Paper Technology International 2024 - Journal - Page 125
PAPERTECHNOLOGYINTERNATIONAL
to get a share of our budgets. Ambitious IT companies, if they
are not Amazon or Microsoft, or a startup, are trying to sell to
pharmaceuticals or 昀椀nancial services instead. My company, however
large, needs to grapple with the challenges of recruiting and
retaining IT staff.
Key supplier conditions
This would lead me to search for suppliers who must satisfy four
conditions:
A. Industry knowledge and associated reference sites that
are willing to vouch for the solution. This is so obvious that would
normally not require elaboration, except that I keep being astonished
by companies that forget it.
B. Local coverage: this is somewhat debatable, and it
would depend on the speci昀椀c circumstances of my own company
(if the company comprised of a single plant in Greece, it would be
unreasonable to expect local support). But equally, if the majority of
the plants were in North America, I would rule out a vendor with an
of昀椀ce only in India, however large that of昀椀ce and however good or
cost-effective the software.
C. Long-term commitment and modesty: I guess this is
best illustrated with a real example. I was peripherally involved
in what was either the biggest or the second biggest SAP
implementation project in North America in the late 1990’s at Kodak
(we were supplying the geometric optimisation). Multiple 昀氀oors in
massive buildings with cubicles populated by external and internal
consultants as far as the eye could see. Like a tennis spectator, I
observed a turf war being waged: a now-long-gone company called
i2 was pitching its scheduling system. i2’s system used a memorybased database and so could schedule many times faster than a
conventional disk-based system such as SAP’s.
Based on this technological
advantage, i2 started pressing Kodak to
forego SAP for this functionality; being
California-based and not a little arrogant, i2
went as far as to say that they would also
replace the material replenishment process
(MRP) of SAP. At which point SAP, who
was never going to cede a huge hole it its
offering, got mad, developed remarkably
rapidly its own distinct scheduling solution
(APO) and basically forced Kodak to choose.
Breadth won over the point solution and
i2 was consigned to history (along with,
irony of ironies, Kodak). The kludge that
APO was remains to this day a poorlyintegrated solution to the rest of the SAP
suite (SAP is moving to a new architecture,
also with a memory-based architecture, but
now used throughout its solution, called
S4/HANA; unfortunately, this requires a
re-implementation, being a completely new
solution…).
In researching this article, I
interviewed the recently retired head of
manufacturing at a prominent ERP supplier.
He explained to me that generic ERP
providers are ‘defocussing functional speci昀椀city’ (his words) as they
are trying to battle against the cloud hosting solution providers, who
are in ‘hostage taking mode’; data, in their view, is more important
than the application that generates it. After all, the argument goes, if
you have the data, you can throw all sorts of data mining techniques
at it.
Part of the evaluation in this regard would be to gauge the
company’s appetite for embracing modern technology that will have
an impact on all our lives. For example, consider the generative AI
dialogue produced by Tietoevry’s recently-announced TrimChat, as
shown in the box. I am impressed by the sophisticated statistical
reasoning and the provision of a trend chart when that was not
explicitly requested.
The long-term commitment of a company to a speci昀椀c
market is extremely hard to judge. With the mountains of
cash available to private equity and to such conglomerates as
Constellation Software (which owns 800+ software companies,
including at least 5 in our sector), very few IT companies can be
certain of who their owner will be in 5 years. The problem is that
very often the playbook of these purely-昀椀nancially-driven companies
is simultaneously to increase fees and stop investing in the future
evolution of the software.
D. Reasonable cost: I would sit down with my Board and
explain to them that spending up to 2% of turnover pa on IT would
be reasonable. Half of that would go to the peripheral functionality
and infrastructure. The remaining half would be split into two parts:
roughly one third set aside for the implementation / small speci昀椀c
projects and the remaining would go towards the subscription of the
core system. Once agreement is obtained, I would use the resulting
budget to 昀椀lter out the candidate suppliers.
125
