Why We Need a Framework

So far, I've introduced some basic BeSci concepts and the critical importance of thinking about things from our employees’ perspective instead of our own. In this part, I will introduce HUMANS, a practical BeSci framework. HUMANS consists of six elements – one for each letter of the word – that can help us think about things from our employees’ perspective. As we have explored in earlier chapters, this is not something that comes naturally to us. HUMANS forces us to consider that alternative perspective and to remove the blinkers that dynamics like the curse of knowledge can bring.

If we've already implemented a particular behavioural intervention, HUMANS can help us understand why our employees are reacting to it in the way they are. If, on the other hand, we are planning a behavioural intervention, it can help us to think about how they are likely to react to it. In both cases, we're trying to understand what might be driving their response to what we're asking them to do.

What Is a Behavioural Intervention?

The HUMANS framework is designed for use with “behavioural interventions” (BIs), which are attempts to influence the decision‐making and, therefore, the behaviour of our employees. The purpose of a BI is to help deliver a “desired outcome”, something we want our employees to do or not do. For example, we might want them to comply with a particular rule.

The principles and ideas I will outline are designed with employees in mind. However, there is no reason you can't use a version of them to influence other stakeholders, such as regulators or customers.

BIs come in a variety of forms. In simple terms, anything that is designed to influence the decisions made by your employees is a BI. Most obviously, that's policies, rules, training, and communications campaigns. If something is intended to be seen, read, or followed by a human, it's a BI. Equally, it includes things that are partially or wholly designed to deter undesirable behaviour. So, we can see controls, monitoring programmes, incentive structures, and disciplinary processes also as forms of BI.

Dress Code

To illustrate the breadth of what can constitute a BI, let's look at a simple example.

The CEO of Big Company (“BigCo”) is deeply concerned by what they see as a “post‐lockdown decline in standards” in what employees wear to the office. Their “desired outcome” is getting their staff to dress more appropriately. To achieve this, they've identified ten BIs intended to help solve the problem.

They begin by deciding that BigCo needs a detailed dress code (BI1) and take inspiration from a 44‐page policy they remember learning about somewhere.¹ They like round numbers, so 50 pages it is! To ensure compliance, they intend to instruct BigCo's security staff to prevent anyone they see in breach of the code from entering the BigCo offices (BI2). Because standards matter at BigCo, the rules apply to anyone working from home (BI3). For that reason, the CEO will also get BigCo's technology team to use artificial intelligence to monitor adherence to the code during video calls (BI4). Anyone failing to comply with the rules will automatically lose one day's salary (BI5).

Internal comms will email all employees to ensure everyone is aware of this new approach (BI6). They'll also put up posters in the BigCo offices showing the difference between “appropriate” and “inappropriate” attire (BI7).² The BigCo CEO also intends to do a video interview for the intranet to talk about their “passion for fashion” and why “smart dress means business success” (BI8). Every employee will also be required to do three hours of a training course called “what to wear at work” (BI9). Finally, BigCo will provide each employee with a $500 voucher – the company will meet all employee tax obligations – that they can spend at a retailer of their choice to buy work clothes (BI10).

Fortunately, I've just described an entirely fictional example that is purely there to illustrate that BIs are a broad concept and the HUMANS framework, therefore, has wide applicability. It's not just for obvious “comms” things like training courses, emails, or posters. We can also consider how employees will or are likely to perceive our entire control environment, including elements like controls.

As well as applying HUMANS to the BIs, we can use it to achieve the desired outcome. In this case, we could ask the simple question of whether the very idea of the CEO imposing dress standards or requirements is something the employees would find reasonable or acceptable. The answer to that question could then determine the selection and design of individual BIs.

Of course, in this case, the example is so extreme that we don't need a framework to tell us that the approach being adopted by our fictional CEO is – I'll be polite here – “overkill”. With the obvious exception of the voucher, which seems entirely out of character. Even if employees accepted the idea of imposed dress standards, the way the CEO approaches the problem is unnecessarily aggressive. Setting a dress code requirement and policing it this way are highly likely to breed employee resentment, particularly among the Security and IT teams required to help implement the regime.

There are a couple of ways the BigCo CEO could potentially use HUMANS. The first would have stopped them from implementing some of the more extreme measures while encouraging ideas like the voucher. By using HUMANS as a design tool, they would better understand the likely success of the BIs they were proposing. That way, they could focus on the ones most likely to be effective and either drop or redesign those that are likely to be less effective or even counter‐productive.

The second way they could use HUMANS would be if they were reviewing existing BIs. Let's imagine that instead of implementing the new regime, they had inherited it from their predecessor. In that case, they'd be using it as a diagnostic tool to identify areas for improvement. Because the BIs have already been implemented, the CEO would also have behavioural data that they could use to inform the analysis offered by the HUMANS framework.

HUMANS Overview

Each element in the HUMANS framework allows us to think about different aspects of how our employees are likely to perceive what we are imposing on them. By prompting key questions, each element ensures that we consider a range of behavioural dynamics, many of which we might otherwise neglect.

Since we are trying to understand things from the perspective of our employees, the key questions are intentionally framed that way round. So, rather than asking, “How helpful will our employees find what we are asking them to do?”, the framework asks, “How helpful do I find what they are asking me to do?”

The framework works on a straightforward premise: if the answer to the key question is positive, then it is more likely – though, of course, by no means guaranteed – that our employees will comply with what we are asking them to do. Conversely, if the answer to the key question is negative, they will be less likely to comply with what we are asking them to do.

If that all sounds rather messy, remember that we're assessing human decision‐making, which isn't always straightforward. Human emotions can be extreme, as well as contradictory! If we want to overcome the flaws of traditional approaches, we need to accept that this is a feature of the way we think.

The six elements and examples of the types of questions they prompt are as follows:

• H = Helpful: “How helpful do I find what they are asking me to do?”
• U = Understand: “Do I understand what they are asking me to do?” and “Do I understand why they are asking me to do it?”
• M = Manageable: “How manageable is what they are asking me to do?”
• A = Acceptable: “How acceptable do I find what they're asking me to do?” and “Do I accept that they have the authority to do so?” The more acceptable they find what we're asking them to do, the more likely they are to do it.
• N = Normal: “Is what they are asking me to do, normal?” and “Are other people doing this?”
• S = Salient:³ “How relevant is what they are asking me to do?” and “How appealing is what they are asking me to do?”

Although the majority of questions sound simple, as we'll discover, there's more to them than we might initially think. You might also have noticed that although the entire framework is subjective, the later elements are a little more subjective than the earlier ones. For example, we will probably find it easier to identify whether our employees are likely to find something helpful than whether they find it salient.

Affordances

As well as using HUMANS to help us assess BIs, we can also use it to explore affordances. Affordances are a concept from design that describes potential alternative uses of objects.⁴ In this case, we can think of affordances as alternatives to compliance; not what we would like people to do, but what they could do.

By using HUMANS to think about affordances, we can see how the alternatives to doing what we want stack up and therefore how tempting those might be, relative to our desired outcome. This allows us to play devil's advocate and see how they are likely to perceive the options of noncompliance or partial compliance.

What the HUMANS Framework Is Not

To get the most out of HUMANS, we need to understand its limitations. Knowing what it isn't and can't do will make us more aware of what it is and can do.

The framework is not a checklist that you can slavishly follow. It is a tool that you can use to identify dynamics that might be relevant to the outcome you are looking to achieve. In some cases, all elements will be helpful. In others, one or two might be all you need.

The individual elements in the framework are neither an exhaustive nor exclusive list of relevant factors you may want to consider. On many occasions, they will counteract each other; on others, they will work in unison to drive an extreme reaction.

The framework, in common with the entire Humanized Risk approach, excludes consideration of incentives and punishment.

The framework is not designed to be a model that weights each element and produces a score at the end. That's not to say you can't or shouldn't try to score them. By all means, please do so, but remember that the analysis you will be doing is qualitative and not quantitative. Any score is therefore highly subjective.