Process Improvement: from Marionette to Machine

Does it feel like when things happen in your organization, one person seems to be pulling all the strings? In other words, does the work feel super labor intensive, and possibly like if one or a few key people were not there, that thing could not get done? Do people  ever say “Oh, we’ll just wait until (so-and-so) gets back, then they can do (whatever it is)?

I was talking to a colleague the other day who told me that she loves to put in standardized processes that will outlast her and can be done by “anyone.” As she works with various groups in her organization, she helps them define what they are doing, improve things to streamline them as much as possible, ensures it all gets documented, and all that the people doing that thing are trained. If the team responsible for the work is fewer than two people, she will even cross-train a few people from other areas so that there is redundancy in the organization for contingencies (or maybe an offsite team building session).

As she was talking, I found myself imagining a before and after visual interpretation of what she was describing. Process Improvement in this way is like going from a marionette – where there is one person literally pulling all the strings to make things happen, to a machine – which works in the same way every time if it is properly maintained.

The marionette is the one person doing everything model. It is the “if that person left this wouldn’t happen” model. The risks of one person doing things are obvious – if something happened to that person, this thing would not happen – and depending on how critical that thing is to your organization, that will dictate how problematic that is. If it is the CEO and how s/he facilitates an innovative strategic session, maybe it’s okay if something took them away because you could outsource the facilitation. But if it is the CEO and s/he is hanging on to sending every invoice, this could be problematic.

Depending on the person doing it, a second less immediately obvious  risk is one of burnout or frustration. If the person manually doing whatever it is does not enjoy repetitive, high touch tasks (thinking back to invoicing) – then assigning that person as the only one responsible increase the risk of them leaving due to frustration, burnout, or boredom with their job.

A third, less obvious risk, is that of opportunity cost. As the one person is doing xyz that takes so long and has so much manual work – what possible errors are committed that are being missed *and* what other work is *not* getting done because this one thing is so labor intensive? The opportunity cost due to time spent on rework (from errors) or just straight-up time in general – taking away from something which may be more important (like Strategic work for the CEO) or allowing the same person to take on more responsibility (manual invoicing or preparing forecasts) can cost an organization in hidden ways over time.

The final and most insidious risk associated with the marionette is the hardest to change and that is because it is in the sheer unwillingness to change by the person pulling the strings. If s/he does not *want* to let go of the manual labor-intensive process, and even worse – is unwilling to describe and document the steps s/he takes due to fear for job security, then organizations are beholden and sometimes stymied by the inability to get away from the one person doing all the work.

There is a time and place for the marionette version of a process, however. When companies are new, small, and/or in startup mode, often tasks are divided and conquered by a few people, or just one in the case of a solo-preneur. Standardizing, documenting, training, cross-training, and continuously improving processes is not even relevant until a company passes to the “viable” stage, or Crosses the Chasm, to use terminology from Geoffrey Moore. This is one reason why AJC tends to *not* work with small or startup companies, they are not yet viable enough to have a strong ROI from Process Improvement efforts, to move to the more “machine-like” version of a process.

Once an organization has crossed that chasm, however, and every company will have its own point of reckoning, Process Improvement becomes a must, and hopefully not one that is realized only after the person pulling the marionette strings leaves the company for one reason or another. Building the machine does take effort. Truly, the first step is design – as in understanding what is needed (this is mapping the current state), then designing a more streamlined future state (this is mapping the future state). Next,  determine what is required to be done to get from current (marionette) to future (machine). Those things must be prioritized and evaluated for feasibility and cost. Some things can be done easily and inexpensively, and some are more time consuming, complex, or expensive (think ERP system). However, even if some of the steps themselves remain manual, intentionally designing them and documenting those steps, while recognizing that at some point, they could be iteratively improved with an even “better” machine (like an ERP system), gives the organization confidence that it could sustain further growth over time.

Once the “reality state” is determined, this is the improved state with the feasible improvements, then it needs to be documented. Documented processes can be as simple as a bullet point list, or more complex like Standard Operating Processes with detailed roles/responsibilities and steps, and a formal signoff, or anywhere in between. We recommend determining a format that is right for the organization, and to the extent possible, being consistent – at least within impact levels (lower risk processes may have less sophisticated documentation). Of course, once all these processes are documented, they need to be managed so people know where to find them, but that is the topic of another article.

Finally, everyone must be trained, which can be as simple as a “Read and Understand” of the documented process, or as complex as instructor-led guided practice of the process. If possible, we recommend at least three (3) people be cross trained on each critical process, such that you have two levels of redundancy for vacations, illnesses, and workload capacity fluctuations. In other words, you don’t want to funnel your high-volume or critical processes through a single lane when there are five lanes leading up to it, and your most valuable customers require the output from it.

Just as machines must be maintained to remain effective over time, so too do processes. At some point, new people will arrive and need to be trained so they follow the standard processes. At other points, new technology or better ideas will come up that can streamline or improve the process further. Many companies want to grow, and as they do, they may require upgraded machines (like the ERP system) and their increased growth or new investment influx of cash finally allows them to consider purchasing something which will further improve their efficiencies.

In summary, the typical lifecycle of an organization from infancy to maturity includes phases for when the marionette style of running a process is appropriate, to when the machine style of process improvement is the right choice to take the organization to the next level. The finesse is in recognizing the shift and transitioning from marionette to machine before it’s too late.