In 1976, the legendary American country music singer and songwriter Johnny Cash released his hit, “One Piece at a Time.” In the song, a Detroit autoworker builds a car in his garage over 25 years by taking home a different car part every few days. When he finishes, the title and license plate clerk asks for the model year and he answers:
It's a '49, '50, '51, '52, '53, '54, '55, '56, '57, '58 '59, '60, '61, '62, '63, '64, '65, '66, '67, '68, '69, '70 automobile.
By the time the clerk types it all out on paper, the car title weighs 60 pounds.
Every time I hear the song on my smartphone’s road trip rotation, I think about how often companies, especially small and medium businesses, start building their information systems with great intentions. They begin with one product and then sometime later, they add a second component from elsewhere. Then they bolt on a third, and so on, until a few years later they have a collection of parts—in this case, different systems from different vendors—not ever fully intended or designed to work together.
Jonny Cash’s song should be the global anthem of system architects and consultants who spend their careers daisy-chaining disparate software solutions into large and expensive systems that are cumbersome to maintain and operate.
When SMBs embark on technology projects to support their growth and innovation—whether it involves ERP, HCM, CRM, SCM, PLM, EPM, or any of the other TLAs that define enterprise software—the end objective of having a connected solution is often ignored or forgotten. Instead, there is often a blind thrust into a single project that is overly focused on one initial problem or business need.
Consider a typical small or midsize organization undergoing tremendous growth. One day, the company realizes it needs a better way to manage all the HR tasks associated with the people they are rapidly hiring. The business focuses on an HR solution without considering the rest of the company. Eventually a human capital management (HCM) package is selected and implemented. The software is in the cloud, so the company’s employees quickly experience the benefits of not using last-century, on-premises software that is always out of date.
Once again, a specialized functional requirement ends up dominating the conversation. A second vendor wins the ERP business. Good news: ERP and HCM are now in the cloud. Bad news: the two systems don’t play well together; user experiences differ; the same data is stored differently; and information is not synchronized in real time.
So, a consulting firm is retained to get the ERP software to share finance data and other information with the new HCM system. And while both systems advertise open connectivity to synchronize data exchange, a big problem is identified: the data schemas and data dictionary are very different. (Sounds exhausting, right? I’m exhausted just typing it out. '49, '50, '51, '52…)
Soon, more consultants are engaged and a new software solution is identified to transform and align the data. It’s starting to get expensive; resources are limited, and a critical third project, supply chain management, is postponed— even though it is desperately needed to replace a primitive, out-of-date system. (Back to spreadsheets to fix that problem.)
SMBs usually have cash flow constraints and can only support a handful of IT projects at a time. Money for growth is diverted. Panic sets in and the company now has to postpone growth plans while the IT infrastructure is sorted out. With stagnation, the company’s future is threatened and its competitive, first-to-market leadership position is at risk. The optics look terrible and investors start questioning the leaders’ capability to make important business decisions.
This scenario might sound dire and alarmist, but I have seen it happen more than once.
The problem of disconnected clouds highlights the importance of data connectivity. Cloud data exchange by itself is not difficult, but when done in large quantities, it gets expensive. A bigger problem involves data structures and definitions. If one system calls one type of data using one nomenclature, and then the same data in the second system is stored differently, how valuable is all the enterprise information when it remains stuck inside cloud silos?
Consider something as simple as a machine bolt. Some companies, like automotive manufacturers and appliance makers, use millions annually. In a previous career, I was consulting at a supplier to an American manufacturer of automotive parts that used large quantities of bolts. Between their various systems they had managed to create 16 spellings of the word “bolt”.
Over 30 percent of bolt purchases recorded were duplicates of another record. The company could not determine their actual consumption and inventory quantities. Their systems could not figure out that BLT M10 was the same as BOLT M10, BOL M10, or B M10. ('53, '54, '55, '56…) The lost procurement savings were staggering.
This example shows why a suite of software—built from a common schema with a central data dictionary structure—is the best starting point for SMBs who want a system for all their enterprise data. Rather than stitching together systems designed by different vendors (including vendors who compete with each other and thus don’t jointly coordinate data structures), it’s much more efficient and cost-effective to utilize a suite from a single vendor.
This approach skips over software integration and data sharing issues, because suites are designed to work together from a common, shared data infrastructure. This not only saves on deployment costs but vastly improves data quality. High-quality data generates more accurate reporting and confidence in the information, while bypassing the need for many additional expensive consultants.
When Johnny Cash built his car, he did get it to work— eventually. But the challenges were intense, including when he had to connect the transmission to the engine:
The transmission was a '53
And the motor turned out to be a '73
And when we tried to put in the bolts all the holes were gone
I wonder how many different ways he spelled “bolt”…
Like Johnny’s car, IT systems often don’t connect as advertised during the project design phase, despite how good it all looked on the white board.
So the next time you think about picking HR software from vendor A, ERP software vendor B, CRM from vendor C and so on, consider the challenges in Cash’s song. Ponder how your business can get a true enterprise system by selecting a vendor that delivers a complete solution with uniform data structures in a common schema. Avoid the a-la-cart method, which is fun for a potluck party but has no place in your enterprise systems.
SaaS should not just represent Software as a Service. It should also be Software as a Suite. Because the best answer, when asked by your investors to describe your IT, is to point to one connected solution from one vendor that utilizes one data schema with one data dictionary that can be described on one slide or a short elevator ride.
It sounds great in a hit song, but a Johnny Cash answer for his “psychobilly Cadillac” will not go over well in the boardroom—unless you know how to play guitar and sing the answer. '57, '58 '59,'60…
For a closer look at the advantages of a cloud suite, download this ROI report from Nucleus Research.