Consultancy Skills: The Observational Analysis
You should know how to perform a basic observational analysis for a client. The data will often highlight exactly where you focus your efforts.
Welcome to my consultancy newsletter. If you want to build a thriving consultancy practice in your niche, I want to help. Please subscribe to my newsletter and check out my best articles.
As a consultant, it’s tremendously useful to have a variety of tools at your disposal to analyse and diagnose challenges.
One of the most powerful (and least widely utilised) is observational analysis. Observational analysis, as you can guess, is when you observe what is happening and report back your findings.
This is most useful in three situations.
Checking processes are followed. It’s common for an organisation to have the right processes in place but still struggle to achieve success. In this case, you need to check if the processes are being followed correctly (and how consistently they are being followed). For example, you might observe how sales calls are being made, and how customer service is being handled. You might evaluate each incident against the processes/training and see what’s happening. This will determine if you need to change the processes or focus on training and supervision of those processes.
Looking for improvements. You might also want to determine if there is scope for improvement in how the processes are being followed. By observing what’s happening you might identify patterns and incidents which can lead to methods to improve what’s happening.
Create a benchmark for success. You can also use this process to create a benchmark for success internally. If the processes are currently being followed 82% of the time, improving that is a reasonable target to aim for. This has the additional benefit of setting up a subsequent project to track improvement. You might also be able to use this data to have an idea of industry standards and what organisations should aim for in the future.
In short, observational analysis is one of the most valuable services you can often perform for clients. It gives you data which is difficult to gather through any other channel and provides the client with an unbiased view of what’s really happening.
How To Do Observational Analysis
Observational analysis might conjure up images of you standing on the factory floor or looking over someone’s shoulder as they work. No one wants to do that. And, in my experience, you probably won’t.
It’s more likely that you’re doing one of three things:
Reviewing past behaviour. This is where you review a record of things which have happened already and take your data from that. For example, this might include viewing recordings of sales calls, a client’s past engagements with customers, emails distributed on the mailing list etc…In this process, you gather and go through an existing record of what’s happened and make your observations from that.
Testing processes yourself. This is where you test processes yourself to see how well they are being followed. Perhaps the most famous example of this is the mystery shopper. But there is no reason you can’t become a customer, go through an onboarding experience, apply to join the company, contact customer support or follow any number of steps which would replicate the experience customer or employee experience relevant to your scope of work.
Joining future engagements. This is where you join a client’s future activities to observe what’s taking place. You might attend internal meetings, training sessions, or other activities relevant to your scope of work to see whether the processes in place are being followed or not.
While in the past option 3 was most common, today I suspect options 1 and 2 are most prevalent.
If you’re going to do an observational analysis, you should undertake the following steps.
Deeply understand the context. If you’re not very familiar with the context, it’s easy to come to the wrong conclusions or report the findings incorrectly. You might want to have a couple of test examples you share with a client to understand why something has occurred and ensure you’re interpreting the incident correctly.
Be very precise in what success/failure looks like. You can’t be vague or too subjective about this (especially if you’re not the only person undertaking the data gathering). You should be very precise about what success or failure (or grading looks like). You should highlight the key things to look for or create specific questions to answer (i.e. did the waiter ask about dietary restrictions? yes/no) to make it clear whether there was a successful outcome or not.
Collect data: Gather as much data as necessary to provide a comprehensive view. This might involve multiple observation sessions over different times or situations. Make sure your data reflects the broader experience. For example, it would be silly to test customer service by making twenty calls in the same hour about the same product. You might (awkwardly) find yourself engaging with the same customer support rep each time or discover that one product has better support than others.
Recording your findings. There are plenty of templates to choose from, here’s a simple example.
In this template, you can see each incident is recorded by date, time, type, whether processes were followed or not and the explanation of the outcome. The success score can either be a binary yes/no or a grading/percentage. I’d strongly recommend also highlighting any particularly interesting stories or incidents which would be useful for presenting the results later.
Turn Observations Into Data
The ultimate goal of this is to turn observations into data. Based on this, you should be able to do three things.
Provide a score/benchmark. You should be able to tell the client that “[processes] were followed 85% of the time”. If you’ve collected a considerable amount of data, you might even be able to break it down by different categories (dates, times, type of incident etc..) to see if there is any particular area of success or improvement which you can learn from.
Provide reasons or share patterns. Whenever there is a failure, you should be able to look at the notes and highlight any common patterns or observations. The notes are a crucial part of this. You should be able to explain very clearly why things are happening and identify specific areas of improvement.
Develop the method of improvement. This might be better training, quality control, or simply improving the processes themselves. What matters most is you can use your notes to laser focus on the areas where the client can improve the quickest.
I’d also suggest a follow-up check. I’d also suggest recommending a follow-up check to measure improvement over time. Six months is usually a good time period to test the impact of improvements.
Go As Deep Into The Weeds As You Can
I can’t stress enough how valuable it is to go as deep into the weeds as you can with a client to really observe what’s happening or why. It’s dangerous to make high-level recommendations without really observing what’s happening in the ground.
An observational analysis isn’t relevant to every project for every client, but I’d certainly propose it in a large number of situations because it equips you with an understanding and a dataset from which you can make very specific recommendations.