Better Product Managers, and Product Management

There’s a new post over at On Product Management about blenders, and what they tell us about simplicity.

blendtecNow, as it happens, my blender of choice is not a simple blender.

I’m a dedicated - but very non-fussy/pragmatic - gourmet cook, and I love my BlendTec.  (You may recognize the name from the “Will It Blend?” YouTube videos, which are brilliant.).

Simplicity is one way to think about it. Designing the whole product experience is another.

Saeed writes:

The usage scenario goes something like this:

  1. Place the contents to be blended into the blending  container
  2. Blend for 10-15 seconds (maybe 20 seconds in extreme cases)
  3. Pour the contents out of the container

That certainly sounds like the types of usage scenarios I typically read in Product Requirements docs.  But it illustrates the difference between “designing the product” and “designing the whole product experience.”

Read the rest of this entry »

   

If you start recruiting via social networks and social media, talent will expect you to respond more rapidly. Different modes of communication have different expectations of “real-time” response.

If your initial business model assumptions were incorrect, you will need to adapt your execution plan. It may be possible to still succeed if you can provide a lesser version of the product/service but with lower costs, faster time-to-market, higher scalability.

If you collect people’s personal data, you will need to clearly communicate what you do with it and what will happen to it if your company doesn’t survive. If you suddenly close up shop with no warning, you’re going to have a lot of angry, alarmed ex-customers.

“If someone’s going to cannibalize your business, better it be one of your other businesses.” - Getty CEO Jonathan Klein (Getty Images acquired iStockphoto, a crowdsourced low-cost stock photo provider).

If your code fails the “random person test”, then the only people who can point out flaws in your assumptions are other engineers. At a previous company, a lot of engineers complained about the effort involved in converting our entire codebase to Model-View-Controller.  They complained a lot less when they realized I and other product managers could see potential pitfalls when 10% of the code was written instead of 100% done.

If you change your environment, be prepared for your environment to change. “Professional” changes can move you away from the grubby-but-effective roots that got you to launch products and close deals.

If you want something done but there’s an obstacle blocking you, there’s usually a way around it. Funny story about looking beyond the obvious.

   

Here’s why I won’t be using Hunch:

hunch_fail

Anatomy of a bad user experience - step by step:

  1. Read an article about Hunch, remembered that I had a beta account.
  2. Thought of a question that I wanted an answer to.
  3. Went to hunch.com and tried to type it in - you CANNOT ask a new question.  If what you typed doesn’t match a pre-existing topic, there is no call to action to create one.
  4. Logged in - maybe only logged-in people can create a topic?
  5. Clicked on “Create new topic”
  6. Got the above image - basically “no, you can’t do what you want until you jump through these hoops.”

I understand that the system needs to be “seeded” with information.  This is not the way to do it.

Any product where the first user experience actually uses the words “Come back when you’re done” needs an immediate rethinking.

   

It’s hard to argue with the concept of the Minimum Viable Product (MVP) - build the smallest possible representation of your product vision first - then validate it, get feedback, and refine it.

But how can you balance the Minimum Viable Product with maintaining credibility?

Companies who do not employ iterative development processes may not understand that “version 0.1″ is a mere fraction of what “version 1.0″ will be.  If you’re selling into an enterprise company, you may not get multiple meetings where execs can see your vision unfolding.

When the rallying cry of MVP is “if you’re not embarrassed by your first version, you spent too long on it”, how do you balance that with looking serious enough to get meeting #2?

Prepare: Go direct to the end-users for early testing and research

In a B2B2C situation, you can do a lot of quick and dirty research with your customers’ customers.  You don’t have to tell the customers you’re doing this (it’s very easy to solicit for “users of Company X” on Craig’s List or similar venues).  If you want to be stealthier, you could ask for “users for Company X’s competitor”.

When I worked at Yodlee, my team frequently tested the first (”most embarrassing”) rounds of prototypes directly on end consumers and iterated based on that feedback.

This also allowed me to go into meetings and say “I’ve talked to your customers and this is what they want” - very powerful.  Then we had the opening to say, “Now tell us what YOU want” without sounding clueless.

If you’re working with enterprise software that will be used internally, it’s more difficult, but still possible, to contact employees directly.  It’s easy to find names of people using LinkedIn, and if you compensate them for their time they really won’t care that this is in preparation for a presentation to their bosses, as long as you maintain their anonymity.

Prepare: Invest some time in elegant but minimal design

You don’t need to have your product fully thought out - it’s not even possible - but what you give your customers to react to, can still look professional and polished.   Someone with great UX design and CSS skills can make a great-looking — but minimalist — usable prototype - not as quickly as an “ugly” one but still 5-10x faster than it would take to code it.

“Looks good” is a surprisingly strong indicator of credibility.  A pleasant neutral color scheme, good use of whitespace, and of course, lack of typos, gives customers faith that whatever else you build will also be high-quality.  (I have tried the opposite - presenting essentially the same demo but without this minimal good design, and the difference in reception is enormous.)

Alternatively, you can use a tool like Balsamiq, which gives you “hand-drawn” looking stencils so that you can make UI elements look consistent and legible - but still clearly “works in progress”.   (Another option is to create stencils for Visio or Omnigraffle - a good option if you have a lot of visual elements specific to your industry that you’ll need to reuse over and over.  I had one of my designers create a whole set of stencils, which allowed product managers and other “non-visual” types to sketch out prototypes that looked good.)

Present: Set expectations.

Emphasize: this is early and their feedback can help direct the product. Use flattery (”we’d especially benefit from your feedback because of your expertise in X”).  Follow up with a thank you and synthesis of their feedback (”we understand that your main needs are X and that you’d be looking for something more Y”).  Ask for the followup (”can we contact you next month/quarter to show you the next revision?”).

   

By now we’ve all seen the numbers showing that 90% of Twitter’s content is contributed by 10% of their users.

  • First thought: why is this a surprise?
  • Second thought: why is no one talking about how to change these numbers?

What we should be asking is:

How can we break the 90-10 “rule”?

It can be done.  Un-scientific research of my Facebook friends shows that over 30% of them update their status at least weekly.  (My Facebook friends include a large number of early mainstream/late mainstream technology adopters, so I suspect that’s a fairly representative number.)  Lithium, who powers enterprise online communities, says that their audience participation rate is also around 30%.

What is it about Facebook and well-run online communities that converts three times as many users from passive participants into active contributors?

  1. Sponsors. Most people join Facebook at the urging of friends, who provide some context as to “what is this, and what do you do here”. You listen to them because you trust them.
  2. Unofficial Mentors. Thriving online communities have naturally-emerging leaders - people who’ve been around, know some history, and take pride in helping new arrivals find information and understand the “unwritten rules” of the community.  Lithium actively encourages development of these “super-members”.
  3. Context. People join Facebook to stay involved with friends.  People join online communities because of a desire to learn or participate in a shared interest.
  4. Don’t-Make-Me-Think Contribution. Moving the community norm towards shorter, more frequent updates encourages spontaneity - people don’t feel the pressure to compose, re-read, and make sure they sound “smart enough”
  5. Critical Mass. Immediately seeing people you know, or an activity indicator like “418 posts in the past 24 hours” signals users that it’s worth their time investment to contribute.
  6. Immediate Perceived Value. The shorter the time between arriving and finding content that is informative, rare, useful, or about you, the more likely people are to stick around.

Twitter obviously has the #4 (140 character limit, no editing) and #5 (3700% growth last month) covered.

But if Twitter wants to keep competing with Facebook to be the platform for expression, they’ve got to incorporate more of #1, 2, 3, and 6.

Sponsors: Allow people to invite friends using the Facebook (and LinkedIn) social graphs.  Encourage them to add at least one “tag” about each friend and include “suggested people to follow” based on that tag with the invitation.

Mentors: I have no doubt that there are thousands of Twitter-ers who would volunteer to “mentor” a new Twitter joiner, especially if there was some cursory demographic matching.  (I would happily spend an hour or so with a new-to-Twitter product manager, for example.)

Context: Oh, I know the whole point is that Twitter is “whatever you want it to be!”  Yeah, sometimes the serendipity needs to be toned down - too many choices are paralyzing.

Allow new users to choose a “flavor” of how they think they want to use Twitter and show them suggested people to follow who use it that way.  This could be powered off a completely user-populated database - little work for Twitter themselves.

Immediate Perceived Value: Finding people outside your social web but inside your “interests web” is incredibly valuable.  Twitter should buy one of those user-populated directories already and MrTweet and make it an official part of the service.