Patchwork systems are more than just an annoyance.
No one sets out to create a kludge. It just… happens.
Sometimes it happens because businesses have no real plan. For example, many small businesses turn to the humble spreadsheet as a multipurpose tool, making it work for anything from customer management to tracking stock, and it’s likely to work, even for business-critical functions. But once a business grows past a certain point, it’s likely to be a burden, impeding further growth and potentially causing a great deal of pain.
But even businesses that set out with a plan are not immune to this problem. The best intentions evaporate when a business grows and new tools and services are required.
When web applications and websites are launched, there is functionality that is necessary, that which is optional, and that which is not even considered. Over time, businesses will need to add functionality that they once considered optional and perhaps that they never even considered. Sometimes, more staff are required to keep these services ticking over. For example, a business may not have Metaverse features in its website right now or even have the slightest idea what that might look like… but there is a non-negligible chance that these may be introduced in the next few years. No one starts a project with seven layers of observability and analytics—these are only added after a couple of failures.
Over time, the idea of starting again and implementing a more efficient way of doing things becomes more and more daunting. The patchwork of systems and workarounds may be nearly breaking at the seams, but it works. The inefficiencies may be bad but are nothing compared to risking something new.
But the costs of patchwork systems go beyond the obvious—and even the obvious costs are probably too much for any sensible business.
THE COSTS AND HIDDEN COSTS OF PATCHWORK SYSTEMS
Dealing with multiple vendors, technologies, platforms, and solutions can be irritating and time-consuming. But it also comes with real costs.
When developing web applications, dealing with a patchwork system means building infrastructure, in-house DevOps tools, and hiring additional employees. Inefficiencies include coordinating tasks across multiple vendors. Like an “impedance mismatch” in electrical engineering, plugging a five-volt circuit into a nine-volt power source is wildly inefficient.
Managing multiple contracts and multiple vendors means businesses cannot take advantage of consolidated bills and a single point of contact if, for example, their monitoring tools, ticketing tools, content delivery networks, etc, are all from different places. If shared responsibilities are spread around many different entities, it can be hard to nail down who is actually at fault if something goes wrong. Not having a common reference point means that something as simple as “how many people visited my website” may have many answers.
These inefficiencies are the most obvious. But there are others.
Patchwork systems can make it difficult to innovate. One might imagine that taking the absolute best of breed of every disparate element of your system, the resulting work would represent the absolute best. But the mismatch is often sand in the machine, adding cruft that needs maintenance. Businesses that are spending time maintaining their websites and applications are not developing them. There is also a cognitive-load element to this: can you become an expert in 15 different systems? And are you actually creating any value for yourself or your customers by being an expert of this very specific DevOps tool? The more interfaces there are between systems, the more chance there is of failure—and that requires planning for failure, and the likely solution is often to add more systems, further eating into opportunity time.
Security is an issue: multiple solutions require multiple upgrades, and businesses may be reluctant to move to the newest version of anything because of the potential for conflict. Of course, whether services are upgraded depends on whether all of the different parts of a patchwork system are being tracked and potential patches being flagged—parts are easily forgotten and may end up end-of-life without the business realizing. Failing to keep up with patches means being hacked—not if, but when. Businesses that run patchwork systems are, by their very nature, running systems that are vulnerable to attack and all of the costs that go along with that, such as paying ransomware demands or the time spent reverting to backups.
FIXING THE PATCHWORK
Sometimes a patchwork system is inevitable, but there are still good choices that can be made. What tools can be eliminated? Can there be tradeoffs? For example, when introducing a new, great-value SaaS tool, what can be retired in order to maintain some balance? If you can find a tool that would bring you 90% of the value you derive from the stack of tools – sometimes that is the rational choice. And if you actually miss those features, he “last 10%” can often be trivial on a simple system, whereas the cost of a patchwork tends to become exponential.
When you don’t consider integration costs and the cost of mismatch, it could look on paper that a fully-integrated solution is more costly. But this is not an apples-to-apples comparison. Just as there are hidden costs to the patchwork, there are hidden gains in fully integrated systems.
Studying the question of web application development, Analyst house Forrester interviewed multiple businesses and used an average “composite business” to make some estimates on costs for a generic web project—taking into account the global effort, including time spent as well as investment in tooling (and time around tooling)—when the environment was that of what we described above as the patchwork. With an average of fifteen different tools within the project scope. They found that the cost of implementing a new platform was substantial in time and money—two months to complete, and internal costs close to $140,000. This was before the costs of the new platform itself, at just short of $375,000.
In comparison, they found that over a three-year project lifespan, choosing a fully integrated solution could build up savings to over $2m. These are significant numbers – and evidently, the “composite business” model they chose might not look like yours. Still, these are big enough numbers to warrant a hard look. There are actually tradeoffs, and
sometimes disentangling the patchwork is not actually feasible. But when it is, the gains Forrester saw are not only in dollar amounts but also in terms of faster innovation and better flexibility.
Businesses don’t create patchwork systems through bad decisions or poor planning. It’s something that any business that scales will have to reckon with as they outgrow what they started with. But it’s important to recognize and fix the issue before it stifles growth and becomes a costly burden.
