Staff retention and measurements

My colleagues produce art, and I apply engineering methods to measure that art. This simple discipline has become part of our staff retention efforts.

A consultant typically has a simple set of tools to work with. They might have a basic template for a document that they will prepare for a client, or some template illustrations that can be re-used across clients. In the hands of the untrained practitioner, these templates might seem like a simple exercise of filling in the blanks on a form, or re-arranging box and line illustrations. The consultant uses these basic tools only as a guideline, alongside a series of conversations or workshops with customers. Those conversations are an inherently artistic, creative process - to use the existing tools and their individual experience to create something new for a client.

At the individual level, each project that a consultant takes on appears unique.

At the level of the professional services manager, however, trends emerge. A pharmaceutical company, a manufacturing company, and a retail company all sound different, but their organizations might want the same fundamental things. Trends emerge when the focus is on measurable inputs and outputs, which requires avoiding meaningless buzzwords, like "knowledge transfer" or "leveraging", in a Statement of Work document. For quoting purposes, inputs would be those things that are important to measure, like the number of computers the client has, or what software they want deployed.The outputs could be bespoke documents, or training, or computer programs.

Once the inputs and deliverables have been defined and controlled, these pieces of art can be commissioned on a fixed price basis, rather than having to create a new estimate for every customer.

My current simple method is to apply a Monte Carlo analysis to the completed time sheets for different types of engagement, and then to run the analysis enough times to get within an acceptable statistical error rate. There's an Excel spreadsheet involved, but that's not the point. If you have ten previous customers who wanted to deploy a piece of software on 250 computers, and ten previous amounts durations, you can come within a reasonable degree of accuracy for how long the eleventh engagement will be. That duration would be a primary basis for the fixed price.

We used to have a nine day program for sale. When we sat down and reviewed a year's worth of time sheets, however, we found that the program actually required ten days. A simple analysis using averages or medians wouldn't have shown that pattern, or that my colleagues had been working the hidden tenth day in their hotel rooms, in the evenings or late in the morning. That extra time could quickly take a toll on morale of any professional services organization - a non-specific dread of not having enough time to meet the customer's goals. The sense that they're not producing art, and that they can't afford to take risks due to a lack of hours in the contract. Imposed frequently enough on consultants, a pattern of misquoted engagements helps contribute to staff turnover, which is an awful outcome for a professional services firm.

It's possible to use engineering disciplines to reduce that risk of staff turnover. A boutique firm can get by with timesheets that record the actual time to complete work, and a simple Excel workbook that tracks the duration of projects. There also has to be trust that consultants can file accurate time sheets for both billable and non-billable hours. The discipline of periodically reviewing the data is the hardest part. Using a simple Excel Monte Carlo analysis might cost up to an hour to review the data bi-weekly, and it has to be a priority. This is a great deal when compared to the time spent on reactive staff retention or recruiting methods.

The role of consultants