It makes so much sense. Buying energy efficient products. Lower running costs — thanks to higher efficiency — mean we end up paying less over the lifetime of the product. What’s more, even if the purchase price is a little higher than the less-efficient alternative, there are often hefty rebates kicking around, which means we don’t even have to pay more at the outset. It should be so simple. Yet, we still tend to opt for the less efficient option.
In previous posts, we’ve shared the results of experiments we’ve run at Enervee that show real promise in getting us all to make more efficient choices. We’ve shown that our proprietary Enervee Score, which ranks all products on a scale of 0–100, makes a significant (and positive!) difference to the products consumers choose within stated preference studies. And we’ve also seen that our customized energy savings predictions — that show what you’d save by buying a specific energy efficient machine compared to the average machine in its class — can also drive more efficient choices, under certain conditions.
Hold on, under certain conditions? To be precise, we’ve found the energy savings information — in the form of dollars saved — influences decisions with financially-stressed shoppers and leads to better choices. This makes sense — if we’re concerned about money and it’s top of mind, it’s very likely showing us a personalized projected dollar saving from an energy efficient purchase will make a difference and influence our preference. In this instance, energy savings in dollars becomes salient as an attribute or characteristic that steers our choice.
But that’s the only condition under which we see this piece of information (shown as a simple piece of arithmetic) making a difference. Which raises the question: can we make the dollar amount of energy savings salient for everyone? If we could make that happen, it would be another lever at our disposal to nudge, jolt or engage people to buy energy efficient.
How might we do that? One idea comes seemingly from a left-field area of research within psychology: fluency. To explain this better, take a look at these three mental puzzles, and quickly jot down your answers to each one.
All done? Great. We’ve put the answers at the bottom of this post. How did you do?
If you got one or more of them wrong, don’t worry too much. In the many many times this study has been run, typically more than 90% of respondents get at least one wrong. This seems strange, considering the sums are actually pretty straightforward. One clear argument (championed by Nobel prize winner Daniel Kahneman) for the alarmingly high error rate is extremely simple: we make mistakes because they’re so easy to make with these questions. For what looks like a simple task, we think a cursory glance at the information is enough to get to the answer. But of course it’s not — the questions are designed specifically to have us fall victim to this cursory approach. Kahneman refers to this cursory approach as our System 1, or ‘fast and frugal’ thinking mode. When we’re happy to use this mode (which in most cases, we are), we don’t catch the mistakes we make, because we allow our fast and frugal approach to call the shots — we’ve no reason to challenge it or to test the results we end up with.
But here’s a thing. Look at the same questions again, but this time in a font that is harder to read.
When people see the questions — the same questions, remember — written like this, guess what? The error rate falls to 35%. The same questions, and — to all intents and purposes — the same people answering the questions.
This is a ‘disfluency effect’.
As the name suggests, disfluency is the opposite of fluency, and describes the ease (fluency) with which we can process information. Think about your favorite household brands, for example — the chances are most of them have been designed with highly fluent names in mind, specifically to make them easier to read, recognize and recall (and to lead to purchase). But interestingly, when it comes to recognition and recall, if you already use a certain brand, the last thing that brand wants is for you to start thinking too hard about its merits compared to a rival brand. Fluency potentially leads to more superficial (and habitual) behaviors, for the reason they encourage us not to stop and think again and challenge the status quo.
But enough of toothpaste, kitchen cleaner and your favorite breakfast cereal. Let’s get back to our mental arithmetic examples above, and their potential impact for energy efficiency.
The font used in the second version of the mental puzzles above has been found to be a highly disfluent font (it’s Haettenschweiler, in case you’re interested, whereas the font in first case is Helvetica — a hugely popular font specifically because it is so fluent). Why does a disfluent font have this effect? Because the very fact that it’s hard to read means we cannot try and solve the puzzle with a cursory glance, with us instead having to concentrate on the information. This in turn activates our more deliberative, ’System 2’ thinking mode, which is the more processing-happy mode. And when that mode is active, our increased processing ability means we tend not to be caught out by the questions.
It’s important to stress this is not a question of smart people versus less smart people. It’s simply a question of activating a ‘reserve’ processing mode that we all have waiting in the wings, but that we try hard not to call on, for the simple reason that it’s more work. But when something’s hard to read — disflient — then we’ve no option but to flick the switch and work a little harder.
So if changing the font is enough to get people to make fewer mistakes with a series of simple mental puzzles, could doing the same thing be enough to get us to do the sums correctly with respect to energy efficiency savings?
Could presenting our energy savings information (which is, in effect, a small sum that we’ve already done for consumers) in a ‘disfluent’ font trigger more deliberative processing of the information, which would then make the result salient and constructive in the choice?
We thought it was worth testing.
We created a custom version of Marketplace, and once again asked consumers to state their preferences from a number of washing machines. Everything on the custom version mimicked the experience on the normal Marketplace, except that we presented the energy savings information in the same font and style as the original studies above (FONT).
It worked. Changing the font to one that is disfluent led to more efficient choices being made.
In other words, we find support for the argument that making energy efficiency less fluent — in terms of the dollar savings calculation from a more efficient purchase — causes us to focus and process the information more carefully, which in turn causes us to ‘recognise’ the financial argument for making a better buying decision.
Counter-intuitively, increasing disfluency leads us to do the math right when it comes to understanding the benefits of buying energy efficient.
There’s one more result worth sharing. As in previous studies, we ran this as a 2x2 design, meaning we also saw what happened when this disfluent approach was combined with our Enervee Score. As discussed above, without the Enervee Score, we see energy savings in dollars now working and leading to more efficient choices. But when it’s combined with the Enervee Score, we don’t see it having a significant effect i.e. the Enervee Score seems to deliver the effect when they’re both together.
Is this disappointing? Not at all. We’ve always argued that the Enervee Score acts as a strong proxy for energy efficiency, in that it both quantifies the efficiency of a product and allows us to instantly compare that product with every other relevant product. In this respect, the Enervee Score could create an energy efficiency heuristic — a mental short-cut to making a smarter choice. We know that a heuristics-based decision making style suits our System 1 thinking mode, and is the opposite to a ‘System 2’, systematic decision-making style (where we knuckle down and process the information). And we also know that we will always opt for an easy decision-making style if we can get away with it. Daniel Kahneman sums this up brilliantly: ‘Thinking is to humans as swimming is to cats. They can do it, but prefer not to’. Which explains what we see with our results: with no easy way out, we process the information and make a smart choice, but when the easy option is introduced (the Enervee Score), we revert to being the cursory decision-makers we all are at heart. Hence no effect when the two features are presented together.
This is all further proof that these two features of Enervee’s Marketplace — unique features, we should add — can not only both deliver more energy efficient preferences, but do so via different underlying mechanisms. Two horses, for two courses.
So a question: Is one better than the other?
No, we don’t think so. It’s more constructive to view them as two separate levers with which to drive better purchasing. Two distinct approaches to nudge us to buy energy efficient. So the better question may be this: when might each lever be more effective? We’ve already seen that the type of consumer can moderate their effectiveness, but what about other contextual factors? Does it depend on why you’re buying (remodeling versus a panicked purchase due to breakdown)? Does it depend on what you’re buying (TVs versus fridges) or the influence of specific product brands? These are all good questions, we feel, and we’re getting to them, one by one.
At Enervee, we’re focused on nudging us all to buy more energy efficient. That’s quite a journey we’ve committed to, and it means exploring, mapping and signposting every viable route we find along the way. It also means we try to be as robust as we can in terms of uncovering routes that we think can help us lead the way. And in this latest study, we believe we’ve found another signpost to help us. Interestingly — paradoxically in fact — it may work as a signpost specifically because it’s just so difficult to read.
The (all important) answers:
Question 1: 5¢, Question 2: 5 minutes, Question 3: 47 days