Interesting Learnings from Choice Architecture
Many of us, in our jobs, have to present choices to people and guide them towards making the best choice. Surprisingly, the way we present this information makes a significant difference to the outcome. This comes from the concept of nudging, where you present multiple options and guide the client in a certain direction, while still preserving their ability to choose something else. This is often better than not giving them a choice in the first place. Or giving them a choice in a completely neutral manner like the result of an SQL query (humans aren’t robots).
Let’s see some concrete examples to understand this better. I’ll present examples from my work as a tech advisor to CXOs, since this is the context I know best:
I offer multiple types of engagements. I could present them as a list:
Engagement 1
Engagement 2
Engagement 3
Engagement 4
But I instead present it this way:
I present my primary offering first and foremost, with others behind a link, to draw attention to the first. I do this because most clients would benefit most from the Outcomes-based engagement. Defaults are powerful — many people go with them.
The outcomes-based engagement in the above screenshot has more terms than are presented on the website. These terms are in the agreement. This is an example of progressive disclosure — don’t vomit all the information at the outset. Instead, divide it into chunks, and share each chunk as needed. This comes from UX design where, for example, a flight booking site doesn’t show every bit of information about each flight in the search results page. Only the important information is shown and the rest is behind a click. When you share a certain piece of information is an important aspect of choice architecture.
I’ve stopped offering certain options unless they ask, because I’ve found that proactively offering them confuses the conversation, and confused buyers don’t buy. Not offering a certain choice is an important aspect of choice architecture. Don’t look at each option in isolation and justify it — that will result in too many options, making it hard for the important ones to stand out. Instead, think of the overall experience.
A middle ground is offering a choice but discouraging its use. For example, I recently told someone:
Let’s work together this way: <details>.
<pause> There’s another option, which I don’t recommend because it will produce worse results. That would be <details>. As I said, I don’t recommend it because it will produce worse results for you.
In this case, I didn’t want to go so far as to not offer the second option, because the client may have a valid reason to pick it. But I wanted to guide them towards the first by telling them up front that the second option produces worse results. And to protect myself from criticism later if they do choose this option.
Notice the phrasing in “I don’t recommend it because it will produce worse results.” If I just said, “I don’t recommend it”, it’s not convincing. Why? Because it’s less profitable for me but better for the client? Who knows? Simply saying “it will produce worse results” leaves the conclusion that you don’t recommend it implicit rather than explicit, and explicit is always better. You’d be surprised by how many people don’t get it when it’s implicit.
Further, repeating the point before and after:
There’s another option, which I don’t recommend because it will produce worse results. That would be <details>. As I said, I don’t recommend it because it will produce worse results for you.
reinforces it. I also said it in a tone of voice without enthusiasm. Your non-verbal communication should be in sync with your verbal.
Tell them what you’re going to tell them, then tell them, then tell what you’ve told them.
- Aristotle
Don’t offer options that are dominated by other options. Dominated means better (or at least no worse) in any dimension. For example, don’t say:
You have a choice in billing. You can go with a guaranteed price of $100K.
Or you can go with hourly billing, at $175/hour. Set aside $100K to begin with, and if more is needed, I’ll let you know.
In both cases, all the outcomes you’re looking for will be achieved, and it will take 1 year.
Nobody would go for the second option since it can cost more while not delivering the project any sooner. Just eliminate it and offer the client only sensible choice(s).
Offer 3-4 choices to maximise conversion rate. Research has found that when consumers are offered a choice between (say) two types of sauces, more of them buy than if offered only one sauce. And when 3-4 choices are offered, even more buy. But don’t take it any further — offering 5 or more choices confuses customers and decreases the conversion rate. Before I send a consulting proposal, if there are fewer than 3 options, I stop and think whether I can add more. On the other hand, if there are 5 options, I identify one option the lead can do without and remove it.
Is there a difference between offering
A
B
C
and offering
C
B
A
?
Yes — people pay more attention to options presented earlier. So, in my consulting, when I present a proposal, I list the most desirable option first.
When you offer multiple choices, mark one as the default and say, “We’ll go with this option unless you say otherwise”. This reduces cognitive load while not reducing their choice. People want contradictory things — they want choice, but they don’t want to put in the effort to choose. Having a default option and not requiring any action to choose it reduces the burden of choice.
Two companies offered triple-pane windows insulated with argon. One company told customers that they’ll save $100 a month. Another told them that by not having well insulated windows, they’re losing $100 a month. The second company had significantly more sales. So you should think about whether to highlight gains or losses.
Present choices in terms of outcomes for the client, not technical details. For example, when I recommended Python to a client over Java, I told them: “This will result in faster time to market. And greater agility — you’ll be able to change things faster when needed. And a lower cost in terms of salaries.” Many engineers make the mistake of talking about technical reasons like “Python is a dynamically-typed language, where we don’t ask what class an object is; we ask what messages it responds to. This is called duck typing and…” The problem is that the client doesn’t care about any of the concepts you’ve introduced: dynamically-typed, class, object, messages, ducks. Ask yourself, “So what?” and tell the client the impacts it will have on their project.
Should you present a project as costing $1K per month or $12K per year? They’re the same, right? Wrong — in the latter case, you’re communicating that they should commit for a year1. Use the right units.
Sometimes it helps to frame things as a discount. For example, if you were to say, “I charge $100/hour, except for interviewing candidates, for which it’s $175”, it may come across as arbitrary. The client may think or ask, “Why should I pay more for interviewing?” Explaining that interviewing is depressing and so you charge more may not be convincing, because why should the client pay more if you dislike something? To avoid this problem, present it as, “I charge $175, but for <this kind of work>, there’s a discounted rate of $100.” Discounts are completely up to the seller, so people can’t question them. If you go to a restaurant and there’s a discount on Citibank cards, do you demand why it’s not offered for HDFC? And if leads do debate the discount, I respond with, “If you don’t like the discount, I can remove it and charge everything at $175. Do you want that?” This is an example of framing: we’ve framed $175 as the standard price, and $100 as the discounted price, rather than framing $100 as the standard price, and $175 as an elevated price.
Don’t break down a price, such as by saying “This projects costs $100K, of which $30K is for architecture and $70K is for implementation” unless you want people to be able to pick and choose, such as by saying, “I don’t want architecture, so I’ll just pay $70K for implementation.” This doesn’t make sense, since architecture is foundational, so don’t break it down. Just say, “This project costs $100K.” Sometimes, not sharing certain details is an important aspect of choice architecture.
Sometimes it helps to avoid saying “no” if you can instead say, “Yes, for an increased price.” For example, if you want 3 months to do a project, but they insist on 2, ask yourself, “If I were paid $100K, would I do it in 2 months?” If the answer is yes, then ask yourself, “If I were paid $80K, would I do it in 2 months?” If yes, ask yourself, “How about for $60K?” Eventually the answer will be no, and by conducting this dialogue with yourself, you’ll discover your price. Then tell them, “The project costs $40K normally and it will be done in 3 months, but if you want it in 2 months, it will cost you $80K.” This is better than saying, “No, it will take 3 months”. For some reason, the word “no” puts people into argument mode. They’ll turn off their mind and start arguing, “I want it! I want it! I want it!” It’s like telling a child he can’t have ice-cream. Saying “yes, but it will cost $80K” shuts them up.
Social proof, such as testimonials, are another part of choice architecture. If other people have chosen me and gotten good results, it’s likely the lead will, too.
Ranking and rating
You may like:
The billing can still be monthly. Billing and pricing can be monthly.