The EVA Challenge: Implementing Value-Added Change in an ...
The EVA Challenge: Implementing Value-Added Change in an ...
The EVA Challenge: Implementing Value-Added Change in an ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
<strong>The</strong> Need for a W<strong>in</strong>n<strong>in</strong>g Strategy <strong>an</strong>d Org<strong>an</strong>ization 41<br />
of org<strong>an</strong>ization is best. Where the firm faces a more complex environment,<br />
<strong>an</strong>d that environment c<strong>an</strong> be segmented appropriately to<br />
address the complexity, the decentralized org<strong>an</strong>ization would be the<br />
form of choice. If no effective segmentation is possible, the functional<br />
org<strong>an</strong>ization becomes the default choice.<br />
In the case of dynamic org<strong>an</strong>izations, some form of lateral, or<br />
even cross-functional, relations may be required to help generate<br />
needed <strong>in</strong>formation or promote <strong>in</strong>tegration across segments or<br />
functions. Lateral relations r<strong>an</strong>ge from <strong>in</strong>formal contacts between<br />
segment m<strong>an</strong>agers or functional m<strong>an</strong>agers, to <strong>in</strong>tegrators (corporate<br />
staff or group executives), to the most complex form: a matrix org<strong>an</strong>ization,<br />
with its dual authority <strong>an</strong>d dotted-l<strong>in</strong>e relationships. Dunc<strong>an</strong><br />
cautioned that the least <strong>in</strong>vasive form of lateral relations that<br />
does the trick is the right one. Matrix org<strong>an</strong>izations c<strong>an</strong> become<br />
very difficult to m<strong>an</strong>age, <strong>in</strong> part because they have the unfortunate<br />
side effect of obscur<strong>in</strong>g accountability.<br />
For most org<strong>an</strong>izations of <strong>an</strong>y signific<strong>an</strong>t size or maturity,<br />
segmentation <strong>an</strong>d decentralization are the <strong>an</strong>swers. Truly <strong>in</strong>divisible<br />
org<strong>an</strong>izations, except small ones, are rare. Segmentation may be<br />
along geographic, process, product, <strong>in</strong>dustry, or customer l<strong>in</strong>es. How<br />
do comp<strong>an</strong>ies embrac<strong>in</strong>g the three different value discipl<strong>in</strong>es that<br />
we have discussed h<strong>an</strong>dle segmentation?<br />
Cost leaders focus on streaml<strong>in</strong><strong>in</strong>g processes to m<strong>in</strong>imize cost;<br />
thus, they tend to st<strong>an</strong>dardize operations as much as possible. If<br />
processes are not particularly <strong>in</strong>terdependent, as is the case with<br />
m<strong>an</strong>y commodities, creat<strong>in</strong>g separate bus<strong>in</strong>ess units for the several<br />
processes is likely to produce the best results. If processes are <strong>in</strong>tegral<br />
to the product, as <strong>in</strong> the production of <strong>in</strong>ternal combustion<br />
eng<strong>in</strong>es, segmentation by product is likely to deliver the highest<br />
value.<br />
Product leaders, by contrast, are product <strong>in</strong>novators. <strong>The</strong>y<br />
will generally operate better with a more loosely knit, flexible org<strong>an</strong>ization<br />
divided along product l<strong>in</strong>es, where each segment has all