Chapter 1: The Path to Disaster
flat, and the balloons are deflated. Sales now has to find the quantity of customers that the company
claimed it could find when it first wrote its business plan. Sure, Sales may have found a couple of
“beta” customers, but were they representative of a scalable mainstream market? (A mainstream
market is where the majority of people in any market segment reside. They tend to be risk-averse,
pragmatic purchasers.) Time after time, only after first customer ship do startups discover that
their early customers don’t scale into a mainstream market, or that the product doesn’t solve a high
value problem, or that the cost of distribution is too high. While that’s bad enough, these startups
are now burdened with an expensive, scaled-up sales organization that is getting frustrated trying to
execute a losing sales strategy and a marketing organization desperately trying to create demand
without a true understanding of customers’ needs. And as Marketing and Sales flail around in
search of a sustainable market the company is burning through its most precious asset—cash.
At Webvan, the dot-com mania may have intensified their inexorable drive to first customer ship,
but its single-minded focus was typical of most startups. At first customer ship, Webvan had close to
400 employees. It hired over 500 more during the next six months. By May 1999 the company opened
its first $40 million distribution center, built and scaled for a customer base it could only guess at,
and had committed to 15 other distribution centers of the same size. Why? Because the Webvan
business plan said that was the goal—regardless of whether the customer results agreed.
3. An Emphasis on Execution Instead of Learning and Discovery
In startups the emphasis is on “get it done, and get it done fast.” So it’s natural that heads of Sales
and Marketing believe they are hired for what they know, not what they can learn. They assume
their prior experience is relevant in this new venture. Therefore they need to put that knowledge to
work and execute the sales and marketing programs that have worked for them before.
This is usually a faulty assumption. Before we can sell a product, we have to ask and answer
some very basic questions: What are the problems that our product solves? Do customers perceive
these problems that as important or “must have?” If we’re selling to businesses, who in a company
has a problem that our product could solve? If we are selling to consumers how do we reach them?
How big is this problem? Who do we make the first sales call on? Who else has to approve the
purchase? How many customers do we need to be profitable? What’s the average order size?
Most entrepreneurs will tell you “I know all the answers already. Why do I have to go do it
again.” It’s human nature that what you think you know is not always what you know. A little
humility go far. Your past experience may not be relevant for your new company. If you really do
know the answers to the customer questions, the Customer Development process will go quickly and
it will reaffirm your understanding.
A company needs to answer these questions before it can successfully ramp up sales and sell. For
startups in a new market, these are not merely execution activities; they are learning and discovery
activities that are critical to the company’s success or failure.
Why is this distinction important? Take another look at the product development diagram.
Notice it has a nice linear flow from left to right. Product development, whether it is intended for
large companies or consumers, is a step-by-step, execution-oriented process. Each step happens in a
logical progression that can be PERT charted, (a project management technique for determining how
much time a project needs before it is completed,) with milestones and resources assigned to
completing each step.
Yet anyone who has ever taken a new product out to a set of potential customers can tell you
that a good day in front of customers is two steps forward and one step back. In fact, the best way to
represent what happens outside the building is more like a series of recursive circles—recursive to
represent the iterative nature of what actually happens in a learning and discovery environment.
Information and data are gathered about customers and markets incrementally, one step at a time.
Yet sometimes those steps take you in the wrong direction or down a blind alley. You find yourself
calling on the wrong customers, not understanding why people will buy, not understanding what
product features are important. The ability to learn from those missteps is what distinguishes a
successful startup from those whose names are forgotten among the vanished.
Like all startups focused on executing to plan, Webvan hired a vice president of merchandising, a
vice president of marketing and a vice president of product management—three groups that were
The Four Steps to the Epiphany
oriented around executing a sales strategy, not learning and discovering customer needs. Sixty days
after first customer ship these three groups employed over fifty people.
4. The Lack of Meaningful Milestones for Sales, Marketing and Business
The one great thing you can say about the product development methodology is that it provides an
unambiguous structure with clearly defined milestones. The meaning of alpha test, beta test, and
first customer ship are pretty obvious to most engineers. If the product fails to work, you stop and fix
it. In stark contrast, sales and marketing activities before first customer ship are adhoc, fuzzy, and
absent measurable, concrete objectives. They lack any way to stop and fix what’s broken (or even to
know if it is broken, or how to stop at all).
What kind of objectives would a startup want or need? That’s the key question. Most sales
executives and marketers tend to focus on execution activities because at least these are measurable.
For example, in sales, the number one thing that matters is revenue. Sales uses revenue as its
marker of progress in understanding customers. Some startup sales execs also believe hiring the core
sales team is a key objective. Others focus on acquiring early “lighthouse” customers (prominent
customers who will attract others.) Marketers believe creating corporate presentation, data sheets,
and collateral are objectives. Some think that hiring a PR agency, starting the buzz and getting on
the cover of magazines at launch are objectives.
In reality none of these are true objectives. Simply put, a startup should focus on reaching a deep
understanding of customers and their problems, discovering a repeatable road map of how they buy,
and building a financial model that results in profitability.
The appropriate milestones that measure a startup’s progress answers these questions: How well
do we understand what problems customers have? How much will they pay to solve those problems?
Do our product features solve these problems? Do we understand our customers’ business? Do we
understand the hierarchy of customer needs? Have we found visionary customers, ones who will buy
our product early? Is our product a must-have for these customers? Do we understand the sales road
map well enough to consistently sell the product? Do we understand what we need to be profitable?
Are the sales and business plans realistic, scalable, and achievable? What do we do if our model
turns out to be wrong?
Webvan had no milestones that said stop and evaluate the results (2,000 orders per day versus
8,000 forecasted) of its product launch. Before any meaningful customer feedback was in hand, and
only a month after the product started shipping, Webvan signed a one billion dollar deal (yes,
$1,000,000,000) with Bechtel. The company committed to the construction of up to 26 additional
distribution centers over the next three years.
Webvan had leaped right over learning and discovery in its rush to execution. There is a big
difference between a process that emphasizes getting answers to the fundamental questions that I’ve
listed above and a process that uses the product development model to keep early sales and
marketing activities in sync with first customer ship. To see what I mean, consider the product
development diagram from the perspective of people in sales and marketing.
5. The Use of a Product Development Methodology to Measure Sales
Using the product development diagram for Customer Development activities is like using a clock to
tell the temperature. They both measure something, but not the thing you wanted.
Figure 1.2 shows what the product development diagram looks like from a sales perspective. A
VP of Sales looks at the diagram and says, “Hmm, if beta test is on this date, I’d better get a small
sales team in place before that date to acquire my first ‘early customers.’ And if first customer ship is
on this date over here, then I need to hire and staff a sales organization by then.” Why? “Well,
because the revenue plan we promised the investors shows us generating customer revenue from the
day of first customer ship.”
Chapter 1: The Path to Disaster
Figure 1.2 The View from the Sales Organization
I hope this thinking already sounds inane to you. The plan calls for selling in volume the day
Engineering is finished building the product. What plan says that? Why, the business plan, which
uses the product development model to set milestones. The consequence is that selling isn’t
predicated on discovering the right market or whether any customers will shell out cash for your
product. Instead you use product development to time your readiness to sell. This “ready or not, here
we come” attitude means that you won’t know if the sales strategy and plan actually work until after
first customer ship. What’s the consequence if your stab at a sales strategy is wrong? You’ve built a
sales organization that’s burning cash, cash that needs to be redirected in a hurry. No wonder the
half-life of a startup VP of Sales is about nine months post first customer ship. “Build and they will
come,” is not a strategy, it’s a prayer.
Webvan had this problem in spades. After first customer ship, Webvan had a nasty surprise
waiting for it. Customers refused to behave the way the Webvan business plan said they would. Six
months after Webvan’s June 1999 launch, the average daily volume of orders was 2,500 orders per
day. Sounds pretty good? Not bad for a startup? It was. Unfortunately, the Webvan business plan
had forecast 8,000 orders per day, a number that was necessary for the company to achieve
profitability. This meant that its distribution center (designed to process product volumes equivalent
to approximately 18 supermarkets) was operating at less than 30% of capacity. Oops.
6. The Use of a Product Development Methodology to Measure Marketing
The head of Marketing looks at the same product development diagram and sees something quite
different (see Figure 1.3). For Marketing, first customer ship means feeding the sales pipeline with a
constant stream of customer prospects. To create this demand at first customer ship, marketing
activities start early in the product development process. While the product is being engineered,
Marketing starts creating corporate presentations and sales materials. Implicit in these materials is
the “positioning” of the company and product. Looking ahead to the product launch, the marketing
group hires a public relations agency to refine the positioning and to begin generating early “buzz”
about the company. The PR agency helps the company understand and influence key industry
analysts, luminaries, and references. All this leads up to a flurry of press events and interviews, all
geared to the product launch date. (During the Internet bubble, one more function of the marketing
department was to “buy” customer loyalty with enormous advertising and promotion spending to
create a brand.)
Figure 1.3 The View from the Marketing Organization
At first glance this process may look quite reasonable, except for one small item: all this
marketing activity occurs before customers start buying—that is, before Sales has had a chance to
actually test the positioning, marketing strategy, or demand-creation activities in front of real
customers. In fact, all the marketing plans are made in a virtual vacuum of real customer feedback
- Create Marcom
- Hire PR Agency
- Early Buzz
- Create Demand
- Launch Event
The Four Steps to the Epiphany
and information. Of course, smart marketers have some early interaction with customers before the
product ships, but if they do, it’s on their own initiative, not as part of a well-defined process. Most
first-time marketers spend a large part of their time behind their desks inside their building. This is
somewhat amazing, since in a startup no facts exist inside the building, only opinions. Yet even if we
get the marketing people to get out from behind their desks into the field, the deck is still stacked
against their success. Look at the product development diagram. When does Marketing find out
whether the positioning, buzz, and demand creation activities actually work? After first customer
ship. The inexorable march to this date has no iterative loop that says, “If our assumptions are
wrong, maybe we need to try something different.”
This “marketing death march” happened at Webvan. In its first six months of business, Webvan
acquired an impressive 47,000 new customers. However, in those six months 71% of the 2,000 orders
per day that were coming in were from customers who had already used the service. This meant
Webvan needed more new customers, and it needed to reduce the number of customers who ordered
once and then never used the service again.
These facts contradicted the marketing assumptions in the original business plan. As happens in
most startups, those assumptions were wrong. Yet Webvan had scaled its spending (particularly on
building and operating large distribution centers) on these unverified guesses.
7. Premature Scaling
Having Sales and Marketing believe that by first customer ship, come hell or high water, they need
fully staffed organizations leads to another disaster: premature scaling.
Startup executives have three documents to guide their hiring and staffing; a business plan, a
product development model and a revenue forecast. All of these are execution documents – they
document spending and hiring as if success is assured. As mentioned earlier there are no milestones
that say “stop or slow down hiring until you understand customers.” Even the most experienced
executives succumb to the inexorable pressure to hire and staff to “plan” regardless of early customer
In Webvan’s case premature scaling was an integral part of the company culture and the
prevailing venture capital “get big fast” mantra. Webvan spent $18 million to develop proprietary
software and $40 million to set up its first automated warehouse before it had shipped a single item.
Premature scaling had dire consequences since Webvan’s spending was on a scale that ensures it
will be taught in business school case studies for years to come.
As customer behavior continued to differ from the predictions in Webvan’s business plan, the
company slowly realized that it had overbuilt and over-designed. The business model made sense
only at the high volumes predicted on the spreadsheet. The average daily volume of orders was
significantly below the capacity the company needed to achieve profitability. To have any hope of
achieving favorable gross margins, Webvan had to find a way to substantially increase its volume,
the number of customers, the number of orders placed by its customers, and the average order size.
8. Death Spiral: The Cost of Getting Product Launch Wrong
Premature scaling is the immediate cause of the Death Spiral. Premature scaling causes the
burn rate to accelerate. Sales, salaries, facilities, infrastructure costs, recruiting fees, and travel
expenses start cutting into the company’s cash flow. The pressure for revenue grows exponentially.
Meanwhile the marketing department is spending large sums on creating demand for the sales
organization. It is also spending “credibility capital” on positioning and explaining the company to
the press, analysts, and customers.
By the time of first customer ship, if the company does not understand its market and customers,
the consequences unfold in a startup ritual, almost like a Japanese Noh play. What happens when
you fully staff sales and marketing and you haven’t nailed who your customers are and why they
should buy your product? Sales starts missing its numbers. The board gets concerned. The VP of
Sales comes to a board meeting, still optimistic, and provides a set of reasonable explanations. The
board raises a collective eyebrow. The VP goes back to the field and exhorts the troops to work
Documents you may be interested
Documents you may be interested