What is business process automation?
Business process automation is not just about software. Nor is it exclusively modern. We can see it at work in the assembly lines of the industrial revolution; in the daily habits of great men (and women) of history; and across all industries and company sizes today. In my own life, I have used BPA to run a landscaping business, to sell real estate loans, to start a software company, and to turn sales teams into well-oiled machines.
BPA does not care about the color of your collar. It only cares about your system, the outlines of which arise from these three questions:
- What is your end goal?
- Which actions would lead to that goal?
- Which data dictates those actions?
BPA begins with answering these three questions and ends with the implementation of a system that you design to make meeting your goal more efficient. Sometimes, "efficient" might mean fully automated and hands-free, and sometimes it might just mean streamlined but still human. Either way, you benefit from BPA: save you time and money by eliminating repeatable tasks from your to-do list, which makes you more productive and allows you to hone in on your higher-value tasks.
Lest you still think BPA is only be possible with software, let me say this. In my experience, I have found that the largest opportunity for BPA consists in systems that combine both automated and human pieces--and in a minute, I will give you case study examples of all three types: business process automation software, non-software, and hybrid.
The Eisenhower matrix: BPA like a president
This guy knows what I am talking about.
Before the red-pill-blue-pill Matrix, there was the Eisenhower matrix, which was the time management system President Eisenhower used to decide which tasks to do when. For anyone new to thinking about how business process automation can help his/her business, I recommend applying the Eisenhower matrix to your own life--to a typical day, week, or month. This makes is easy to see which tasks you are doing, which tasks you should be doing, and which tasks you should target for automation.
"What's urgent is seldom important. What's important is seldom urgent." - President Dwight Eisenhower
Eisenhower's military and political career spanned almost five decades, from 1915 to 1961. Back then, there was no software. Yet here was a man who understood the value of BPA.
He divided all his to-dos into four quadrants. Q1 and Q2 were the highest-value uses of his time and were thus never the target for BPA. In Q3 and Q4, BPA mostly took the form of if this, then that rules given to delegates, instructing them what Eisenhower would like them to do in various instances, and in the form of simply granting autonomy to smart delegates. Then, although irrelevant to BPA (but still really important), he simply deleted many tasks that he did not esteem important--that did not, in his estimation, lead to his goal(s).
Note: had software existed back then, he might have thought differently about never targeting Q1 or Q2 for BPA; because it is not clear, today, with the reliability of well-designed BPA software, why "important" and "is my specialty" would be mutually exclusive. Indeed, if an important task leads to an important goal, and automating that task allows you to do more of it, then it would make sense to permit BPA to also target Q1 and Q2. Heck, that is how these many of these "unicorn" software companies today arrived at billion-dollar+ valuations: they applied BPA to really important tasks so they could be done more frequently and cheaper.
The Eisenhower matrix is, of course, not BPA itself. Rather it is a visual way to organize your to-dos into separate buckets of varying importance and urgency. With it, you can either apply it to your life in general, or you can get really granular and apply it to a single goal--perhaps a key business goal like, "know when to deploy more resources, and where," based on what you gather by your data-collection tasks. Focusing on that one goal, your matrix would consist only of data-collection tasks and tasks to route that data to the managers who use it to deploy resources. You might find, for example, that certain data-collection tasks are so routine and non-specialized that you really should not be doing them. Someone--or something--else whose time is cheaper...that is your ideal delegate for those tasks. In other cases, you might find that certain data-routing tasks should not be done by anyone, because today's software makes it so easy to automate just about any data-routing task.
BPA case study # 1: irrigation
We will start with an example that is dead-simple to understand.
A business owns its office building, which is flanked by a lush lawn of grass. To keep the grass lush, it needs water. Sometimes it rains, and sometimes the irrigation system needs to takeover. But the first challenge is getting the rain and the irrigation system to communicate so they do not both try to do the job simultaneously. Too much water at once could drown the lawn.
The business could install an old-school rain gauge like this one and hire a guy to write down the fill-level every day and drop off that notepad of readings to the front desk. Then they could hire another guy to monitor those daily readings at the front desk for a fill-level low enough to warrant him walking outside and turning on the lawn irrigation. Would this process work? Sure. But it could be far more efficient than this.
Alternatively, the business could install an automatic rain gauge that routes the daily reading to an automatic irrigation system. When the reading drops below a certain moisture level, the irrigation would automatically turn on a water the lawn. This second process not only eliminates two wage-laborers, but it makes the data-routing step almost instantaneous and free of potential human error.
As a former landscaper myself, I can tell you that the only stakeholder that the first option is good for is the landscaper and his kid with ivy league college aspirations. The business pays more than it should, because it failed to identify an easy target for business process automation.
BPA case study # 2: work-site safety
Now we will look at a bigger example, one that actually affects us all.
Construction sites haphazardly maintained can be hazardous to workers. Having systems in place for vigilantly monitoring these sites is not only the right thing to do as far as worker safety is concerned, it is also required by law. A construction site does not just need to be safe; the law also requires a thick paper trail of monitoring data to preemptively prove it--well, nowadays, metaphorical paper: often in the form of the spreadsheet.
“I’m slow, unproductive, and hard to analyze. Yet you still rely on me every day.” - your spreadsheet
A large power company supplies all of the electricity to residents in its region. To keep up with growing demand, it has to constantly plan and break ground on new facilities for its workers to generate new electricity for the grid. Not only are these construction sites, but they construction sites with massive, potentially dangerous pieces of equipment necessary to generate and transmit electricity. This company has an excellent track record on safety: few work-site incidents and copious, daily reports on hundreds of safety indicators.
Decades ago, when this system for data-collection and reporting began, there were only a handful of safety indicators being tracked. An Excel spreadsheet was probably the right tool back then. But overtime, as facilities grew more complex, the number of safety indicators that management felt like it needed to track daily grew. Also, as the number of faculties grew, the data management job for this one area of the business became labyrinthine, actually requiring close to a hundred weekly man-hours across the company just to get a report in upper-management's hand.
From collection to management, this is the path data would take. Site managers would oversee data-collection each day on several dozen safety indicators per site. The data would be recorded manually--meaning typed in--into an Excel spreadsheet containing several worksheets, one for each safety category. In all, twenty different workers would add data from their own site to their own spreadsheet. Then, each day, the workers would email the spreadsheets to their respective site managers. Then the site managers would email their spreadsheets to one employee at the central office, whose sole job was to receive these daily spreadsheets, copy/paste the data into a single spreadsheet, and then email the master spreadsheet to upper management.
Why do I say that this case study affects us all? Because this is how many large, safety-sensitive companies--like power companies--manage their data, and since they are spending hundreds of thousands of dollars a year on a dinosaur data management process, we, the consumers of things they produce--like power--have to pay more. Dinosaur data management does not make anyone safer. Indeed, it might make those site workers less safe because of the immense margin for human error and the significant lag in getting site safety data to the decision-makers with the authority to resolve issues before they balloon into actual danger.
Spreadsheets without business process automation are where hard-earned data (and dough) go to die.
What if all site managers entered daily data into a single data-collection app they accessed from an iPad; if that data automatically transmitted to a single database; and that data was automatically organized for reporting according to preset rules. And what if that report was submitted to upper-management within seconds of data collection? Not only is this BPA version a plausible alternative to the pony-express version described above, but it also would save that company close to a hundred man-hours per week: significant time and money saved.
James Young, co-founder of Terraine, has been building data management apps for these types of companies for two and a half decades:
"When American Electric Power hired us to improve its data management system, we were able to save them hundreds of data entry manhours per year and reduce its error rate by 95%. How? With effective software apps. Nowadays, analyzing and visualizing data with Excel is like commuting to work in a horse-drawn carriage. It will get the job done, but there are faster, cheaper ways to get there."
When it comes to managing data, far too much time is spent collecting it, and far too little is spent deriving meaning and insight from all of the data you have worked so hard to collect. The first step to world-class data management is automating what can be automated—and not just collecting data, but also sharing and analyzing it.
BPA case study # 3: home inspection
This last example belies the tedium of buying a home. Everyone involved feels that tedium, even your home inspector.
Home inspectors have an immense amount of specialized knowledge and training, which makes sense, since their job is to enter a home for the first time and, with an hour or two, divine every little defect with a property. To do that, they climb ladders, crawl under buildings, scrutinize every inch of every room, and take copious notes and pictures of everything they find. In the end, the home buyer gets a fifty-page+ report, written in a meticulous legalese, that details every defect highlighted in the pictures and notes. A twenty-year old, two thousand square foot house might require 100 pictures and--notice--an hour or two of audio notes. Yes, many inspectors record notes with a voice recorder, which is easier than carting around pen and paper and trying to write legibly while crawling through dark, tight corners or over steep roofs.
The manual process of completing a single inspection follows this sequence: drive to property, conduct inspection, later review pictures and audio notes, transcribe that info to inspector shorthand, and then transcribe that shorthand to official legalese and send the report. The average inspector spends about twenty hours per week doing transcription work, from review to report. During those twenty hours, all of that specialized knowledge and training sits on the shelf collecting dust. This is a scenario ripe for an Eisenhower matrix.
What about this system instead? Inspector collects data as before, but once he gets back in his truck to travel to his next inspection, he sends the audio and image files from that job to an offsite assistant to begin transcription. That assistant does not have or need the inspector's specialized knowledge. The assistant only needs the knowledge necessary to do the transcription. How to methodically proceed through the raw data and what format to convert it into.
Here is some math for comparison.
If the average home inspector earns $400 per inspection and takes an average of two hours to "collect data," then his time on-site is worth $200 per hour. If he spends two hours offsite transcribing and generating a report, then his overall rate drops to $100 per hour.
An audio transcription service like Scribie charges $45 per hour of transcription. If the inspector has an hour worth of audio for a two-hour inspection, and he could pay another person $55 to take the text transcription from Scribie and pair it with the photos in the proper format according to a preset of rules and templates, then his earnings jump back up to $150 per on-site hour. That is a huge time and cost savings!
Three business process automation examples
Now, I promised that I would I give you case study examples of all three types: business process automation software, non-software, and hybrid. Can you guess which is which?
The irrigation example illustrated pure software business process automation, the work-site safety example hybrid BPA of human and software, and the inspection example a pure human BPA.
What other BPA examples can you think of?
Ideas for your business or someone else's...
Want to automate your business?