Loops and Ladders

Giving and Getting Feedback

About This Lesson

We’ve described the process of giving and getting feedback as a form of iterative, collaborative learning that leverages complementary points of view. The effectiveness of such collaborations depends on these shared values:

  • Focus on valid information that’s relevant and actionable
  • Mutual respect for each other’s capacity to learn, and
  • Informed choice, which is based on the conviction that the giver of feedback isn’t telling the recipient what to do

Behaviors that contradict these behaviors are likely to trigger defensiveness that undermines effective learning. After all, even valid information that is delivered disrespectfully is hard to hear, much less act upon.

Sometimes we receive feedback in the form of beliefs or conclusions that can be difficult to reconcile with our point of view. In this lesson, we’ll explore a technique for decomposing such high-level feedback into the data and meanings that inform it. Having a framework for generating such revealing questions is powerful and can enhance the efficiency and effectiveness of the effort you invest in discovery and learning.

The Ladder of Inference

The Ladder of Inference developed by Chris Argyris is a tool that helps us slow down, examine, and understand our own judgments as well as those of others. Without understanding, we don’t have a choice. By broadening our understanding and realm of choice, we are more likely to create successful strategies.

Ladder of Inference

The Ladder of Inference

The lower rungs of the ladder consist of observable data and experiences—imagine that which you could capture with a video camera. Because we simply don’t have the capacity to pay attention to everything, we are constantly sifting and selecting data. As you travel up the ladder, you put selected data in context and give it a name. Finally, you translate your contextualized data and experiences into beliefs and judgments that, in turn,  guide your actions.

When we receive feedback that represents beliefs and judgments at the top rungs of the ladder, we can use the framework to prompt questions that reveal the underlying data, context, and assumptions. Sometimes, we find that we’re not selecting the same valid data. Other times, we draw upon the same data but come to different conclusions because we view it through different lenses. In any case, learning how to navigate up and down the Ladder of Inference can help you get the most out of the feedback you receive.

An Example of the Ladder in Use

For example, what if you had asked me about feedback on your new e-commerce website, and I told you, “Your e-commerce website is terrible.” Armed with this new tool, you recognize that my statement is high on the Ladder of Inference. It’s a judgment statement that doesn’t reveal much about how I reached my conclusion. Without further digging, my high-level feedback doesn’t provide much guidance about how you might go about improving your site. You may conclude that something is amiss, but it’s hard to know what exactly to do.

In this situation, you might ask, “What did you see or experience with the website that resulted in that conclusion?” In effect, you would be inviting me to climb down the ladder and share the observable data I selected and the context in which I was making my assessment. Your question may prompt me to explain, “I found it really difficult to navigate. It seems to me that almost all of the navigation on your website is based on text-based links. Furthermore, the default font is tiny. Because I do almost all of my shopping using my phone, it’s really difficult to select the option I want. The links are just too small and fidgety for me to easily select them with my fingers.” Now you know that what I saw was hyperlinked text, and I’d missed the dropdown items in your navigation menu. Furthermore, you realize that the context of my comments was related to my desire to use my phone rather than my desktop computer when shopping. “Small and fidgety” are names I gave to my observations and experiences. Taken together, my conclusion that the “website is terrible” makes more sense. It’s certainly more actionable.

The OODA Loop

The “OODA Loop” (sometimes known as the “Boyd Loop”) refers to a recurring learning cycle composed of Observe, Orient, Decide, and Act:


The Boyd Loop consists of a recurring cycle of Observation, Orientation, Deciding, and Acting.

  • Observe means sensing and selecting data.
  • Orient means drawing upon your previous experiences, cultural traditions, and other accumulations to make sense of your observations.
  • Decide means to form a hypothesis about the behavior of the system in question, and
  • Act means to test your hypothesis and, as a consequence, generate more observable data.

The upshot of creator John Boyd’s work is that going through the loop fast and frequently—even when starting with low-quality information—yields superior results. Likewise, seeking feedback early, often, and from a wide variety of perspectives will help you achieve superior results faster and less expensively. That’s an underlying reason why our friend Jake Cook’s “3 × 5 × 10 Conversations” technique is effective: it provides a practical and productive way to “fly through the OODA loop” quickly and inexpensively.

Wrapping the Ladder into a Loop

Interestingly—and encouragingly—the OODA Loop and the Ladder of Inference seem like two ways to see the same phenomenon. The lower rungs of the Ladder correspond nicely with the Observe-Orient quadrant of the loop. Similarly, the higher rungs of the ladder fit with the Decide-Act quadrant. In a way, going through the OODA Loop is akin to climbing down and up the Ladder of Inference.

Wrapping the Ladder into a Loop

The Ladder of Inference fits nicely with the OODA Loop.

Seeking feedback from diverse, complementary perspectives helps us cycle through the learning loop more frequently. By developing the capacity to climb up an down the Ladder of Inference enables us to cycle quickly and productively. The net result is more efficient learning and a higher likelihood of success.

Next Steps

In our next and final conversation, we’ll wrap up this series on giving and getting feedback. We’ll discuss how the tools and techniques we’ve presented are designed to help us slow down and think hard in a manner that builds upon our automatic and fast-acting intuition and emotions.

Key Ideas

  • The lower rungs of the Ladder of Inference are observable data and experiences that we select and name. The higher rungs represent the meanings, beliefs, and judgments that we derive from the lower rungs.
  • When we receive high-level feedback, we can ask questions to climb down the Ladder of Inference in order to understand the data upon which such feedback is based. Furthermore, we can use the framework to better appreciate how there exists data that we’d not previously selected as well as contexts in which familiar data might yield novel conclusions. In the process, we can generate deeper understanding that expands the domain of choice.
  • The Observe-Orient-Decide-Act Loop is a model of learning that corresponds nicely with the Ladder of Inference. Cycling through the OODA Loop fast and frequently—even when our starting information is poor-quality—helps yield superior results.
  • Techniques such as “Benefits and Concerns” help facilitate and enhance our traverse through the learning cycle.

Read the Video Transcript

Laura, in our previous conversation, you described the set of values that are required for productive, collaborative learning. First, you talked about our need to focus on valid information. That’s information that’s relevant to the decision making at hand. Second, it’s important that we demonstrate mutual respect, and that is a shared conviction that each of us is capable of learning. Third, we value informed choice and agency. That is, the giver of feedback isn’t telling the recipient what to do. They’re just providing a perspective that will help improve the nature of the decision-making process.

We also talked about how defensive reasoning can undermine our efforts to learn through feedback. You gave the example of when we unilaterally shelter the recipient of feedback from our honest perspective we in effect say, “You can’t handle the truth.” That’s disrespectful because it implies that we don’t think the other person has the capacity to learn from the valid information that we have.

Are there other ways in which we unwittingly undermine effective feedback and learning?

When we’re giving feedback we may want to be wary of phrases like, “well obviously” or “clearly,” simply because they may belie our respect that we’re trying to convey. Bill Noonan, my colleague and first coach in productive reasoning, put it to me this way. He said, “Laura, if you can add the word ‘stupid’ to the end of your sentence and it fits, then it doesn’t matter how valid the information is that you’re sharing. It’s not going to be heard because you’re signaling some disrespect.” So these values that underlie collaborative learning and effective feedback are really intertwined: valid information, informed choice, and mutual respect.

And it’s not just about common courtesy or being politically correct. There is there’s a fundamental reason to pay attention to these things, right?

Well, sure. We want our feedback to be valuable and useful to the person. And so if we inadvertently deliver feedback in a way that raises the defenses of the person asking us for feedback, then it’s much harder for them to hear what it is we’re saying because they’re so focused on how it is we said it. It may undermine our usefulness and credibility and, I guess, long-term social capital as somebody who can provide help.

In addition to some of the tools that we’ve already talked about in this series, are there any other frameworks that you would suggest that would help us receive and give feedback more effectively?

I have found tremendous help in Chris Argyris’ “Ladder of Inference.” The premise that Argyris put forward is that we are incredibly good, highly skilled at making inferences, and, therefore, we do it automatically. So the Ladder of Inference is a tool that helps us slow down and recognize that our inferences are not facts.

So let me just describe the Ladder of Inference, and then we’ll talk about some examples. The Ladder of Inference says that the first rung on the ladder is to select directly observable data. And here the notion is that directly observable data is what a video camera would pick up. It’s something that could be recorded, videoly or audioly. Even though we try to observe everything, we don’t. We fixate on certain things, and so we first select data.

The second rung is that in our minds we put that data into context. And context can be cultural, it can be organizational, or it can be social. We’ll unpack that a little bit more in a minute.

The third rung on the Ladder of Inference is that we name what we think we saw. So if the data I select is that somebody walked in the room at 10 minutes after 9:00, and the context I put it in is that there was an email that said we were going to start the conversation at 9:00, then the way I might name it is, “This person is late.”

Then the rungs above that are more where we get into judgment. So I might judge being late as really rude. And then with that judgment, I may act on my inference that that person is being rude and choose to dismiss something that person would say or not welcome them into the conversation. So that’s the basic premise.

The value of the Ladder of Inference is when we recognize that we’re hanging on a top rung. Just like you can imagine, when you’re hanging on to a bar far above the ground of directly observable data, it can feel really good to slow down and let your feet find lower rungs and back down the ladder, rather than simply fall onto the ground with our judgments.

I have found the Ladder of Inference useful in giving feedback. Because the way I approach it is to start with what I selected. So I say what I saw. Then I walk through my interpretation of what I’m focusing on up to the judgment or the opinion.

It’s also very helpful in eliciting feedback because if you ask someone for feedback, and they give you a very high-level inference or generalization like, “Your website is just not usable,” then you don’t have to leave the conversation there. You can use the Ladder of Inference, without necessarily explaining it, to back down the person and help them explain to you what they saw. So you can say, “Well, what were you looking at when you first concluded that the website was not usable?” That invites them to talk about a particular screen. Maybe it was the screen was loading really slowly. So you can walk them back up to that Ladder of Inference.

Now the beauty of the ladder is that it helps us recognize that there’s more data out there than the first piece of data we selected. And there are many more inferences to draw from the data that we select than the single one that we ended up with. So that’s been a real help to me.

So I imagine that by being able to navigate up and down the ladder if you will, you can test to see whether the information is valid, that the conclusions that people are making are based on the same things that you see.

That’s exactly right. Again, my colleague and coach Bill Noonan described it as dueling ladders, when people were hanging on to the top rungs of their ladders, and the conversation devolves to kind of a, is not, is too. And when you know how to back down to the lower rungs of the Ladder of Inference, it often comes out that you’re looking at different things. Or you’re looking at the same thing but putting it into different contexts. “Well, I thought this is what we were doing when we were pointing to this.” And you say, “No, no, no. That wasn’t my intention at all.” And so it helps you move to a more productive conversation. It’s productive because you explore a broader realm of what each of you can see and also what others can see that you can’t and how they interpret that. So, it’s productive because you’re learning and expanding your realm of choice.

Previously you had introduced a tool called the OODA Loop, standing for observe, orient, decide, and act—a form of learning cycle. What relationship, if any, do you see between the OODA Loop and the Ladder of Inference?

The take away from John Boyd’s OODA Loop was that cycling through the observe, orient, decide, act, cycle faster, even on worse information, and cycling through more times, leads to better outcomes. I see that as related to when we ask someone for feedback, we are asking them to take us through their OODA Loop. What did they observe? How did they make sense of it? And what would they recommend that you do?

So the parallel is that cycling through the OODA Loop more times is akin to asking a wide variety of people for feedback earlier on in our endeavors before we over-invest in a particular aspect or procedure of our business, whether we’re working with vendors or a particular website configuration. That’s one of the reasons I think that the OODA Loop is really related to the Ladder of Inference.

If you curve the ladder into a cycle, you can see that when we observe something, we’re really on the bottom rung. We’re selecting data, and orienting is moving up the rungs. We’re putting that data in context and naming it. And then when we decide what to do, we’re at the top rung. So I envision that ladder folding around the OODA Loop.

Throughout the course of our conversations, we’ve talked about a set of, what on the surface seem to be, fairly abstract concepts—things like the Johari Window, the OODA Loop and the Ladder of Inference. Interesting, but to what extent are they really practical?

The social scientist Kurt Lewin said, “There’s nothing so practical as a good theory.” More colloquially though, one of my colleagues said that “You can know a lot of things intuitively, but when you have a framework to help you understand why your intuition is right, you become much more powerful.” In the entrepreneurial context, these tools for eliciting feedback and giving feedback help us learn more efficiently. That’s faster, but also more effectively. As we learn more effectively, we increase the speed and the degrees of freedom we have for acting in ways that help us achieve success as we define it.

Lesson tags: learning