A lot of our prospective clients ask us if we can force comments, create alerts or tasks when there are safety violations on their checklists. We don’t offer these features, not for technical reasons, but for liability and management reasons. I’m going to use this blog to explain our thinking on this subject.
Alerts and tasks sound great on paper, the reality is that they open you up for additional unnecessary liability and work. It has a lot to do with how and when restaurants conduct checklists and the nature of our business. Let’s start from the beginning.
An alert is a way of drawing attention to an issue, but it doesn’t require that you take any action. Because it doesn’t require you to take action, it is often ignored. Also, and this is a theme throughout this topic, there is a limitation in computer logic that may create false alerts.
Let’s discuss false alerts quickly by using the example of 50-degree mayonnaise on your line. If this mayonnaise had been in a cooler all night, and you took it’s temperature, and it was 50 degrees this is a critical violation and probably a sign that your cooler is broken and everything that goes along with a broken cooler.
There is also another example where 50-degree mayonnaise is not a critical violation. Some mayonnaise is shelf stable and can sit at room temperature indefinitely until opened. At that point, it needs to be refrigerated, and you have 4 hours to get that mayonnaise down to a safe temperature. If you were out of mayo and opened a new container and stocked up your line, then this would not be a critical violation until that Mayo had been in the danger zone for 4 hours or more.
How would a computer know this? It can’t know that the Mayo was safe or not safe it can only look at the temperature and create an alert based on whether or not that temperature is in or out of range. But in this case, the alert is a false alert, it is busy work that requires a person to look at something that isn’t an issue. This is one question out of possibly 50 to 150 questions. We have several clients with 150+ item line checks. How many real vs. false alerts could be generated on a 100 question line check per shift?
Think of yourself in this situation, how many false alerts would you look at before you stopped looking? Look at your cell phone and your app badges, those little numbers that tell you that there is something in the app that requires your attention. How often to you see those and think, I need to do something about this?
In our opinion alerts are useless because: they don’t drive accountability at the user level. Also, the lack of context that the systems have and the dynamic conditions that exist in a professional kitchen make it hard to reduce false alerts.
Forcing Comments when a temperature is out of range, or a safety violation is discovered is another thing that feels like a good idea but when it is done has some potentially negative consequences. Forcing a comment is extra work for the person conducting the checklist. It is extra work that is only incurred on questions when there is something wrong.
When I type in a 42-degree temperature, I have to do this extra work but when I type in a 39-degree temperature I don’t. Have you ever heard of the Hawthorne Effect; it posits that people act differently when they know they are being observed. Have you heard of the Lazy Ass effect; where people are lazy and if they don’t understand the importance of what they are doing might be tempted to alter answers to not have to do as much work, such as lower temps by a degree or two to not have to enter a comment. Have you heard of the I Don’t Want to Get in Trouble Effect; where a person doesn’t want to be the person who answered the question that was obviously wrong so much so that the app forced me to explain what was happening?
All of these effects are real and happen. Look at how many people pencil whip their paper checklists today because they know, no one can catch them. Our concern is that by forcing comments, we are reinforcing a negative and incentivizing people to take the easy way out and not to give us accurate data. Data accuracy is of paramount importance to completing checklists, especially when they have to do with safety.
In our platform, we allow people to enter whatever temperature they recorded with a thermometer without any prompting for a comment or the creation of an alert. When they submit their checklist, the score of the checklist may be altered based on optional scoring rules but that is for each client to decide. We encourage our client’s to train their teams to enter comments explaining why a temperature was out of range, but it is not mandated. Training to enter a comment is a small but important difference between mandating and managing to this standard.
It is a lot like the reverse psychology I have to use to on my 3-year-old. If I want her to stop doing whatever she is doing that is going to cause me to spend thousands of dollars at the urgent care. I can tell her to stop, she won’t listen to me and will continue doing it or modify her behavior just enough to have me move on. This in my mind is like the mandating the comment because I’m forcing it to happen and it is a negative interaction, one that she would like to avoid.
If I go to her and say “hey, we aren’t going to watch Princess Sophia if you keep jumping on the bed.” She will stop jumping immediately because she made the decision herself and because she wanted something and she sees it as a positive interaction. That is what we want from the person completing the checklist. We want them to identify unsafe conditions an let us know what actions they took to fix those issues voluntarily and with praise from management.
There is also value to the organization in seeing which of your manager’s are following through on these types of issues. It provides insight into your managers work performance and provides opportunities for training and coaching.
If you are going to use tasks to measure your compliance and to prove that you are addressing all safety issues, then you can’t do it halfway. It’s an all or nothing proposition. It becomes a standard at which you have to manage to, 100% or nothing. Here is a scenario that could happen when using tasks.
Most line checks and temp logs are conducted right before service starts for a shift. We often see line checks being completed up to 10 minutes after a restaurant is open for business. It is a common occurrence that a restaurant could get slammed right as it opens and that the manager who just conducted the line check might not have time to complete and close all tasks before they are called away to run their shift.
You now have a situation where you identified a potential food safety issue, notified a manager, but did not address it before the food was served to customers. In reality, that manager may not have time to get back to their computer or tablet and close those tasks until the restaurant has slowed down several hours later. You know, and I know that the restaurant may have fixed that issue before service or that the food wasn’t in the danger zone or any other reason that a restaurant professional would know.
How would that look to the media or a lawyer who is trying to sue you for getting their client sick? I think that it would be used against you. Tasks work great for knowledge workers who are at their desks and computers for their entire shift and can quickly get tasks resolved and close them. Restaurant managers are in constant motion during their shift and are wrong if they are in the office during service; their job is to be managing out in the restaurant. Tasks for restaurant managers that are time sensitive could pose issues for a company from a liability perspective.
Another weakness of tasks in the restaurant industry has to do with a number of questions and locations. Let’s say you want tasks to go to your district/area managers when restaurants have a critical temp issue. If I’m an area manager with 50 locations, our area managers back at Quiznos had 50 or more locations. You conduct 4 to 5 temp logs a day; you get one temp task per temp log, and you could be looking at 250 tasks a day that needs to be addressed and closed. It isn’t uncommon to have a 1 item that is in the danger zone on a 20 or 30 question temp log or line check.
Once again you have to close these tasks if you are managing by tasks. There is no halfway; you can’t not close tasks if that is how you are tracking compliance. Managing the resolution and closing of all these tasks becomes untenable for larger organizations.
At OpsAnalitica, we replace alerting, forced comments, and tasks with summary reports. Summary reports allow our inspectors to conduct inspections quickly and then in the background we group like issues together and email them to area managers on a schedule. These reports allow the area managers to look at the issues and the comments and use their judgment on how they are following up with their restaurants without overwhelming them with communication.
Ultimately the goal of using an automated checklist app is to collect great operations data and to run safer restaurants. You don’t want to do anything that is going to take away from those goals or puts you or your organization into a situation where you were trying to do the right thing, but you increased your liability.