  Today, people walk by the “virtual” newstand also called Google search or Facebook news feeds.  As a result, “[n]ewsrooms have already adapted to Google’s power as a driver of Web traffic.  Editors in recent years have been trained to develop headlines that are ‘search engine optimized,’ and have learned other tricks to ensure they aren’t punished by having stories appear lower in search results.

One direct result of Google and Facebook becoming gatekeepers is the industrial organization “hold-up” implications.  “Mike Dyer, chief strategy officer at the Daily Best, said the publisher has learned from its dealings with Google, including how economic arrangements can evolve over time. ‘We saw the terms ultimately change,’ he said.  ‘Who is to say that couldn’t happen again?'”

Perhaps more interesting from a consolidation and vertical integration of gatekeepers becoming content creators.  In terms of platform economics, economists would call this “envelopment.”  In the context of the article, the threat is that Facebook moves from a gatekeeper through its news feeds, into an editorial role:  “Anxiety over the social-media giant’s intentions in news publishing has risen in recent weeks ahead of the expected announcement of an initiative called Instant Articles, through which Facebook will publish content from major outlets like the New York Times, National Geographic and BuzzFeed rather than linking back to their sites.”  How large a leap is it then for Facebook (or Google) to become a content provider itself.  The threat of envelopment – when the gatekeeper and clearinghouse becomes a content competitor was the heart of objections to Google’s purchase of ITA (the air travel information platform).  This is the heart of the European Union’s suit against Google in the area of ecommerce and comparison shopping.

As an example of design, prototype, test, and iterate – the lean cycle, the article  notes that “The Instant Article” program isn’t the first time Facebook has delved into news partnerships  In 2011, it was joined by the Washington Post to create Social Reader, a kind of newspaper for Facebook that aggregated news from a variety of media sources into a single feed.  But some users complained that the program deluged users with spam by sharing any post opened with all a reader’s friends.  Traffic fizzled after Facebook tweaked its algorithm and the effort ended after about a year.”

The article also provides updated facts about the evolution in the gatekeeping industry (Facebook v. Google).  The article cites a Pew Research study that finds   “39 of 50 news sites now have more traffic coming from mobile users than from desktop users.”  Facebook is also extremely successful in the mobile ad space:  Last year, Facebook “accounted for 35% of the moile display advertising market, followed by Google Inc. which had 12% according to eMarketer.  Overall, including search and other forms of mobile advertising, Google had 37% market share to 18% for Facebook.”  A table shows Routing Visits:  Share of traffic referrlas to about 300 news sites – trends from 2012 through March 2015.  Google has held relatively steady at around 22.2% but Facebook has exploded from below 10% in 2012 to 26.9% as of March 2015.

Bitcoin experiment at NADAQ

Hope, Bradley and Casey, Michael J., “Bitcoin Technology Gets Nasdaq Test,”, WSJ 5/11/2015

1. What is “settlement” from the perspective of a market like Nasdaq?

2. What are the current problems with how Nasdaq (and other markets) “settle” a trade?

3. How would bitcoin address this problem?

4. What are the selection criteria for a solution to the problem of “settlement” in trading platforms?

The article describes an experiment at NASDAQ to test the role of bitcoin in shifting the industry from a T+3 settlement structure where the final transfer of funds and securities occurs three days after a trade, to T+0 or real-time settlement.  The article is interesting not because offers a tremendous amount of detail into the problem context and issues involved.  Rather, it is a starting point for a discussion along several different dimensions.

First, as a question of design, what is the actual problem?  That settlement takes too long.  Why is the length of time for a settlement a problem?  What is the objective that markets hope to achieve by reducing the time it takes to settle a trade?

Second, what are the criteria behind a settlement solution?  The article raises several dimensions of security.  The dimensions include “availability” (in the security sense of identification/authentication, access control, and availability).  This means that the platform needs to be reliable (think “up-time” and capacity to handle the necessary volume).  Identification/authentication.  Note that the bitcoin infrastructure is intended to be autonomous and private.  The identification/authentication issue here is not that someone knows the actors in the system but that the actors remain reliably independent, autonomous, and private – key on independent – so that collusion and market manipulation is not possible.  That is an interesting computational theory question.

Third, is an issue related to the “Lean” design process.  Namely, how does one “test” a “minimal” solution like updating a trade platform?  Can a solution for a problem of this complexity even be “tested” in a credible way at a small scale?  What does it mean to define a “minimum viable product” in this context?  As a manager, what is the strategy for rolling out a test and, assuming the test is successful, for rolling out a new solution?

I need to speculate on this more; no time right now.  Thoughts?

Market and Product Testing 2

Jonofsky, Adam and Loften, Anugs, “Repeat Crowd Funding Efforts Pay Off,” WSJ, 4/16/2015.

Questions.

[1] How is a crowd funding site like an innovation tournament or an idea tournament?

[2] In what ways is a crowd funding site a good (or bad) selection mechanism?

[3] Recognizing how people are “irrational” in the Dan Ariely sense of the term, what can a crowd funding platform provider do to improve the quality of the screening signal of a crowd funding campaign?

[4] How are decision biases exacerbated or mitigated by crowd funding platform choices?

[5] Why would a repeat crowd funding entrepreneur receive better results than first timers?

Crowd funding sites are mature and better accepted these days.  Recent research highlighted in the WSJ suggests that entrepreneurs who return to crowdfunding sites for new projects succeed or exceed their goals at a (significantly) higher rate than first time posters?  Is this “successful” first time entrepreneurs are more likely to succeed? or any entrepreneur, independent of their initial success?  Why might this be?  Is this merely a reflection of the adage that practice makes perfect (or of the tournament idea that you have to throw lots of ideas at the wall to get something to stick)?  This adage would hold in the case of second- or third-time entrepreneurs succeeding independent of their initial success (or lack thereof).  Alternatively, one could argue for a herding effect-reputation effect that would suggest that *successful* first timers would naturally do better in part because their initial success is a signal to potential investors and (at least in part) because their initial success is merely an indicator of their inherently higher potential for success as determined by natural selection; that is to say – these people are just “winners.”  A third explanation, saying nothing about natural ability but some combination of luck and practice, reinforces the idea that it is not practice that makes perfect, but that “perfect” practice makes perfect.  Rather, that the practice effect is only helpful in so far as the prior experience was actually useful.  In fact, the cited research suggests the latter – that *successful* returning entrepreneurs do better – significantly so.  One also wonders whether the opposite is true – namely, that a failure the first time around is actually *less* likely to do better than a first-time entrepreneur.  Some entrepreneurs like to talk about why failure is good; and how future success was sown in the ashes of their initial failures.  Is this born out in crowd funding campaigns as well?  One also has to wonder (if possible) whether the effect holds for follow-on projects that are somehow related (e.g. see the idea of horizon one versus horizon two innovations in the context of Terwiesch and Ulrich, Innovation Tournaments).

Independent of the “repeat” nature, one wonders whether a platform can mitigate any of the decision biases, and whether the effect (mitigation or otherwise) is exacerbated in the case of returning (un)successful founders.

Designing and selecting concepts for photos and phone contacts

Geoffrey A. Fowler
“Organize Your Pictures … and Your Phone Contacts,” WSJ, 4/23/2015

Questions.
[1] the problem of “cataloging a lifetime of photos”, is a very broad job-to-be-done.  In what ways does this general job-to-be-done differ for different individual users?

[2] Describe the personas of at least two different user types and how the job of organizing a lifetime of photos is different between the two of them?

[3] What questions might you ask a user to discover what type of user that they are; what variation of the job-to-be-done matches that specific user?

[4] How would you prioritize the different needs within this single job-to-be-done

[5] How do you think the priorities change depending upon the user persona?

The app review articles written by WSJ articles are great mini-case studies in design that students may approach from any number of stages in the design process.

At the beginning of the design process is understanding user needs and the job-to-be-done.  Notably, the job-to-be-done is more than just the end result.  It is also only about “what” and not at all about “how.”. In other words, it is possible to describe the job-to-be-done and it’s context without slipping into “how” to accomplish the task.

Consider questions about cataloging photos.
– what equipment do you own (both camera and photo storage device)
– how many photos you have to sort
– is this a task done from work, home, commute?
– do you do his regularly or in big purges?
– why do you organize photos (how do you view – timeline? Maps? Facial identification?)

When you analyze your results, you can create a persona around the constellation of needs and their priorities.  In the article, they name:  the “apple” user, the “anything but apple”, the professional…

The author identifies different products and explains why particular products are better matched with different personas where a persona represents a particular constellation and prioritization of needs.  The “matching” of personas to products may also be seen in the context of a concept selection matrix (see Ulrich and Eppinger on the Product Development Process).  If every product represents a concept, then “matching” a concept to a persona is essentially “ranking” all product concepts with respect to that particular user persona.

One can also see this as an exercise in opportunity identification.  FIrst, by identifying needs, the exercise essentially defines the innovation space for a specific problem (job-to-be-done).  Second, in the context of a review that matches multiple products to different personas, one could also interpret this as a type of “gap” analysis or the articulation of a “blue ocean”.  By categorizing existing products in the multi-dimensional space defined by the personas and needs that are identified, the reviewer also identifies the gaps – where the existing products fall short in satisfying the needs of existing personas as well as opening the space to defining different user classes.