PDF Eng

9 -8 1 2 -0 9 7
REV: AUGUST 24, 2012
FRANK V. CESPEDES
THOMAS EISENMANN
STEVEN G. BLANK
Customer Discovery and Validation for

Entrepreneurs
The purpose of a business is to create a customer. – Peter Drucker
Sixty years ago, Peter Drucker recognized innovation and customer creation as core
entrepreneurial functions.1 Starting a business requires more than product development and
marketing. But customer discovery and validation—the first steps in Steve Blank’s customer
development process—remain essential activities.2 Identifying your initial customers and
determining whether the opportunity you are addressing is important to them are crucial for
entrepreneurial success.
Marketers have developed research tools to assist with these activities. The tools are typically used
by established firms introducing new products or line extensions into existing markets. But
entrepreneurs face special challenges. Technological uncertainty and new-to-market possibilities
mean that key assumptions of familiar market research methods—knowing the variables for which
you seek data and having respondents who are able to articulate their preferences in relation to
known benchmarks—often do not apply in early-stage situations. Further, resource constraints mean
that entrepreneurs face time and money pressures when conducting necessarily iterative research.
Hence, entrepreneurs need more than market research in the traditional sense. They need a
process for turning their intuitions and enthusiasm into market-driven facts that can be used to
corroborate, pivot, or abandon an initial business model. The process requires (a) coherent and
falsifiable hypotheses about potentially relevant variables and (b) timely and affordable means for
investigating those hypotheses.
This note provides suggestions for practical customer research techniques suited for early-stage
ventures. It stresses the importance of learning, action, and failure in customer discovery and
validation. This may feel uncomfortable in an MBA program where the emphasis is legitimately put
on (a) comprehensive analysis aimed at establishing optimal solutions versus “good-enough” actions
aimed at moving to the next stage and (b) avoiding mistakes versus learning from frequent failures.
But the mindset we encourage in this note reflects startup reality.
________________________________________________________________________________________________________________
Senior Lecturer Frank V. Cespedes, Professor Thomas Eisenmann, and Lecturer Steven G. Blank (Stanford University) prepared this note as the
basis for class discussion.
Copyright © 2011, 2012 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-
7685, write Harvard Business School Publishing, Boston, MA 02163, or go to www.hbsp.harvard.edu/educators. This publication may not be
digitized, photocopied, or otherwise reproduced, posted, or transmitted, without the permission of Harvard Business School.
This document is authorized for use only in Thomas Cassel's EAS5450 Engineering Entrepreneurship I, Fall 2024-Cassel, Halak, Babin at University of Pennsylvania from Aug 2024 to Feb
2025.
812-097 Customer Discovery and Validation for Entrepreneurs
What Do You Want to Know about Potential Customers?

A hypothesis is more than an assumption. It is a testable proposition that, if answered, has
implications for decisions. Many entrepreneurs—and many managers in large corporations—define a
research issue as a broad area of ignorance. In effect, they say, “There are things that I don’t know.
But when the research results come in, I’ll know more and then I can figure out what to do.” This
boil-the-ocean approach makes it likely that the research will be expensive and time-consuming, and
will produce findings that are not actionable.3 Instead, startups—precisely because they face resource
constraints—need to be especially disciplined about their hypotheses in two areas:
 Customer Problem and Product. Entrepreneurs must formulate initial hypotheses about
potential customers’ problems and the venture’s solution. These hypotheses are core to
the customer value proposition component of a venture’s business model.
 Demand Creation and Buying Process. Entrepreneurs also require initial hypotheses
about how potential customers will find out about a startup’s product, who will buy it,
and who else will influence the purchase. These hypotheses are core to the Go-to-Market
component of a venture’s business model.
Only the entrepreneur can develop these hypotheses, and only a very confident or foolish founder
would ignore time-tested lessons about how best to do so. We first review some concepts helpful in
generating and framing hypotheses, and then discuss techniques for validating the hypotheses.
Customer Problem and Product

The first step in any customer discovery process is to verify that a market could exist for the
venture’s proposed offering. In other words, have we identified a problem that potential customers
want solved, and does our solution meet those needs? The goal here is not to determine how to build
the product. Rather, the goal is to determine whether the product should be built. In developing
hypotheses about this issue, it is helpful to keep certain distinctions in mind:
Problems versus Products This crucial difference is often confused in startup situations. As
Drucker noted, nobody really pays for a product—they pay for satisfactory solutions to their
problems.4 Harvard professor Theodore Levitt taught generations of students that “People don’t buy
2-inch drill bits, they buy 2-inch holes.” Levitt coined the term “marketing myopia” to describe the
common mistake of defining a business by product features rather than by customer benefits; for
example, assuming you are in the railroad rather than the transportation business or, more recently,
the newspaper versus the news business. Myopic managers fail to properly perceive market
opportunities, potential substitutes, and customer motivations to buy or not buy. 5
Entrepreneurs often are driven by a product vision and—consistent with the entrepreneurial bias
for action—eager to dive into product development. Steve Blank emphasizes the need for a customer
development process—to be run in parallel with a startup’s product development process—to verify
customer demand for a proposed solution. Blank notes that in most startups, “Product features
emerge by vision and fiat against unknown customer and market requirements… The job of
Customer Development is to get the company’s customer knowledge to catch up to the pace of
Product Development.”6
During customer discovery, it is important to understand the benefits at the heart of the venture’s
customer value proposition. This means confirming that potential customers have a strong unmet
need for the venture’s proposed solution—even if they may not be fully aware of that need.
2025.
Customer Discovery and Validation for Entrepreneurs 812-097
Customer discovery also entails fleshing out assumptions about how proposed product features do
or do not translate into customer-desired benefits, often based on direct observation of specific users’
experiences. The entrepreneur’s task is to develop verifiable hypotheses about questions such as:
What basic problem does the customer face in this aspect of their personal or professional lives? What
result(s) would yield more satisfaction than their current means for dealing with the problem? What
are the functional, emotional, and/or social dimensions of a satisfactory solution?
Benefits versus Status Quo Psychologist Daniel Kahneman won a Nobel Prize for
research conducted with his late colleague Amos Tversky on how people assess market choices.7
Their work reveals three characteristics of decisions about new products. First, customers evaluate
new products relative to a reference point, usually the products they already own or consume.
Second, people view improvements relative to the reference point as gains and treat any
shortcomings relative to the reference point as losses. Third—and most important for
entrepreneurs—perceived losses have a much greater impact on behavior than similarly sized gains,
a phenomenon Kahneman and Tversky called “loss aversion.” For instance, repeated studies across
cultures have shown that most people will not accept a $100 bet where there is only a 50% chance of
winning. In fact, the potential gains must be two to three times the potential losses before most
people will find a $100 bet attractive. Loss aversion leads to a bias for the status quo, which has been
demonstrated in research about choices regarding stock investments, automobiles, jobs, and labor-
management negotiations (e.g., “give-backs”), among other areas.
This status-quo bias means that most potential customers will not adopt a new product unless the
gains it promises far outweigh the perceived losses from switching away from a familiar solution.
How much better must a new product be? This is an empirical and situation-specific issue. But
studies have demonstrated that entrepreneurs are prone to overconfidence biases. 8 John Gourville
has shown that developers of new products, on average, overvalue the expected benefits of their
innovations by a factor of three, compared to the benefits ultimately delivered. Given that potential
customers overvalue losses by a similar factor, Gourville discerns a big mismatch between what
innovators think they are offering to potential customers and what those prospects will require to
change their behavior.9 He sees this mismatch as contributing to flawed decisions and high average
failure rates for innovations.
Due to these biases, gaining in-depth understanding of status-quo behavior is a crucial task of
customer discovery. New products often require consumers to change behavior and incur switching
costs in the form of transaction costs (e.g., activation fees), learning costs (e.g., how to use a new
software program), or obsolescence costs (e.g., the perception of my CD collection as a stranded asset
if I switch to downloading). These costs increase the perceived magnitude of “losses” in relation to
status-quo behavior and weigh heavily in customer decision making.
Lead Users versus Mainstream Customers Lead users are the early adopters of a new
product or service, willing to buy prototypes and accept evolving performance promises. Geoffrey
Moore emphasizes that early adopters are often visionaries with different purchasing behavior than
pragmatists—the later adopters who are crucial for scaling a business. In Moore’s account, new
ventures typically fail to recognize this and, lacking a plan for “crossing the chasm” between early
and mainstream customer behavior, frequently see sales stall after a burst of initial growth.10 But here
we are concerned with initial customer discovery and validation where, by definition and necessity,
the point is to identify early adopters. If a new venture does not survive the short term, its team need
not worry about the longer-term challenges of crossing the chasm. Indeed, research aimed at
identifying the needs of mainstream customers can actually be counterproductive if it leads an
entrepreneur to neglect the unique requirements of crucial early adopters.
2025.
What characterizes lead users? There is no simple answer to this question; research in a variety of
contexts has generated multiple and often conflicting criteria. 11 Simply labeling one’s first customers
as visionaries or evangelists is tautological and does not help much in establishing targeting criteria.
But Blank outlines a useful way to think about lead users, suggesting that they should have the first
four—and preferably all five—of the following characteristics: (1) the potential customer has a
problem, (2) she understands that she has a problem, (3) she is actively searching for a solution,
(4) she has cobbled together an interim, often home-grown solution, and (5) she has or can acquire a
budget to purchase a better solution. Blank says, “These people are … the ones [to] rely on for
feedback and for your first sales; the ones who will tell others about your product and spread the
word.”12
Consistent with Blank’s criteria, the process of customer discovery requires hypotheses about
those who—compared to the average or mainstream prospect—see your solution as providing far
superior benefits than status-quo solutions. To identify these potential customers, an entrepreneur
must understand what is (and is not) behaviorally compatible with their established usage systems.13
Entrepreneurs will encounter less resistance to innovations if their product requires only modest
change on the part of potential early adopters. For example, it was easier for Intuit to sell check-
writing software to consumers who already owned PCs.
Cultural factors are often relevant in identifying early adopters. In fact, some argue that the U.S.
has a productive history of entrepreneurship in part because the country has a larger number of lead
users willing to try new products in many categories. American farmers, for example, are among the
world’s fastest adopters of genetically modified crops. Silicon Valley likewise benefits from the fact
that U.S.-based companies tend to adopt new information technologies more quickly than their
counterparts in other countries do.14 Of course, this does not hold in all categories; Scandinavians, for
example, are often lead users of “green” solutions. But community norms often have implications for
lead-user identification.
Demand Creation and Buying Process

Hypotheses about customer problems help assess whether there could be a market for the
venture’s idea. But a “market” never buys anything; only customers do. Thus, you need to clarify
assumptions about how potential customers will find out about your product or service and how
they will buy it.
For many consumer products, purchase activity follows a simple hierarchy-of-effects model. First,
consumers become aware of the product. Second, given awareness, some are interested enough to
make a trial purchase. Third, given a trial, some like the product enough to repeat purchase it and
devote some portion of their future category purchases to the new brand. As shown in Figure 1,
companies must determine the tools best suited for moving consumers from one stage to the next.15
Figure 1 Simple Hierarchy-of-Effects Model and Typical Major Influences
Stage 0: Stage 1: Stage 2: Stage 3:

Unaware ------------------- Aware ------------------------- Trial -------------------------- Repeat
Advertising Product concept Product quality

Samples Distribution After-sale service
Coupons Price Switching costs
2025.
 For ventures in which personal selling is more prominent, the consumer hierarchy-of-effects
model is less useful. Especially for products that will be part of a firm’s capital budget,
multiple parties are typically involved in the buying process. Many models are potentially
relevant in these situations, but the common core is the notion of a “buying center” that must
be understood and navigated. This refers to the roles that customer participants can assume in
a purchase decision. Six roles, some of which may be filled by the same person, are
encountered in many B2B situations: 16 Initiators recognize that some company problem can
be solved or avoided by acquiring the relevant product or service.
 Gatekeepers may have the title of buyer or purchasing manager and are paid to keep track of
vendors’ offerings. They control the flow of information and often access to other decision
makers in the relevant buying center.
 Influencers are those who have a say in whether a purchase is made and what is, or is not,
bought. Influencers may not be able to approve a purchase (affirmative power), but in some
cases they can exercise veto power over a potential purchase decision or vendor. Technical
support units (e.g., IT departments) often also have influence over purchase decisions,
focusing on issues such as vendor reputation and ease of implementation.
 Deciders can say yes or no to the contemplated purchase. Often, deciders are not the ones
who formally sign off on purchases or negotiate price or other terms and conditions; that may
be left to others. In general, there is only one decider per sale, but it may also be a set of people
such as a team, board, or committee. Deciders tend to focus on economic concerns such as
total cost of ownership, fit with budget, return on investment, and productivity impact.
 Purchasers are concerned with actually obtaining the product or service. Corporate or
divisional purchasing departments usually fill the purchaser role, but for some products,
users may be authorized to make purchases.
 Users actually consume the product or service, and typically seek the best solution for their
problem in terms of criteria such as product performance, ease of use, and reliability.
Understanding these roles is critical, since the various parties to a purchase will typically have
different ways of assessing value. During customer discovery, it is important to recognize that
current users in a category may be the source of the venture’s idea, but users are often not the
deciders or purchasers, whose support is necessary to acquire a customer and make a sale.
How Can You Validate Demand?

The rest of this note focuses on ways to validate hypotheses about customer demand. In doing so,
we make two assumptions about the situation confronting most startups and the mindset appropriate
for validating demand. First, few entrepreneurs get it right on their first try. Most invest in many bad
ideas before they succeed. But it is better to fail and iterate before big sums are spent on building and
launching a new product. Yet this requires entrepreneurs to admit they are wrong, overcoming their
natural ego-defensive tendencies.
Second, entrepreneurs usually need both qualitative and quantitative validation techniques, and
often cycle between the two types. Qualitative techniques such as focus groups allow probing for
additional information based on respondents’ statements or behavior, and are therefore especially
useful when the entrepreneur has only a rough sense of what he or she is looking for. (See Appendix
A for a description of focus groups and how entrepreneurs can use them.) Qualitative techniques are
2025.
also useful whenever quantitative research generates surprising results—for example, low adoption
for a feature expected to have strong appeal. However, qualitative research is subject to multiple
interpretations and vulnerable to biases and blind spots. By contrast, quantitative research seems
more objective and exact, but it only yields the right answers when one asks the right questions.
To oversimplify, qualitative research surfaces questions that are then answered through
quantitative research; quantitative research can surface anomalies that are then explored through
qualitative research. Effective entrepreneurs know when they have learned enough to stop cycling
between these techniques and either declare a hypothesis validated or pivot to a new one.
Below, we discuss various research techniques and order them in terms of when they are first
likely to be used by a startup: (1) customer surveys, (2) letters of intent, (3) usability tests, (4) market
trials, (5) split tests, and (6) Net Promoter Score. Exhibit 1 summarizes the strengths and limitations
of these techniques and the settings for which they are best suited.
The tools on this list are not mutually exclusive. Some can be used in combination; for example,
split tests may be conducted during a market trial. Nor is the list exhaustive. In particular, an
important technique in startup settings—face-to-face interviews with potential customers—is
complex enough to warrant a separate note (“Customer Visits for Entrepreneurs,” HBS No. 812-098).
We also do not discuss how entrepreneurs can use published data, including information
available on the Web, to learn about potential markets and target customers. In general, we find that
many first-time entrepreneurs—especially those with MBAs—spend too much time trying to
precisely size a market. Investors do want evidence that a startup is pursuing a large and growing
market, but most investors also know that estimates of market size are subject to manipulation
because market boundaries are rarely clearly defined—a fact that keeps an army of antitrust lawyers
busy. Investors also know that new markets are rife with uncertainty, so market-size projections
based on published sources are often guesses. Therefore, most investors prefer to see real data on
customer reactions, collected through research methods such as those described in this note.
Customer Surveys
Entrepreneurs often survey prospective and current customers to solicit information on topics
such as frequency of category purchases, preferences regarding product features, purchase intent
for/satisfaction with current category offerings, and the appeal of different methods for learning
about and procuring products. In general, customer surveys are better suited for disconfirming early-
stage hypotheses about demand than for validating them. If an entrepreneur hypothesizes that a
certain customer segment will find the startup’s offering attractive but a well-designed survey shows
that prospects in that segment have little or no interest in that offering’s key features, then the
entrepreneur probably should pivot to a new hypothesis. If, on the other hand, a survey shows strong
segment interest in the features, there is still more work to do.
Results from customer surveys should be “handled with care” for many reasons. Respondents
often have difficulty articulating their preferences about the separate features of products because
those features are experienced as a bundle. (Conjoint analysis, described in Appendix B, is a survey
technique that addresses this shortcoming.) Likewise, surveys do not easily evoke the loss aversion
that may deter customers from abandoning familiar products to adopt new ones.
Nonetheless, customer surveys are widely used by startups because they are inexpensive to
implement—especially with online tools. To maximize the learning value of surveys, entrepreneurs
should follow certain practices. Outlining those practices is beyond the scope of this note; readers
2025.
may wish to consult a white paper on the topic by SurveyMonkey. 17 In brief, thoughtful design is
important in three areas:
 Formulating Specific Questions. Common mistakes include crafting confusing or leading

questions. For example, “How do you view the convenience of our customer support
system?” is a leading question; less leading would be “Does our customer support system
meet your needs?” In general, it is easy to write biased and misleading questions (just look at
political surveys) and difficult to write unbiased and useful ones.
 Survey Length. Response rates and the thoughtfulness of responses both decline when
surveys are too long or repetitive. Pilot testing the survey with individuals who are
representative of the target respondents is a crucial step for ensuring that the survey’s length
and the sequence of questions are appropriate and that the questions are worded clearly.
 Recruiting Respondents. Entrepreneurs sometimes fail to recruit survey respondents who are
representative of target customers. Instead, they rely on convenient sources such as business
school classmates and Facebook friends. Likewise, they often fail to consider the sample size
necessary to make statistically significant inferences about survey responses.
Letters of Intent
A letter of intent (LOI) can be useful for validating demand before the startup has developed its
product offering. After exposing a potential early adopter to the product—perhaps a crude prototype
or even just a sales pitch explaining the planned product features—an entrepreneur asks the prospect
to sign a letter of intent. In the LOI, a potential customer agrees to try/buy a proposed product with
specific features, once it is ready. An LOI is obviously less of a commitment than an actual sale would
be. However, in the words of one entrepreneur, “While legally non-binding, customers who have
signed an LOI are less likely to disappear or make themselves hard to get a hold of.”18 Consequently,
if few prospects will sign an LOI, then hypotheses about customer demand have been disconfirmed
and an entrepreneur must pivot to a new value proposition. Conversely, if most prospects asked to
sign an LOI do so, then initial demand has probably been validated. Also known as a “memorandum
of understanding,” such an agreement typically is legally non-binding.
Entrepreneurs can learn about customer needs by negotiating LOI terms with early adopters (see
Exhibit 2 for a sample LOI in a B2B context). But early adopters often press for custom features that
are not on the startup’s initial product roadmap. Entrepreneurs must sort out when such features
would be valued by other customers and when they are the idiosyncratic needs of a single prospect.
Usability Tests
Usability tests normally involve asking current or prospective customers to complete specific tasks
(called “use cases”) with a working prototype or an actual product version. The goal is to identify
potential problems with ease of use. Usability tests are conducted during all stages of product
development. Late in the product development process, usability tests may involve collecting
quantitative data (such as the time required to complete registration; or the number of clicks on a
given tab) in order to improve specific design elements (e.g., whether buttons should be more round).
But when an entrepreneur is still validating demand hypotheses, the objectives for usability
testing are different. The goal is to rule out the risk that a flawed initial product design will
inadvertently lead an entrepreneur to conclude that demand for the venture’s proposed solution is
low. Put another way, if early demand is weak, it is crucial to know whether prospective customers
2025.
(a) have little need for a solution to the problem that the venture has targeted or (b) do want a
solution, but simply cannot figure out how to use the venture’s initial product, because it is badly
built or designed.
Qualitative testing with a few subjects is often adequate to meet this objective. For example, Jakob
Nielsen of Sun Microsystems showed that 80 percent of the major problems with early versions of
websites can typically be found with samples as small as five. 19 In such tests, subjects’ interactions
with the product are observed closely and often recorded on video. Subjects are usually asked to
“think aloud” about the choices and barriers they encounter as they proceed through an assigned
task. Moderators must set the stage properly, emphasizing that the goal is to test the product, not the
subject’s abilities. Likewise, it is crucial to avoid biasing the results by offering your own opinions or
by coaching subjects (“Here, let me show you how that works…”).
The experience of the Dropbox team shows how useful—and emotionally wrenching—usability
tests can be. Founder Drew Houston recounted how not one of five mainstream consumers recruited
from Craigslist could install and use an early version of the startup’s online file storage service.
“Watching them fail,” he recalled, “was excruciating… probably the most painful day we had as a
team, but afterward, we created a list of 70 things to fix.”
Usability testing is readily affordable, even for bootstrapping entrepreneurs. For Web-based
products, online services now allow entrepreneurs to outsource the task of recruiting subjects and
conducting usability tests; for example, for $39, UserTesting.com will deliver an online video of a
typical consumer “thinking aloud” while completing a specified task.
Market Trials
By first testing products with a subset of their intended customer base, entrepreneurs can validate
demand and also refine product features and marketing approaches before investing in a full-scale
launch. In consumer packaged goods (CPG), a test market limits a new product launch to one
demographically representative city. The test offers those consumers a finished product at a regular
retail price. The firm’s sales, marketing, operations, and customer service functions are all involved in
the test, which is used to refine marketing plans (e.g., ad messaging, store promotions, pricing) and,
in most instances, to validate consumer demand. Firms frequently abandon new product concepts
following lower-than-expected initial sales and repurchase rates in a test market.
Market trials are widely used by entrepreneurs. For example, the founders of Rent the Runway,
an online service that rents designer dresses, conducted a trial for a limited set of paying customers a
few weeks before their public launch and gleaned lessons that affected product features (e.g.,
minimum viable inventory, or the need for “stylists” to give advice on dress selection and to explain
how the service works), marketing (e.g., the need for a waiting list to manage inventory availability),
and operations (e.g., the inadequacy of their initial order-processing system).20
There are three key issues to consider in designing a useful market trial:
 Customer Selection. Customers in a market trial should be representative of early adopters.

In B2B ventures, the time and cost of managing a trial often means that only a few dozen
testers—and sometimes far fewer—may be recruited. If the venture’s product impacts a
mission-critical customer activity, ideal test customers can be tricky to recruit: The impact may
mean high potential value but also more reluctance to participate because of the risk to a key
customer activity. Likewise, participating testers often have a vested interest in customizing a
product to their needs. A common challenge is dealing with the “feature creep” that can result
2025.
from market trials. For this reason, Blank argues that the goal “is not to gather feature
requests so you can change the product. Instead, your purpose is… to find [whether there are]
customers for the product you are already building.” 21
 Timing. A market trial is defined as “a real world test of a system after it has passed all
laboratory tests.”22 But that may not be true for a minimum viable product (MVP), where
certain features may not be present and performance bugs often still are. Early trials involving
MVPs may blur the distinction between regular products and product simulacra—prototypes
and other simulated experiences that capture some but not all of a product’s planned features,
often by relying on makeshift, temporary technology. Such blurring may entail trade-offs that
are explored in greater detail in a companion note, “Hypothesis-Driven Entrepreneurship:
The Lean Startup” (HBS No. 812-095). In brief, early feedback that comes from putting an
MVP in the hands of real customers can be very valuable before investing heavily in product
development or marketing. However, an early market trial with an incomplete product can
have reputational and competitive consequences that must be managed.
 Information Flows. Entrepreneurs must control information flows to and from a trial site. In
B2B settings, non-disclosure agreements can raise concerns when testers are being recruited.
Also, customers may find information requests to be intrusive. For example, assessing the
product’s economic value to a prospective customer may mean going beyond immediate
users and gaining access to a range of people within the potential buying unit.
How does an entrepreneur know when a market trial is succeeding? If customers are charged a
regular price, as in CPG-style test markets, then conversion-funnel metrics can be used to validate
demand. Following the hierarchy-of-effects model in Figure 1, a conversion funnel shows the rate at
which prospects who are contacted with a marketing message or by a salesperson take action (such as
registering at a website or agreeing to take the sales call), then the rate at which these “activated”
prospects become paying customers, then the rate at which those customers engage in actions
showing deeper commitment (such as referring friends), and finally, the rate at which customers
repurchase the product. In “Startup Metrics for Pirates,” investor Dave McClure tells entrepreneurs
building Web-based businesses to track “AARRR”: Acquisition of a cohort of website visitors; their
Activation into first-time users; Repeat visits by users; their Referral of new prospects; and their
Revenue generation.23 Conversion funnels can be calculated for businesses that rely on mass
advertising methods or direct sales. They are especially easy to monitor for Web-based businesses, for
which Google Analytics and other tracking services can provide real-time data.
How high must conversion rates be to validate demand hypotheses? This varies by business. In
established markets, competitive benchmarks are often available. In new or existing markets,
entrepreneurs should be able to estimate the lifetime value (LTV) of a typical early adopter using data
on the marketing costs incurred and the revenue collected at each stage of the conversion funnel. If
LTV is positive, net of the average customer acquisition cost (CAC), then demand has likely been
validated in a market trial. Even if LTV minus CAC is negative for trial customers, many startups can
improve conversion rates through ongoing optimization efforts, so a reasonable extrapolation of
conversion rate trends may show that demand is adequate to continue investing.
Split Tests
In split tests—also called A/B testing after the practice of assigning letters to each variation—
different versions of a product (or a product simulacrum) are offered to separate samples of
customers at the same time. Reactions to a baseline control version are compared to reactions to one
2025.
or more test versions that alter the variables under study; for example, product features, marketing
messages, or price. Catalog companies have long sent different catalogs to different customers to test
optimal product selection and CPG companies routinely test different versions of products. Split tests
are easy to conduct online, where changes can be made quickly and metrics can be readily recorded.
For Web-based ventures in particular, most elements of a website can be split tested.
Spilt testing is useful for optimizing product design and marketing programs. Such optimization
efforts typically entail experiments that are modest in scope. For example: Will conversion rates from
website visitor to registered user improve if we make the “Try It Now” button bigger or change its
color? For early-stage ventures that are still validating demand, however, split tests often entail
bigger changes in product features or pricing, because the goal is to prove whether anyone will use a
proposed feature, not to increase usage rates by optimizing a feature’s design.
For example, Grockit, an online test-preparation startup employing peer social learning, used split
tests to validate demand for core features. The founder’s original hypothesis was that more is better
when it came to features that provided Grockit’s users with social communication tools. But split
tests indicated steeply diminishing returns for those extra features and showed that a key to user
engagement was a combination of social and solo features. So Grockit added a solo-study mode, with
game-like quests and levels of proficiency.24
When conducting split tests, entrepreneurs should follow the basic principles of experimental
design, and can find guidance online if they lack statistical training. 25 In particular, it is crucial to (a)
create a control group, (b) ensure random selection for control and test samples, (c) use big enough
samples to get statistically significant results, and (d) use proper statistical methods in drawing
inferences about results.
Also, the small scale of a startup’s user base may limit the rate of learning because only a few
experiments can be conducted at any one time due to sample-size requirements. At the online dating
startup Triangulate, for example, founder Sunil Nagaraj discovered that, “With the modest scale of
our current user base, you can’t run too many experiments in parallel, because you can’t split the base
too many times. [This] puts a gate on the amount of development work possible during any given
period.”26
Net Promoter Score

Net Promoter Score (NPS) is a measure used to track overall customer satisfaction at the product
level. Developed by Bain & Company and Satmetrix Systems, NPS asks a product’s customers, “On a
scale of 0 to 10, how likely is it that you would recommend XYZ to a friend or colleague?”
Respondents are usually also asked an open-ended question about the main reason behind their
score. Those who score 9 or 10 are termed “promoters,” those who score 7 or 8 are “passives,” and
those who score 0 through 6 are “detractors.” NPS is the percentage of customers who are promoters
minus the percentage who are detractors, and it has been shown to correlate with measures of firm
growth and profitability. Compared to rivals, companies with a high NPS have higher customer
retention rates, lower customer service costs (due to fewer service problems and to the fact that
longer-tenured customers have more experience using the product), and lower customer acquisition
costs (due to stronger word-of-mouth referrals). Established companies use Net Promoter to identify
product or service problems and to spur improvement initiatives.
Startups that have not yet validated demand but have launched a product—either in a market trial
or at full scale—can also use NPS. By examining NPS by customer segments, they can test hypotheses
about which types of customer will find the venture’s product most appealing. Why not simply
10
2025.
observe initial adoption rates by segment? Because for many products, satisfaction depends on
ongoing product use. NPS can give an early indication of a product’s performance before repurchase
rates are known. Likewise, word-of-mouth referrals impact customer acquisition costs. NPS can help
predict whether referrals will occur before they actually happen. Finally, because NPS can be
implemented quickly and inexpensively to deliver fast feedback, it is well suited to lean startup
priorities.
Conversely, by tracking NPS over time, entrepreneurs can avoid premature scaling—i.e.,
investing aggressively before validating their business model hypotheses. If a startup’s NPS
improves steadily and passes a threshold level, an entrepreneur can conclude that her product
satisfies the target market.
Average NPS varies widely by industry, but across sectors, very high scores tend to range
between 60 to 80 (e.g., 60 for JetBlue, 70 for Amazon, 72 for Apple, 78 for USAA homeowners
insurance, 82 for Trader Joe’s). According to Netpop Research, several leading Internet services have
lower NPS scores; e.g., 50 for YouTube, 36 for Facebook, 33 for Groupon, 30 for LinkedIn, 29 for
Craigslist, and 11 for Twitter.27 According to Reichheld and Markey, the average U.S. company has
an NPS score between 10 and 20. But due to the “status quo” bias cited earlier, most startups must
offer a significant improvement over incumbents’ products and must therefore focus on relative rather
than absolute NPS. This often implies an NPS at least 10 to 20 points higher than the scores of their
incumbent rivals.
Conclusion
Customer discovery and validation are intertwined and iterative. Therefore, entrepreneurs must
approach these tasks as a continuing process, not a one-time research project; as a motion picture of
learning and adjustment, rather than a one-time data snapshot. Also, because entrepreneurs face
resource constraints—especially with respect to time—they must be selective in applying the research
methods described in this note. The methods have different strengths and weaknesses and
entrepreneurs must choose the best combination for their ventures’ situations. They also must know
when to say, “That’s enough.” As one venture capitalist notes, “most startups don’t starve, they
drown” in too many ideas pursued simultaneously. Consequently, entrepreneurs should use research
to set priorities and make decisions, not simply to collect more information.
Finally, new ventures are ultimately about today and tomorrow in a market, not yesterday.
Therefore, entrepreneurs should supplement any research results with direct experience, judgment,
and—even as they listen closely and learn from others—a healthy distrust of conventional wisdom.
After all, other people—including this note’s authors—often “like to give good advice as consolation
for the fact that they can no longer set bad examples!” 28
11
2025.
Exhibit 1 Summary of Entrepreneurial Demand Validation Techniques Discussed in This Note
Surveys
Benefits: Low-cost and relatively easy to implement; an accessible start to learning about potential
customers and/or relevant product features.
Limitations: Deceptively difficult to write and ask non-biased objective questions; can you recruit
respondents who are representative of early-stage target customers?
Use setting: Gathering initial and general information about category purchases and preferences;
typically better for disconfirming early-stage demand hypotheses than for validating them.
Letters of Intent
Benefits: A means to increase purchase/use commitment of potential customers and learn about the
presence or absence of demand for the proposed product offering.
Limitations: Not legally binding; risk of gathering one-off preferences and “requirements” that are
not representative of other customers.
Use setting: Early concept stage of a venture’s product and customer development process.
Usability Tests
Benefits: Can help entrepreneur determine whether prospective customers have problems with
current product design, versus a lack of demand for the proposed solution in any form.
Limitations: Helps entrepreneur to understand specific product-usage requirements, but not the
nature or scope of relevant customer needs and propensity to purchase.
Use setting: Early prototype stage of product and customer development.
Market Trials
Benefits: Can help to validate demand and refine product features and/or marketing messages to
initial target customers; an opportunity to gather, measure, and analyze conversion funnel metrics.
Limitations: Need to determine relevant customer/market scope for trial to be useful; time and cost
of managing information flows with trial customers, especially in B2B ventures.
Use setting: When venture is beyond early-prototype or MVP stage of product development.
12
2025.
Split Tests
Benefits: Helps to identify key product and/or marketing variables; especially easy to conduct and
assemble split-test data in online ventures.
Limitations: Requires coherent, consistent experimental design procedures and sample sizes.
Use setting: Beyond concept and prototype stages, when entrepreneur is investigating choices about
specific product features and/or marketing variables (such as price) and their impact on demand.
Net Promoter Score
Benefits: Track overall customer satisfaction with venture’s product and customers’ willingness to
refer; help to test hypotheses about which customers find the venture’s product most appealing.
Limitations: Customer “status quo” biases require high relative NPS to validate demand, but how
high is often unclear.
Use setting: Post introduction and launch, when venture has a product offering that has been
purchased and used by target customers.
Source: Casewriters.
13
2025.
Exhibit 2 Sample Letter of Intent
STARTUP agrees to provide, without charge or fee, its XYZ platform on or before DELIVERY DATE
to CUSTOMER, which will provide the following features:
 Functionality XYZ and ABC (one-sentence description of each)
STARTUP recognizes the sparseness of detail in the aforementioned features and therefore agrees to
update CUSTOMER on development progress on the following dates:
 DELIVERY_DATE minus six weeks
 DELIVERY_DATE minus four weeks
 DELIVERY_DATE minus two weeks
In exchange, CUSTOMER will [make itself available on the] aforementioned dates to provide
feedback and guidance as to functionality and requirements. During and after the development
process, CUSTOMER agrees to fairly evaluate the product for use as their XYZ platform for a free
trial ending DELIVERY_DATE plus four weeks. Additionally, CUSTOMER agrees to use STARTUP,
at a 12-month rate of $XYZ/month, as their XYZ platform if STARTUP satisfies their needs.
It is the intent of CUSTOMER to enter into a formal agreement with STARTUP by

REASONABLE_DATE. Due to the sensitive nature of the discounted pricing, CUSTOMER hereby
agrees not to disclose pricing information and the terms of this Letter of Intent to any other person or
entity that is not an employee or associate of the CUSTOMER.
Agreed and understood by:

Designated STARTUP Representative Designated Customer Representative
Author’s Comments:
1) We could have titled this a Memorandum of Understanding (“MOU”) as opposed to an LOI,

but we wanted to strongly signal that they are demonstrating intent to buy by signing
something called a Letter of “Intent.”
2) Brief description of functionality. This is simply to get both parties on the same page and set
expectations.
3) Development process updates. This was to ensure us face-time with our customer. Should
our customer get busy, we didn’t want to be easily forgotten or ignored.
4) Pricing and price discovery. Our next LOI recipient got a significantly higher price.
5) Intention discovery. Do they actually want to buy our product? Well, let’s test it very plainly
with “It is the intent of CUSTOMER to enter into a formal agreement with STARTUP, by
REASONABLE_DATE.” We have de facto started the sales cycle.
6) Terms. Can we sign them up for 10 months? Three months? Month-to-month? Per usage?
7) Disclosure. This is our effort to keep our customers from chatting with another about the
terms of the LOI as we go about price and intention discovery.
Source: Patrick Vlaskovits, “Seller Beware: Customers Have Their Own Agenda,” December 15, 2009, “Market by Numbers”
blog, http://market-by-numbers.com/2009/12/seller-beware-customers-have-their-own-agenda/, accessed
November 20, 2011.
14
2025.
Appendix A: Focus Groups

In its classic form, a focus group is 8–10 users or purchasers in the relevant category (such as
consumers talking about apparel, corporate controllers talking about budgeting software, or patients
talking about healthcare services). Groups typically should not include close friends because they
tend to reinforce each other’s views, which can skew the discussion. Participants are brought to a
common location—ideally, a quiet room with a circular table—and often compensated for their time.
A moderator conducts the session, which typically lasts 60–90 minutes and may be recorded. At
sessions conducted in market research firms’ offices, there may be a one-way mirror so others can
observe the discussion. Discussants are informed of such arrangements and promised anonymity.
Online video technology has reduced the cost of conducting focus groups and makes it possible to
conduct the activity, with appropriate group interaction, across different locations.
Running a focus group requires skill, so entrepreneurs sometimes rely on market research
professionals to perform that task. When they do so, the entrepreneurs should attend the sessions:
Being immersed in a problem, they often glean insights that hired professionals miss. But
entrepreneurs can also run their own focus groups by observing some guidelines:
After a brief introduction in which the moderator explains the objectives and ground rules and
asks the participants to introduce themselves, the discussion typically begins with each person
commenting on how they use—or when and how they last bought—the product under discussion.
Then, the moderator transitions to questions through which the participants’ likes and dislikes are
listed. To stimulate conversation at the outset, the moderator may ask the participants to spend a few
minutes writing down ideas. Throughout the session, the moderator may record key themes on a flip
chart or whiteboard for later reference.
It is important to use open-ended and neutral questions to avoid “leading the witnesses.” The
moderator should never disagree with a participant’s statements or take sides in a debate, but instead
ask, “Do others agree?” For entrepreneurs passionate about their concepts, avoiding defensive
reactions to criticism takes discipline. It is also important that everyone has a chance to speak and
react to others’ comments, because a goal of most focus groups is to have current consumers
verbalize their latent feelings about the category. Often, this only emerges through group interaction.
The purpose of focus groups is exploratory learning about “consumers’ opinions, semantic
structure, usage patterns, attitudes, and buying processes.” 29 The term “semantic structure” refers to
how people talk about, frame, and relate to the category. To learn this, entrepreneurs must hear them
speak and watch their reactions, especially in high-involvement categories with emotional, status, or
lifestyle associations. In a venture for a hand lotion product, for example, a male entrepreneur saw 10
female participants casually wait with their hands on the table for the discussion to begin. Seconds
after the topic was announced, all hands were either under the table or out of sight. The discussion
revealed much anger about the fact that, in the country where the focus group was held, women were
expected to do all the housework and still have soft hands. Insights from the focus group resulted in
a changed product.
Such insights can be powerful but, as one entrepreneur notes, in a focus group setting, “The
answer to any question that starts with ‘do you want’ or ‘are you concerned about’ will always be
‘yes’!”30 Thus, even if all participants say they will buy a startup’s product, do not conclude that
demand has been validated. On the other hand, a unanimous negative reaction can probably be
viewed as evidence of lack of demand. In short, focus groups are better for disconfirming hypotheses
than for validating them. Like other qualitative research methods, focus groups are also good for
generating new hypotheses that must subsequently be validated through other techniques.
15
2025.
Focus groups are typically best suited for exploring current usage by mainstream customers, not
lead users. Hence, the technique is most often used for line extensions in consumer goods and new-
feature additions in industrial products. If your venture creates a new product in a new market, then
take focus group responses with a warehouse of salt, because current users are not good at
verbalizing responses about products they have not experienced. 31
16
2025.
Appendix B: Conjoint Analysis

Conjoint analysis is used in new-product development for consumer and B2B goods.32 It breaks a
product concept into relevant attributes and estimates the value that potential customers attach to
each attribute. For example, choosing a health-care facility might depend on attributes such as travel
time, time lag in getting an appointment, waiting time once you are there, hours of operation, and
facility type (e.g., doctor’s office or health-care center). When evaluating a bundle of product
attributes, customers are not good at valuing specific attributes, because they buy and experience the
bundle. In particular, statements about specific attributes may be unreliable when attributes interact
(e.g., travel time to a facility may be less burdensome if the facility is open late; reactions to a $200
price may vary by brand). For these reasons, tools that ask respondents to rate the importance of a
series of product attributes, considered independently, often yield flawed results. In conjoint analysis,
attributes are CONsidered JOINTly.
Conjoint analyses use two basic methodologies. The discrete-choice method asks customers to
choose their favorite bundle of attributes from among a small set (often just a pair) of different
bundles, and to make such choices repeatedly, with the bundle attributes being varied each time. The
single-concept method asks respondents to rate the overall attractiveness of many different bundles
on a high-to-low scale, assessing one bundle at a time. In most situations, the two methods yield
similar conclusions, but each has different advantages and shortcomings. 33 In brief, even if a
respondent has low interest in the category overall, discrete-choice analysis will nevertheless include
all respondents’ input in estimating attribute values—introducing “noisy” data from likely non-
buyers. Conversely, with single-choice analysis, category enthusiasts may give very high ratings to
all bundles, making it difficult to discern the strength of their preferences for different attributes.
With either method, after preference data is collected, mathematical algorithms are used to infer
the value that respondents attach to each product attribute by examining their pattern of preferences
for different bundles. By summing the values (also called “utilities”) across all attributes and
confirming that the total exceeds the proposed purchase price, one can predict the probability that the
respondent will buy a product with a given set of attributes. In most situations, the resulting
prediction is far more reliable than purchase intent data collected in conventional customer surveys
that ask, “On a scale of 1 to 7, how likely are you to buy XYZ?”
Properly used, conjoint analysis can provide validated learning about two core issues confronting
startups. First, it elicits quantitative data for testing hypotheses about a product’s value proposition;
specifically, what product attributes do customers value and which design trade-offs do they find
preferable or acceptable. Second, it helps entrepreneurs understand which attribute combinations are
most and least valued by different customer segments and thus can help identify lead users.
To use conjoint analysis effectively, certain conditions must hold. First, in contrast to focus groups,
the startup team, not the respondent, must specify the attributes to examine. The technique cannot
indicate the absence of an important attribute, so the team must be confident that it has included all
relevant attributes. This is possible with a “better/faster/cheaper” improvement in an existing
product category. However, with a fundamentally new product category, an entrepreneur is likely to
face considerable uncertainty about the nature of both customer problems and potential solutions.
Second, conjoint analysis requires respondents who are familiar with the attributes and the
product category overall. Again, this can be challenging with a fundamentally new product.
Remember the maxim attributed to Henry Ford: “If I had asked people what they wanted, they
would have said faster horses.”34
17
2025.
Conjoint analyses can be powerful but complex to design. Consequently, proper training or
assistance from a market research professional can help to ensure good results.35 Further, conjoint
analysis can be expensive: It may be necessary to retain a market research service to structure and
interpret the analysis, large sample sizes are often required (roughly 200 respondents for each
customer segment evaluated), and each respondent might spend 15–30 minutes answering questions.
Researchers have lowered costs by using online tools to recruit respondents and conduct conjoint
studies, but the technique is often still too expensive for bootstrapping entrepreneurs.
18
2025.
Endnotes
1 Peter F. Drucker, The Practice of Management (New York: Harper & Brothers, 1954), pp. 37-38.
2 Steven Gary Blank, The Four Steps to the Epiphany, 3rd ed. (Quad Graphics, 2007).
3
For the limits of this approach to research, see Alan R. Andreasen, “Backward Market Research,” Harvard
Business Review (May-June 1985).
4 Peter F. Drucker, Managing For Results (Pan Books, 1964), chapter 6 (“The Customer Is the Business”).
5
Theodore Levitt, “Management Myopia,” in Innovation in Marketing: New Perspectives for Profit and Growth
(New York: McGraw-Hill, 1962).
6 Blank, Four Steps, p. 27.
7
For a witty and accessible review of this research and its application to health and environmental issues, see
R. Thaler and C. Sunstein, Nudge: Improving Decisions about Health, Wealth, and Happiness (New Haven, CT: Yale
University Press, 2008). For the core academic research and concepts, see D. Kahneman and A. Tversky, eds.,
Choices, Values, and Frames (Cambridge, UK: Cambridge University Press, 2000).
8
For example, see A. Cooper, C. Woo, and W. Dunkelberg, “Entrepreneurs’ perceived chances for success,”
Journal of Business Venturing 3, issue 2 (1988): 97–108.
9 John T. Gourville, “Eager Sellers and Stony Buyers: Understanding the Psychology of New-Product
Adoption,” Harvard Business Review (June 2006).

10
See Geoffrey A. Moore, Crossing the Chasm: Marketing and Selling High-Tech Products to Mainstream
Customers (New York: HarperCollins, 1991; revised 2002) and Inside the Tornado: Strategies for Developing,
Leveraging, and Surviving Hypergrowth (New York: HarperCollins, 1995; revised 2004).
11The best place to begin understanding this research is the work of Erich von Hippel: The Sources of
Innovation (Oxford University Press, 1988), Breakthrough Products with Lead User Research: Methods for Uncovering
the Ideas and Prototypes of Leading Edge Users (Cambridge, MA: MIT Press, 1998), and Democratizing Innovation
(Cambridge, MA: MIT Press, 2005).
13 Gourville, “Eager Sellers.”
14
See Amar Bhide, The Venturesome Economy: How Innovation Sustains Prosperity in a More Connected
World (Princeton University Press, 2008) and von Hippel, Democratizing Innovation.
15 Figure 1 is adapted from Robert J. Dolan, “Researching and Monitoring Consumer Markets,” HBS No. 592-
088. This sequence of awareness  trial  repeat phases is the common core of most models of buying processes
for nondurable goods. For similar models of the more complex purchase processes associated with consumer
durables (such as automobiles), see J. Hauser, G. Urban, and J. Roberts, “Prelaunch Forecasting of Automobiles,”
Management Science (April 1990): 401-420.
16The information on Initiators, Gatekeepers, Influences, Deciders, Purchasers, and Users was adapted from
Thomas V. Bonoma, “Major Sales: Who Really Does the Buying?” Harvard Business Review (May-June 1982). For
variations on this model and discussions of the academic research underlying such models, see Frederick E.
Webster, Jr., Industrial Marketing Strategy, 3rd ed. (John Wiley & Sons, 1995), chapter 2 (“Models of
Organizational Buying Behavior”), pp. 36-52.
17 This white paper is available at http://s3.amazonaws.com/SurveyMonkeyFiles/SmartSurvey.pdf.
19
2025.
18
Eric Ries, “Case Study: Using an LOI to Get Customer Feedback,” December 23, 2009, “Startup Lessons
Learned” blog, http://www.startuplessonslearned.com/2009/10/case-study-using-loi-to-get-customer.html,
accessed November 20, 2011.
19 Walter Apai, “Interview with Web Usability Guru, Jakob Nielsen,” September 28, 2009, “Webdesigner
Depot” blog, http://www.webdesignerdepot.com/2009/09/interview-with-web-usability-guru-jakob-nielsen/,

20 Thomas R. Eisenmann and Laura Winig, “Rent the Runway,” HBS No. 812-077 (Boston: Harvard Business
Publishing, 2011), p. 8.
22 J. G. Sweetland, “Beta-Tests and End-User Surveys: Are They Valid?” Database (February 1988): 27-32.
23 McClure’s presentation is available at http://www.slideshare.net/dmc500hats/startup-metrics-for-
pirates-long-version.
24 E. Ries, The Lean Startup (Crown Business, 2011), pp. 133-141.
25
See A. Kaushik’s blog “Occam’s Razor”; for example, “Excellent Analytics Tip #1: Statistical Significance,”
http://www.kaushik.net/avinash/excellent-analytics-tip1-statistical-significance/, accessed November 20, 2011.
26
Thomas R. Eisenmann and Lauren Barley, “Triangulate,” HBS No. 811-055 (Boston: Harvard Business
Publishing, 2011).
27 Marketingprofs blog, “YouTube tops Facebook, Twitter in User Satisfaction,” March 3, 2011,
http://www.marketingprofs.com/charts/2011/4549/youtube-tops-facebook-twitter-in-user-satisfaction,
28 La Rochefoucauld, Maxims, ed. Leonard Tancock (Penguin Classics, 1959).
29 Glen L. Urban and John R. Hauser, Design and Marketing of New Products (Prentice-Hall), p. 129.
30
Cindy Alvarez, “10 Things I’ve Learned,” September 8, 2011, “The Experience is the Product” blog,
www.cindyalvarez.com/roundups/10-things-ive-learned, accessed September 12, 2011.
31There is also a general guideline here for all validation techniques: Market type will affect what you need
to know. Blank (Four Steps, pp. 23-26) distinguishes four new-venture types: new higher-performance product in
existing market, new product in new market, new product attempting to resegment an existing market at lower
cost, new product attempting to resegment an existing market with a custom niche solution. Blank emphasizes
that “What kind of a startup are we?” should inform customer discovery and validation techniques.
32
This example, and much of this section about conjoint analysis, is indebted to Robert J. Dolan, “Conjoint
Analysis: A Manager’s Guide,” HBS No. 590-059 (Boston: Harvard Business Publishing).
33 Brian Orme and W. Christopher King, “Conducting Full-Profile Conjoint Analysis Over the Internet,”
1998, Sawtooth Software, Inc. white paper, November 20, 2011, http://www.sawtoothsoftware.com/download/
techpap/internet.pdf, accessed November 20, 2011.
34 For more on the history of this quote, see Patrick Vlaskovits, “Henry Ford, Innovation, and that ‘Faster
Horse’ Quote,” Harvard Business Review blog post, August 29, 2011, http://blogs.hbr.org/cs/2011/08/henry_
ford_never_said_the_fast.html, accessed November 20, 2011.
35 For more about the methodologies and assumptions embedded in conjoint analysis programs, see Urban
and Hauser, Design and Marketing of New Products, pp. 269-273; or David Bakken and Curtis L. Frazier, “Conjoint
Analysis: Understanding Consumer Decision Making,” in Rajiv Grover and Marco Vriens, eds., The Handbook of
Marketing Research: Uses, Misuses, and Future Advances (Thousand Oaks, CA: Sage Publications, Inc., 2006),
chapter 15.
20
2025.

PDF Eng

Uploaded by

Copyright:

Available Formats

PDF Eng

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PDF Eng

Uploaded by

Copyright:

Available Formats

9 -8 1 2 -0 9 7

REV: AUGUST 24, 2012

Customer Discovery and Validation for

What Do You Want to Know about Potential Customers?

Customer Problem and Product

Demand Creation and Buying Process

Figure 1 Simple Hierarchy-of-Effects Model and Typical Major Influences

Stage 0: Stage 1: Stage 2: Stage 3:

Advertising Product concept Product quality

How Can You Validate Demand?

 Formulating Specific Questions. Common mistakes include crafting confusing or leading

 Customer Selection. Customers in a market trial should be representative of early adopters.

Net Promoter Score

Exhibit 1 Summary of Entrepreneurial Demand Validation Techniques Discussed in This Note

Use setting: Early prototype stage of product and customer development.

Net Promoter Score

Exhibit 2 Sample Letter of Intent

It is the intent of CUSTOMER to enter into a formal agreement with STARTUP by

Agreed and understood by:

1) We could have titled this a Memorandum of Understanding (“MOU”) as opposed to an LOI,

Appendix A: Focus Groups

Appendix B: Conjoint Analysis

Adoption,” Harvard Business Review (June 2006).

17 This white paper is available at http://s3.amazonaws.com/SurveyMonkeyFiles/SmartSurvey.pdf.

Depot” blog, http://www.webdesignerdepot.com/2009/09/interview-with-web-usability-guru-jakob-nielsen/,

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.