HP Kittyhawk Case Thru the Lens of Customer Development

Way back in November 2012 I attempted to capture the key points of the HP Kittyhawk Case « Clayton M. Christensen through the “lens” of Steve Blank’s “Market Type” and “Customer Discovery”. /enjoy.

HP Case Table of Contents

Customer Discovery

Recap – Customer Discovery Points


  • Stop selling, start listening
    • There are no facts inside your building, so get outside
  • Test your hypotheses
    • Two are fundamental: problem and product concept
    • { Use Minimum Viable Product (MVP) for product exploration }


  1. Facts are outside the building, opinions are inside.
  2. Solve a problem that customers say is important and valuable
  3. Does the product concept solve that problem?

Exit Criteria

  • What are your customers top problems?
    • How much will they pay to solve them
  • Does your product concept solve them?
    • Do customers agree?
    • How much will they pay
  • Can you draw a day-in-the-life of a customer
    • before & after your product
  • Can you draw the org chart of users & buyers

Discussion / Conclusions

The “Kittyhawk Team” did not do Customer Discovery as described in “Start up Owners Manual. That is they did not employ “Minimum Viable Product” (MVP) strategies to rapidly research, and explore, “Product/Market Fit” with end-customers. They developed a “full” product in house and launched it without “experimenting” on their customers prior to launch. They made one decision, PDA vs. Nintendo in June ’91. They operated as if they were delivering an “enormously better” product into an existing market that understood, and needed, that improvement. They may have had lots of “hypotheses” along the way, but tested none of them over the 3+ year span of this case. For example, “Did they revisit the exploration of a $50 drive from a technical perspective after ’91?”, “If Cheap-Dumb-Drives were so important why didn’t they have a technical team working that angle to find the breakthrough?, “Why did it take until ’93/’94 to explore ‘marketing cost reductions’, like asking Nintendo about the accelerometer?”. What exactly did they explore over that time period? It was “How to sell what they already had”, no changes please.

The “Kittyhawk Team” did Customer Discovery the “old way”. They met over lunch in early 1991 and the “boss” selected the product. A few months later the Kittyhawk Team researches team dynamics and hires a “Can Do” core team ( ie people who can execute). They ensured that “marketing” would stick to the executive plan by hiring an inexperienced MBA, Jeff White, who had little hope of effecting change to the plan. In May they develop a charter that is all “execution”, there is no room for exploration or failure. In June 1991, about 5 months after the boss selected the product, they go to CES to talk to potential customers for the FIRST time. They decide on the mobile PDA market because Nintendo’s request is not technically feasible at that time. They then hire a market research firm to look at PDAs ( while HP are developing the product), but that firm can not “size” the PDA market or determine the leading customer targets. This is not a “RED FLAG” to the team. They stick to the schedule and deliver on time in June 1992. Even with the product in production they refuse to explore the markets that new customers bring to them. HP Kittyhawk then develops a refined version, the Kittyhawk II, for this same PDA market. The case leads us to believe that HP Kittyhawk marketing never discussed the cost of the accelerometer with Nintendo until after the Kittyhawk II ( so somewhere in 1993 to 1994) . In 1994, 3 years later, they are finally deciding if they should explore a new market for their drive. It’s the “PDAs are the market” or that “cheap dumb drive” market the “new” customers are asking about. In those 3 years they’ve never learned much, they just executed and delivered to schedule.

HP KittyHawk Case Timeline
HP KittyHawk Case Timeline

The “Kittyhawk Team” set product milestones as if they were delivering a “better/faster” product into an existing market even though they knew that PDAs were an emerging and unproven market. They did internal market research and launched. They did not believe that market adoption of their product would be an issue. They did not foresee any “chasms”. Even more astounding is that the “Kittyhawk Team” expected the Consumer Market Drive eco-system to operate in similar way to the “Enterprise Market” Drive eco-system that they were a leader in. Thus they began a “Customer Discovery” process only as “Triage”, or damage control, after having trouble selling the “Kittyhawk” drive for 3 years. Thus all of their “Customer Discovery” activity really happened with the “Kittyhawk” and “Kittyhawk II”. Sadly it was a very expensive MVP. Given that they spent an enormous amount of money on Kittyhawk with minimal return they ran out of the corporate goodwill required to experiment further and jeopardized the development of Enterprise Drives. HP DMD would close its doors in a few years time.

So What? (What Could You Have Done in Jeff White’s Place?)

You finished your MBA and got a great job at HP a few months ago. But now you’ve slaved for a few months and figured out that HP.DMD.Sales has no “lead customer” for the PDA market. Not one. In fact they don’t even know who the short list of prospects should be! You’ve figured out that there was no PDA market study, “The big boss” picked the “smallest box” on his desk over lunch. The “big guys” went to CES for a week, are now PDA experts, and made a decision. The project creed was cool at first, “I am going to build a small, dumb, cheap disk drive!”, but it’s now clear the creed is missing a key part, “For Who?” There are markets who would like a “cheap” drive, but HP.DMD doesn’t know those markets and engineering now says that we need a significant breath-through to get there. It’s starting to look like you’ve been placed on the “Titanic”. What can you do? You’re the marketing guy? What could you do now that you have taken this course?

The first thing you could have done is to have a “Customer Discovery Review” meeting in June 1991 prior to launching full product development. At that meeting you could review the “Customer Discovery Exit Criteria. You would present the concept of Market Type, the Business Model Canvases, the “Day in the Life of the Customers”, and org charts of your customers. From this it should be apparent that HPs plan was very weak. A reasonable outcome could have been.
“Delay full development for one month to find a lead customer and take a better look at the Nintendo opportunity”. Actions could be
a) Hire Market Research Firm to identify prospective lead customers in PDA industry
b) Allocate small amount of resources for “Customer Discovery” of cheap drive segment, Nintendo, on marketing and technical levels.

The outcome would most likely be “we’re not stopping, but we’ll do some of that work.” In the hallway expect that you’ll be told – But remember young man we need you to get out and support the sales team with Kittyhawk.PDA. Don’t be afraid, we know what we’re doing. It’s not as bleak as you say.

The second opportunity for you to speak up is after a month into the Market Research work. This is a serious event, a team of people who really know how to do the research couldn’t develop a list of lead prospects! This is a huge red flag. If you can’t force a Customer Discovery Review with this information then you are certainly on the “Titanic”. Thus you get the review scheduled. Now you are calm and you do a “Customer Discovery Exit” Review. At this point you recommend that “HP Kittyhawk.PDA” be delayed until lead prospects can be determined. Make the HP.Kittyhawk supporters work hard for their dream. Hey it might work if you put up a “challenge”. You then note progress on the “Kittyhawk.Nintendo” “skunk works”. The results of the Nintendo trip resulted in information that “accelerometer” is not needed and there are many other places where costs can be cut. The specifications for the Kittyhawk.Nintendo drive are becoming more realistic, but more R&D is still needed for the breakthrough. The Nintendo guys also hinted that other consumer devices may need this and they dropped some names we can follow up on.

The outcome of this meeting will not be pretty. It is likely that people will be fired. Possibly you for bringing this up. But most likely you will survive and your “tough spirit” will be rewarded for thinking more of the team than yourself.

If the decision after this is to plow ahead with no changes then you need to make a serious career choice.

More – Customer Development Manifesto and 9 Deadly Sins

The HP Kittyhawk Team was very close to the “story” told in “Startup Owner’s Manual – Chapter 1 – The Path to Disaster: A Startup Is Not a Small Version of a Big Company” and it committed many of the “9 Deadly Sins” mentioned in this chapter. It’s actions are inconsistent with the Startup Manifesto.

Market Type

  • Kittyhawk – New. New. New.
    (New Technology & New Product Markets & New Customer Types.)
  • New Technology
    • New leading Parameters in Physical Size and Capacity
  • New Product Markets
    • None of the markets discussed were in volume production at that time.
    • Each had not “Crossed the Chasm”
  • New Customer Types
    • HP DMD is totally new to Consumer Market Space.
    • It does not participate in Consumer Market and thus does not have working experience with how to develop products or communicate in this segment.

Business Model Canvases – Existing Drive & New Drive Kittyhawk

HP Server/Workstation vs Kittyhawk Canvases Compared
HP Server/Workstation vs Kittyhawk Canvases Compared


Value Proposition

  • Fastest Access Times
  • Highest Capacity
  • Feature Driven
  • High Price

Customer Segments

Workstation/Server Drive OEMs 80% & HP itself 20%

  • It has a internal customer who was/is a market leader in this segment.
  • The archetype is someone who is very similar to HP DMD. HP DMD and their customers are aligned in their belief of what constitutes a Workstation/Server Product.


  • not really mentioned. but one that is high-end existing feature driven B2B.
  • Likely direct sales.

Customer Relationships

  • Many years of delivering technically leading edge products.
  • Reference Selling and communication of past performance for it’s target segment.


Value Proposition

  • Guess – Industry "leap-frog” “size&capacity” with “3-foot drop”
  • Small
  • Dumb
  • Cheap
  • “droppable”

Customer Segments

  • All of them are a “Guess” because none have “Crossed the Chasm”
  • There are few (if any) internal “HP” customers to help drive the product. They have no “close friends” in the consumer space to “collaborate” with or whom they trust.
  • A TAM/SAM would show that revenue from all these products in best case is 3 Years out and will likely require a few iterations ( Think iPod .. It hit volume with the 4th Generation Model http://en.wikipedia.org/wiki/File:Ipod_sales_per_quarter.svg )
  • There is no existing market product space for the kittyhawk to enter. ( Thus the targets set can not be met. Period. )

Customer Archetype

  • Almost the exact opposite of HP DMD. “HP DMD did see that they needed to build a Cheap Drive”, “I am going to build a small, dumb, cheap disk drive!” – But they didn’t know “who” was going to use it.
  • Thus their archetype “DID NOT HAVE A PERSON IN IT” This is a “RED FLAG”


  • Guess – different than one needed for “Workstation/Server”

Customer Relationships

  • All less than 1 year. Often first time they have met.
  • No experience with “consumer” market segment.


  • sell drive unit (asset sale)

HBR Teaching Notes

Primary Learning Objectives

Students should be able to extract four important lessons from a well-crafted class discussion of the Kittyhawk project.
1. First, in commercializing a disruptive technology such as the Kittyhawk, manufacturers are faced with a fundamental strategic choice. One option is to accept the market’s needs as well-defined, and push the technology of the product to its limit to address the needs of established market. An alternative option is to accept the technology’s current capabilities as given, and then seek a market that will value the inherent attributes of that technology. Research has shown that although most winners in disruptive technological innovation pursue the latter route, the vast majority of established firms opt instinctively for the former (see Christensen, The Innovator’s Dilemma )
Iain – Most firms instinctively chase an “Existing Market” instead of looking for “Product/Market Fit”.
2. The second learning objective is to understand that customer input can be extremely misleading in an emerging market environment. Other methods of market research, which involve inexpensive explorators marketing of preliminary product designs in order to create knowledge about the market, can be much more reliable in synthesizing customer’s opinions.
The concept of Minimum Viable Product as Customer Discovery tool. The Kittyhawk was a very expensive MVP
3. When searching for a new market for a new technology, it is always important for a team to assess not just the probability that it’s own technology may or may not succeed, but also to take into account the risk that additional technologies which are critical for the ultimate end-use application, might not work. In situations such as this, the total probability of failure is the product of the probabilities of failure of each element of technology being developed for the application.
Each customer product market has to Cross The Chasm too
4. When a new technology fosters the development of a new market, an incubation period often exists through which potential customers need to work in order to understand how they might incorporate this new possibility. It is often difficult to shortcut this incubation period, corporate mandates of short payback periods notwithstanding.
New Market Chasms are much larger than one wants. One can’t force it. Again look at length of time MP3 players took to become volume products