The case against robots
Posted: July 8th, 2010 | Author: Bully | Filed under: Resource Management, Workflow and Organization | Comments OffDisclaimer: @jose602 is a robot and he’s okay.
There is something to be said for the production processes forwarded by the industrial revolution. There is also something to be said about knowing where those processes end and individual thought becomes a facilitator of speed.
Corporate America is full of lower level jobs where the individual is expected to perform specialized repetitive tasks for the sake of efficiency. When one of these processes backs up, the solution is to automate or throw people at it. However, there are many areas where we have simply decided that ’specialized’ and ‘repetitive’ means efficient.
Example:
I was recently asked to review a client’s process for handling what is essentially an escalated customer service response mechanism; in order to assist in developing interventions to mitigate a huge backlog. The process had regulatory requirements for timeliness that were not being met and the regulator was coming down on the client for sending out formal requests for extension to the timeframes en masse. In this process, they had broken down the necessary activity into specialized production tasks:
- Sorting – Sorting is a simple repetitive task, so it was given to a lower level staff member.
- Intake – Intake is a data entry and scanning job; also given to a lower level staff member.
- Assignment – Assignment – or distribution of workload – was given to management based on trending and supervisory expertise; but was not considered a time consuming task.
- Research and Response – Research and Response was given to skilled workers hired for writing and critical thinking skills. (this was described as a combined task performed by one tier)
My initial observation was that this structure makes theoretical sense. It was only upon observation and review of the work product that I was able to determine the problem.
The back up was clearly in the tier related to Research and Response, as expected. This is the most time consuming part of the process. The client was aware of this and began adding people to the tier expecting that less volume per person would increase efficiency. There are two problems with this solution:
- Efficiency is an equation = work output/work input x 100. The solution increases neither value because you are simply adding more people with the same capacity for input and output. And you’re going to bleed money. Executives don’t look at the traditional energy equation when they think efficiency; they look at “how much am I getting for my money?”. By the economic measure of efficiency, we change people to dollars: $50 gets you 50 files; $100 gets you 100 files. Do you feel more efficient now?
- Secondly, you will have to perpetuate the model unless the increased volume is a temporary condition. What happens with periodic spikes? You end up workload managing instead of improving capacity of the existing resources.
This solution is temporary and expensive. Increasing the efficiency of the current staffing is a much more desirable outcome.
Further observation of the work functions and product were extremely revealing. The Research function is meant to determine whether rules were followed by both parties and whether subjective professional judgment was appropriately applied. The R&R team was tasked with reviewing the incoming inquiries; determining the nature of the inquiry; and distributing their caseload to subject matter experts for responses that would be placed in a letter to the inquirer. The job description function of this team was to draft the response letter and ensure that it contained a restatement of the issue, a description of the research that was performed, and provide an expert determination of validity.
This is how they did it. They reviewed the inquiry and typed the issue (generally verbatim) into a database. They copied this database field into an email which was sent to a subject matter expert. The subject matter expert would work through a queue of questions and respond based on their knowledge of the rules or circumstance by placing a comment directly into the database record. The team member would then pull up a form letter template that contained paragraphs of introduction and a statement that the inquiry was reviewed by a subject matter expert. Into this template, they would place the issue description and response paragraph from the database.
You see the problem?
The Research and Response team was not performing either task that they were ostensibly assigned to; and their job functions were intellectually far below their employment level.
The responses from other departments were never read by the team member. Rather than becoming a subject matter expert in their own right; they simply cut/pasted information into the form letter and sent it out. Therefore, anytime that a similar case came in, the time consuming step of sending the information to another party for review and awaiting response was needlessly taken.
The team was filled with robots and the robots were slowing down the process because they had been instructed not to think beyond the level of identifying – based on keywords – who else should research and respond.
My instruction to the client was this: Stop automating a process meant for humans and you will see an increase in efficiency.
The client needed to remove the instruction to simply populate templates with other people’s content and allow for the team members to read and absorb the inquiry and it’s associated expert response in order to train it’s employees up. Let the people that you pay to write well thought out responses write well thought out responses; and learn the business rules at the same time. Subject matter experts also often speak in jargon; so a cut/paste response probably wasn’t serving the company very well in the long run either.
There are still many instances where the Research and Response team needs to have a subjective decision reviewed by an expert prior to formulating the response to the customer; but this intervention removed the need for many of the time delays built into the “efficient” process and the need for additional expenditure for temporary assistance. The R&R team members are now able to draft many more responses autonomously and only pull out complex cases for the review process.
Faster production. Same amount of money. Workers empowered. Client happy.
I am not a robot,
Bully