Decision making: choice overload

In previous article we covered context in a startup, including common challenges a founding engineer may face. Today I would like to start a discussion on the decision making, starting with the choice overload. I think it is interesting topic, as it helps to understand the impact the decision making has on individuals and teams. The article is based on the research review paper; “Choice overload: A conceptual review and meta-analysis” by Alexander Chernev a,⁎, Ulf Böckenholt a , Joseph Goodman [1]

“The term choice overload— also referred to as over-choice — is typically used in reference to a scenario in which the complexity of the decision problem faced by an individual exceeds the individual's cognitive resources”[1]. That is a savage definition! Not every day en engineer, or a founding member faces the fact that something can exceed his or her “cognitive resources”. Well, life is not easy, there are always challenges, tasks, decisions to make and some of them are extremely hard! Especially in a startup environment, as the whole idea of building a company from scratch, finding customers and scaling sounds like a trip to Mordor (except no one knows where the ring is). It is a hard pill to swallow, but also worth mentioning that everyone faces this problem.

Consequences

There are consequences to any decision and they can take an indirect form. The theoretical model proposed by the scientists [1] defines consequences of choice overload grouped in 2 dimensions. The first one is “subjective state”, meaning how a subject feels about the decision, including: decision confidence, satisfaction and regret. The second one is “behavioral”, meaning impact of the over-choice on a subject’s behavior. This includes choice deferral, likelihood of reversing a choice, preference for a larger option pool, and the nature of the chosen option. “Compared to individuals not experiencing choice overload, those experiencing overload are (1) less likely to make a choice from a particular assortment, (2) more likely to reverse their initial choice, (3) less likely to display a preference for larger assortments, and (4) more likely to choose an option that can be easily justified”[1].

The subjective state influences a decision maker more than a team. In the presence of regret, or lack of satisfaction, one can retrace the context, to better understand what contributes to the feeling (Covered more in the factors section).  

Behavioral consequences on the other hand may yield a significant impact on a team and an organization. Let’s try to make it more visible with some examples; when a product owner defers the work prioritization, an engineer might decide to work on improving architecture, or just wait for orders instead of working on a business critical feature. Due to a simple lack of decision.

The second example would be a need to build features A and B, both requiring 4 weeks of work. In the first sprint manager decides that A is the most important, and by the end of the sprint 1/4 of feature A is finished. For the second sprint priority shifts to the feature B. Team is again able to deliver 1/4 of the feature B. Ignoring annoyance and cost of context switching, if priorities keep switching, it takes 8 weeks to finish both features. Without the switches feature A could be delivered to a customer by 4th sprint and second by 8th. Two comments here. First of all I am not saying that no one should ever deviate from any plan, I just try to highlight the impact (or cost) of the change. Secondly this is just an oversimplified example that I used to highlight the problem. However it is extracted from my personal experiences.

The third example is a task to build a report based on an existing database, to understand customer behavior. There is a plethora of tools that could be used to build reports including some js libraries, Data Studio, excel, Tableau, and many, many others. In the presence of choice overload an engineer might settle for a tool that is easiest to add considering existing technology stack. An easily justifiable decision, based on assumed time and effort benefit, that disregards benefits of other options.

I think that this behavior also explains tendency to sticking to known solutions, thus blocking learning. My personal (and final) example: I have a few years of experience using Java and Spring, thus this was always my natural (easily justifiable) choice for building back-end applications. One time I was asked to expand a system with pure functional element and I must admit it was bloody hard to change my mind and use lambda with node-js. However it was worth it, because now I have more tools behind my belt.

Factors

The model proposed by scientists, besides consequences, proposes the existence of "antecedents" (causes/factors) of choice overload other than the number of options. This includes choice set complexity and decision task difficulty defined as objective factors and preference uncertainty and decision goal as subjective factors. Below is my humble attempt to translate them to plain English [1].

Choice set complexity

• Presence of a dominant option - when dominant option is present, the choice overload decreases
• Attractiveness of the options - a smaller option pool is preferred when it’s comprised of attractive options. While a bigger option pool is preferred in the lack of the above.
• Alignability of the attributes - refers to a situation when options share the same set of features. In the presence of the alignability, the choice overload tends to be lower.
• Complementarity - it is more difficult to make a decision in presence of complementary option pool, rather than non-complementary, as it is basically harder to wage tradeoffs.

Decision task difficulty

• Time constraints - given less time to evaluate an option, regret increases for larger option pools
• Decision accountability - the related research showed that a decision maker prefers bigger option pools, when he/she expects his decision to be justified, but prefers smaller option pools when the justification is actually required.
• Number of attributes describing each option - considering higher number of dimensions used to describe options complicates choice, thus increasing choice overload.
• Presentation format - proper display decreases the difficulty of choosing an item from larger sets.

Preference uncertainty

• Level of product-specific expertise - for experts, smaller option sets tend to increase choice deferral. For non-experts the situation is the opposite, bigger option pools tend to increase choice deferral.
• Availability of an articulated ideal point - for people with expert knowledge about the domain and product tradeoffs, it is easier to make a decision from a large option set. While in the case of lack of the above, it is easier to make a decision from a small option set.

Decision goal

• Decision intent - takes into account if the actual decision has to be made, or the process is used for example for evaluating or learning. The latter case is expected to lead to a lower choice overload effect.
• Decision focus - in more complex scenarios a decision maker might have to choose not only an option from a option pool, but also to choose among option pools. In the light of the above, researches claim that a larger option pool more likely leads to a choice overload in case of choosing among options, rather than among option pools.
• Level of construal - includes a decision maker distance to the decision making process and it’s outcome. Meaning how important it is for the decision maker. The lower the distance, the higher level of choice overload.

Words small/smaller and big/bigger might mean different things depending on the context. A decision maker can be a subject to the consequences of a choice overflow, regardless of the number of available options. It all depends on the context/factors. As an example to illustrate the fact, I would propose the first hire scenario. In theory it is a yes or no scenario, whether founder starts the hiring process or not. For a founder this is a very important decision, as she/he must wage the pros and cons, include the budget and time constraints. Then a decision on who to hire is a different beast whatsoever. There is rarely an ideal candidate, there are time constrains related to the window of opportunity, when a candidate is has not yet decided on a counteroffer. At the end in both cases the founder is the person accountable for the decision. It is a hard decision after another, with no possibility to making all of them right.

Summary

In the article we shed some light on what are non-obvious effects of a decision making, both on an individual as well as an organization. Furthermore we covered factors impacting the presence and level of choice overload, with hope that it helps in understanding and coping with the decision process.

References

1. Chernev, Alexander & Bockenholt, Ulf & Goodman, Joseph. (2015). Choice Overload: A Conceptual Review and Meta-Analysis. Journal of Consumer Psychology. 25. Pages 333–358. 10.1016/j.jcps.2014.08.002.