New product development 325
shows the decisions required at each stage of
the NPD process, together with an indication of
the information needed to take that decision
and its likely source.
Tzokas identifies six types of uncertainty
pervading the NPD process, namely market,
technological, competitive, resource, product
policy and organizationally-based uncertain-
ties. These can be illustrated and slotted into
the NPD process as shown in Table 12.3.
The models of NPD processes as described
above tend to be idealized and for this reason
may be quite far removed from a specific, real
instance of NPD.
A number of authors have researched to
what extent the prescriptive activities of the
NPD process are undertaken in companies.
Cooper and Kleinschmidt (1986) used a ‘skel-
eton’ of the process taken from a variety of
normative and empirically based prescriptive
processes developed by other authors, and
found that there is a greater probability of
commercial success if all of the process activ-
ities are completed. This finding is confirmed in
another study, which replicates the investiga-
tion in Australian companies (Dwyer and Mel-
lor, 1991). Recent research by Page (1993)
showed that a majority of American companies’
studies do carry out these main activity
stages.
In the Best Practice research studies on
both sides of the Atlantic, however, the better-
performing firms did tend to use cross func-
tional, third generation processes (those with
‘fuzzy and flexible’ decision points) and, in the
UK survey, the better-performing companies
also executed, on average, a higher number of
the possible steps in the process (Tzokas, 2000).
This echoes the advice ‘from experience’ of
Riek (2001), who said that ‘skipping develop-
ment steps to increase speed to market is the
road to disaster’ (p. 311).
A further study of NPD practices in the UK
and the Netherlands has shown that companies
do carry out the activities prescribed by the
BAH model, but that the emphasis of every
stage is on ascertaining the extent to which the
product under development is acceptable to the
market (Hart et al., 1998). This is not surprising,
since the process of developing a new product
is inherently risky, plagued as it is by uncer-
tainty at every stage. Over the process, the
uncertainty is reduced – be it regarding tech-
nology, makeability or potential customer
response.
The conundrum, of course, is that the
greatest level of uncertainty exists at the outset
of the process – when trying to make decisions
regarding these three elements – but until some
development of concept begins, the informa-
tion that can be collected to reduce uncertainty
is less reliable than later on, when a clearer
picture emerges of the type of product, its
features and their likely cost. And yet, despite
the warning reported above about the risks
involved in missing steps, there is still the
persistent view that missing out marketing
research steps in the process may not harm, and
even improve, the entire NPD process. The
well-known and oft-touted example of the
Sony Walkman is proffered by those claiming
that market research tools are redundant in
NPD, and process activities that are research
related simply slow the pace of development.
While it may be desirable to have a
complete process of NPD, each additional
activity extends the overall development time
and may lead to late market introduction. There
can be a price to pay for late market introduc-
tion. For example, Evans (1990) has quantified
the consequence of extending the development
time: delaying launch by 6 months can equal a
loss of 33 per cent in profits over 5 years.
Therefore, a trade-off has to be made between
completing all the suggested activities in the
NPD process and the time which these activ-
ities take.
In the case of true innovations, the role of
traditional marketing and market research may
be less useful. This does not, however, remove
the importance of marketing’s role in NPD. The
PDMA Best Practice research showed that the
best companies used significantly more market
research tools, including the above-mentioned