Lately I’ve been having really interesting conversations regarding the differences and similarities between personal automation and citizen development. Citizen development has been best represented as non-IT individuals developing business solutions leveraging low-code and no-code development tools. This is essentially letting the subject matter experts in a particular domain encode their knowledge so that it can be used to help reduce the use of time for non-productive tasks or to reduce latency in an overall process.
Personal automation is not a term that I have seen defined but references to the term bring up articles about automating tasks on Apple devices. I believe it is a genre of automation that is going to be the next killer application. Personal automation will allow anyone to rapidly automate mundane digital tasks across a variety of device platforms. We’ve already seen the start of this genre with Amazon’s Alexa and other home automation technology. Individuals have developed programs to control their lighting, temperature, viewing choices, etc.
The next wave of personal automation will occur in accordance with personal collaboration tools. We’ve seen the ability to automate with Slack, Microsoft Teams, Airtable and other business applications, but these tools are designed for collaboration in a business setting. Expect that automation will be coming for WhatsApp, Facebook Messenger, iMessage and other personal collaboration platforms in the not to distant future. These automations will allow people to handle the most mundane and annoying task; finding out if you’re available at a particular time.
At work, we have business calendars that we can share internally and Calendly and other similar applications to expose availability of those calendars externally. But what tools does the working parent or the busy student have when a request comes in Messenger, “can you help me out this Wednesday afternoon at 2?” If you’re like me, you have to check the work calendar and the home calendar to make sure there’s no conflicts before you answer. Personal automation would help alleviate this problem.
So, how does this compare to Citizen Development then you ask? Citizen developers’ tasks are focused on business efficiency and productivity. Sometimes, those outcomes overlap or are aligned with their own needs, but often, Citizen development is biased towards business outcomes and business communities.
For example, many people develop complex Excel spreadsheets with macros to simplify a business task they have been assigned. In this scenario, the Citizen developer is engaged in personal automation that also has business value; especially if that spreadsheet is then shared among others in the business to simplify the tasks for all. If, however, Julie’s manager comes to her and asks her to develop a bot—a piece of software that emulates a human—that automates the work she does on invoice reconciliation, then she is acting on behalf of the business alone.
Here’s why this matters. I believe those that become proficient at personal automation will help lead Citizen development at work. Moreover, these individuals will most likely desire to use a common set of tools for both types of automation, which means that they will have significant influence over the tools selection for automation in the workplace. Winning the hearts and minds of those that engage in personal automation may be the gateway to becoming the platform selected for Citizen development in the workplace. And, with a growing emphasis on remote workers, personal automation may very well be the most critical tool in enabling a satisfactory home/work life balance.
4 thoughts on “Personal Automation Versus Citizen Development”
Interesting, with using excel and macros example. what are your thoughts about moving away from macros and using add-ins from AA where the macro logic has been lifted into the corporate enterprise to ensure various company rules, compliance are met and controlled. Of course there is a speed penalty for the add-in that is running the bot to start and return the results.
From a business standpoint, what would you consider as a separator to divide when an automation should be citizen developer or not.
Macros have their place. If the work is isolated to Excel, macros are great for getting work done inside that environment. If data needs to be loaded into the spreadsheet for the macros to work on, that’s where I think RPA is very helpful. Its going to be far more accurate and faster than manual entry.
As for a demarcation for citizen development, I can’t say that is something easily defined. Every citizen developer will have their own level of skill and ability to codify their knowledge. I think all bots can start with the subject matter expert on the tasks they do. They may only be able to capture and record their current method of working requiring a more skilled individual to convert to a bot. Or they may be able to go from start to finish. Moreover, it’s the scope of the use of the bot. If it’s only for their purposes (e.g. personal automation) then I’m favorable to enable them, help them and let them run. If its for a shared community, I would say that it should go through some validation.
absolutely agree. What are your thoughts to allow a Citizen Development eco-system to grow where the CoE can monitor? Specifically, can a product like AA Discovery Bot allow the collection of information within the AA Control Room to allow the CoE to at least be aware any automation opportunity may exists that could mature into a bot controlled by CoE instead of running locally on the Citizen Developers machine.
I think if businesses want to enable citizen developers, there needs to be a good safety net. Too much centralized control may have the resulting impact of turning off individuals from leveraging automation. Think needing to get approval from a board before changing the color of your door on your house. If the CZDev is just automating a single process only germane to them, then I recommend that certain guiderails be put in place to limit exposure, such as forcing passwords to be used from a vault and not directly in bot code. If it’s something that is designed to be shared, then the user should record the bot working and submit it with a request to move into the public domain.