Manipulating Consumers with the Truth

[00:00:00] It is Monday, April 15th. Today, a deep dive into the world of marketing numbers. No, not

[00:00:08] ad metrics or website traffic, but literally how your marketing copy should use comparison

[00:00:14] numbers. For instance, would you say your allergy medicine reduces symptoms in 90% of

[00:00:19] people compared to your competitor which only does it in 80% of people? Or do you promote

[00:00:24] the relative difference? That choice could impact your sales. So which is the right way

[00:00:30] to use comparison numbers in product marketing? That is what our guest Kat Armstrong-Soul set

[00:00:35] out to discover, but before I introduce her, you may have noticed that we've been doing

[00:00:39] these interviews with marketing scientists on Mondays lately. That's partly because

[00:00:43] Monday is the slowest day for news in our space. What we want to know is, do you

[00:00:49] like them? Do you like these exclusive interviews on Mondays or would you prefer the regular

[00:00:54] Daily Report Mondays even if they're a little shorter than the other days? You can vote right

[00:00:58] now by swiping over to the show notes in your podcast app and tapping one of the two

[00:01:03] options or there's a web option as well. If you like these interviews on Mondays,

[00:01:07] go to b.link slash option one. That's the digit one. And if you don't like these

[00:01:13] on Mondays, go to b.link slash option two. Again, that's OPTION and then the

[00:01:20] digit two. And yes, we have tracking to prevent duplicate votes, so no cheating.

[00:01:25] Back to Dr. Armstrong-Soul who with her colleagues have published a research

[00:01:29] paper called Manipulating Consumers With The Truth. Relative difference claims

[00:01:35] in advertising and inferences of manipulative intent. I spoke with her

[00:01:40] last week and started by asking her what they set out to study. We set out

[00:01:44] to really think about the way that people and consumers process numbers in

[00:01:51] advertising and other marketing messages. So the way that marketers make choices

[00:01:56] around how to convey numbers in their messaging and then in turn the way

[00:02:02] that that's perceived by consumers and how consumers may or may not fully

[00:02:07] understand the actual value that's getting communicated based on the way

[00:02:12] it's being presented. And so the primary example in your paper was around

[00:02:16] describing the difference in a percentage. Can you talk about that?

[00:02:20] Yeah, so a lot of times in advertising claims you'll see information about

[00:02:27] improvements or in our case specifically what we look at is risk reduction.

[00:02:32] So for products particularly meaningful ones like prescription drugs or

[00:02:38] safety related products, a lot of times a marketer might want to say why their

[00:02:43] product is X amount better than an alternative. And so specifically in this

[00:02:49] project we're looking at ways of numerically conveying changes in like

[00:02:55] efficacy, how good a product is or how much safer it is or looking at different

[00:03:00] outcomes but looking at a change between using a particular product and the

[00:03:06] focal product that's being advertised or communicated about.

[00:03:10] And so the idea is that if something improves the product by 5% and then it

[00:03:17] improves it again or the competitor is improves it by 5% and then your

[00:03:22] product improves it by 8%, your paper studied what the consumer preference on

[00:03:28] whether they reveal those percentages or whether they talk about the

[00:03:32] difference in percentage.

[00:03:33] Yeah, so there's two basic ways and the paper is kind of jargony so I'm gonna

[00:03:38] try to make it less.

[00:03:40] I appreciate that.

[00:03:42] But there's two main ways to present the differences and it has to do with

[00:03:47] what you were saying about the baseline risk. And so people can make

[00:03:52] choices about whether to just show an absolute difference. So like for the

[00:03:57] focal example we that really got excited about working on the project

[00:04:02] all is a real world example about a youth football helmet.

[00:04:05] And so you could look at the differences of how much better the

[00:04:11] helmet is between the competitor based on let's just hypothetically say

[00:04:15] out of 100 kids that play football if they use a traditional helmet,

[00:04:20] eight of them get a concussion. If they use this new helmet only five

[00:04:25] of them will get a concussion out of 100.

[00:04:27] And so they could make a choice about saying, oh, there's a 3% reduction

[00:04:32] in risk because you're going from 8% to 5% or they could use a percentage

[00:04:37] change format which is called a relative risk reduction.

[00:04:43] And that's often almost always what we see in advertising which would

[00:04:47] be the 8% minus a 5% but then it would be over the original risk.

[00:04:56] And that makes the difference numerically much larger.

[00:05:01] Right. So it would say like a 30% reduction.

[00:05:04] Exactly. So it's the same numbers and they're both mathematically accurate

[00:05:09] and kind of factually true but they look really different.

[00:05:13] Absolute risk reduction information is always categorically

[00:05:19] face value smaller, sometimes a lot smaller as compared to a relative

[00:05:24] risk reduction which is quite a bit larger.

[00:05:27] Which do consumers prefer?

[00:05:30] That's kind of actually a complicated question and actually one of the

[00:05:34] reasons why we really wanted to research it because previous research

[00:05:39] we've done not in this project in particular but and other researchers

[00:05:44] as well have shown preferences for the relative risk format in many

[00:05:50] ways. So if you're just looking at how people evaluate a product they tend

[00:05:56] to evaluate the product more positively when they see the relative risk,

[00:06:00] the larger face value, probably not surprising.

[00:06:05] But also when you're asking like what do they prefer?

[00:06:08] It's not just that it influences how good they think the product is.

[00:06:12] They actually like the format better too because they report that

[00:06:16] it's easier to understand which is a little bit ironic because it's actually

[00:06:23] not possible to fully process that information if you don't have the baseline.

[00:06:28] So it's kind of like consumers like it because it feels easy but it actually isn't.

[00:06:35] Isn't that something that we as marketers are doing now anyway?

[00:06:39] Like are there any marketers that are providing the baseline,

[00:06:41] you know, five percent and now it's eight percent?

[00:06:44] Are there examples of products or categories where that decision is made?

[00:06:47] Yeah, that's a good question too.

[00:06:49] So the FTC encourages, you know, with their guidelines for advertisers

[00:06:55] to always disclose baseline risk information.

[00:06:58] So they recommend that if you are using this kind of format,

[00:07:02] you should be including or at least sharing like a link often in online

[00:07:07] situation.

[00:07:07] And then you'll see like click here for more information, you know,

[00:07:11] on how this claim was created and then you can kind of get access to all the

[00:07:17] actual numbers that go behind it.

[00:07:19] But you also see even embedded within marketing messages sometimes like in

[00:07:23] fine print they'll give information on the actual study itself where the

[00:07:29] numbers came from and that often will include the baseline.

[00:07:33] But that's buried isn't it usually?

[00:07:35] And it's under like a like an asterisk or a link click out?

[00:07:39] Yes, yes it is.

[00:07:40] And you know, connecting back to your first question, even when that

[00:07:44] information is included that that would help a consumer actually take the

[00:07:49] proper meaning from it.

[00:07:51] So say they use a relative risk claim and they say, oh, this helmet reduces risk

[00:07:55] of incursion by 32 percent.

[00:07:57] And then real small down at the bottom, you know, if you zoom in,

[00:08:01] you can read the actual baseline that should help a consumer more accurately

[00:08:07] perceive that claim.

[00:08:09] Actually consumers report they don't like when that's included, even though

[00:08:13] that's what we think is in their best interest to disclose that information

[00:08:17] actually kind of acts like a signal or consumers start questioning and have

[00:08:21] to think about processing and how the numbers were calculated and all that.

[00:08:26] I don't want to sound like the greedy marketer, but that I

[00:08:31] mean, that's consumer preference.

[00:08:33] What sells better?

[00:08:35] Yeah.

[00:08:35] And that's actually probably one of the big takeaways of the paper.

[00:08:40] And actually my some of my co-authors and I have had this conversation

[00:08:43] a lot for decades of like, well, they're both accurate.

[00:08:48] You know, is this, you know, a deceptive practice or not when you

[00:08:54] can show that they're real?

[00:08:56] And, you know, to your really pointed question, the outcomes that we're

[00:09:03] looking at are like, how good is this helmet?

[00:09:05] How likely would you be to buy this helmet?

[00:09:08] The relative risk disclosure is better on those dimensions.

[00:09:14] It's just relative risk meaning which version.

[00:09:16] I'm sorry.

[00:09:16] So the larger version that shows the percentage change without

[00:09:21] the baseline information.

[00:09:23] Um, so the larger claims 30% would, you know, sell more helmets than

[00:09:28] saying a 3%, even though, you know, they're both.

[00:09:32] So it's really the project was specifically looking at, you know,

[00:09:37] could this be considered to be kind of a non deceptive version of

[00:09:41] manipulation of consumers?

[00:09:42] Even though it's good for these marketing outcomes.

[00:09:45] Is it respectful to consumers?

[00:09:48] Is it appropriate when consumers are being, um, their responses

[00:09:53] are being systematically biased by this presentation format?

[00:09:58] While many investors chase a tech boom that has yet to fully

[00:10:01] materialize and major indexes already flirt with record highs.

[00:10:04] It feels like we're running out of those under the radar

[00:10:07] opportunities with the potential to elevate your overall

[00:10:10] portfolio.

[00:10:11] But with Masterworks advisors, you can diversify your

[00:10:13] investment portfolio beyond stocks and bonds.

[00:10:16] And into blue chip contemporary art.

[00:10:18] It's an asset class that's outpaced the S&P 500 for the last

[00:10:21] 28 years by 64%.

[00:10:24] And the process doesn't require millions or art expertise.

[00:10:28] In fact, over the same period, art investing has been an

[00:10:31] effective way to reduce volatility in a portfolio

[00:10:34] relative to rate of return.

[00:10:36] So schedule a call with Masterworks advisors at Masterworks.com

[00:10:39] slash skip.

[00:10:40] That's Masterworks.com slash skip.

[00:10:42] This advertisement relates to the number of investors

[00:10:45] this advertisement relates to the provision of advisory

[00:10:47] services by Masterworks advisors LLC and is not intended

[00:10:50] to offer or solicit investment in any securities.

[00:10:53] Masterworks advisors is affiliated with Masterworks.

[00:11:00] Is there a way to get the best of both worlds where,

[00:11:02] you know, we're disclosing the, I guess it would be

[00:11:06] the absolute numbers, the real numbers, but still

[00:11:09] somehow getting the benefit of that large percentage?

[00:11:12] Yeah, that's that's a really good question.

[00:11:15] I think that some of the research would indicate that

[00:11:19] showing both, and we have this in the paper too,

[00:11:23] that it doesn't hurt the good consumer outcomes of

[00:11:28] purchasing the product to include that baseline information.

[00:11:34] Really our concern as researchers is just more

[00:11:37] the consumers inability and lack of motivation to actually

[00:11:43] process that information.

[00:11:45] And really this project, they're dumb.

[00:11:48] No, no, I don't think so.

[00:11:50] I mean, we also do test something where we teach

[00:11:53] consumers how to do the math.

[00:11:54] I mean, I'm I love consumers and I would never say something

[00:11:59] like that. We are consumers as well.

[00:12:01] Yeah, and I'm a consumer actually.

[00:12:03] So I would never say that.

[00:12:05] But I do think that, you know, we know that

[00:12:07] consumers do a lot of superficial processing.

[00:12:10] We know that there's a large level of enumeracy in the

[00:12:14] population.

[00:12:16] I don't think that means people are dumb.

[00:12:18] I just think it means that they don't have the

[00:12:21] skills and are not properly motivated a lot of times

[00:12:24] to actually implement the skills to do the statistical

[00:12:29] work to to process it properly.

[00:12:32] I know you didn't study this particularly, but do

[00:12:34] you have a gut feel on which industries or product

[00:12:38] categories where this effect might be the opposite

[00:12:41] where consumers might not want this relative larger

[00:12:44] percentage number difference?

[00:12:46] Hmm.

[00:12:47] That's a that's an interesting question.

[00:12:48] I think I guess framing wise, if you were it wouldn't

[00:12:53] be category related, but in any category, if you

[00:12:57] are kind of talking about the difference in the

[00:13:00] opposite light where you'd want to make the

[00:13:01] difference look small.

[00:13:03] Um, yeah, fair find the opposite like for example,

[00:13:06] I'm just thinking of like a price change, you know,

[00:13:09] like where an increase is negative or if you were

[00:13:12] talking about your competitor and so you're talking

[00:13:14] about that they had a decrease.

[00:13:17] I guess you would see the mirror response there.

[00:13:20] What made you want to study this?

[00:13:22] Well, I am a researcher who's really dedicated to

[00:13:25] consumer protections.

[00:13:27] Um, and I have spent a lot of time thinking

[00:13:31] about numerical processing and systematic biases

[00:13:34] that consumers have.

[00:13:35] And I think this one is particularly an issue because

[00:13:39] we see these kinds of claims and maybe less impactful

[00:13:43] like paper towel, quick, you know, uh, more absorbent

[00:13:47] types of increases, but um, you see it a lot.

[00:13:53] It's just really pervasive and prescription drug

[00:13:56] advertising.

[00:13:57] And like I said before, like safety related products

[00:14:00] and you know, those are the things where the

[00:14:03] products are really expensive and the potential

[00:14:07] negative impact for consumers is really high.

[00:14:10] You know, maybe negative side effects or, um, really

[00:14:14] important FXC related information if you're

[00:14:16] talking about heart disease or, you know, all

[00:14:19] sorts of medical things.

[00:14:21] And so to me it feels like a very, um, consequential

[00:14:26] kind of context that you see it in a lot.

[00:14:28] So I guess definitely from a consumer welfare,

[00:14:32] consumer wellbeing, protection angle, um, of

[00:14:36] trying to understand if there even is a way that

[00:14:39] those types of claims can be used in an ethical,

[00:14:42] responsible way as, as marketers versus making

[00:14:47] sure that consumers are making the best consumer

[00:14:50] choices for themselves.

[00:14:52] What surprised you the most about what you discovered?

[00:14:54] I think, you know, for years, the surprise has

[00:14:57] been how historically difficult it's been to

[00:15:01] get consumers to understand the different types

[00:15:06] of, um, numerical claims and, uh, how pointed a

[00:15:13] researcher has to be about getting a consumer to

[00:15:17] really think about the numbers.

[00:15:19] I guess maybe like the, the, uh, avoidance, consumer

[00:15:25] avoidance of wanting to do math and really process.

[00:15:30] I think that that's probably been the biggest

[00:15:32] surprise of like results after results of being like,

[00:15:36] you know, we've explained how you calculate this and

[00:15:39] we've shown how it's used and we see the change of

[00:15:42] numbers and consumers being like, that's too

[00:15:45] complicated.

[00:15:46] I don't want to, don't want to deal with that.

[00:15:49] You had co-authors on the paper.

[00:15:51] Who are they?

[00:15:51] Yeah.

[00:15:52] So, um, my, the first author of the paper is

[00:15:55] Dr.

[00:15:55] Bob Madrigal.

[00:15:56] He was my advisor at University of Oregon and now

[00:15:59] we've been working together for, like I said, over,

[00:16:02] over 10 years.

[00:16:03] Um, and also my colleague Jesse King who is now at

[00:16:10] University of Montana and also was a doctoral student

[00:16:13] with me way, way back when.

[00:16:15] So, you know, like I said, we've been working

[00:16:17] on the project for a long time.

[00:16:20] Well, it's really interesting research.

[00:16:22] It's, I think really important as well to kind

[00:16:24] of see how everything shifts, especially in a world

[00:16:27] where inflation and shrinkflation is making those

[00:16:29] numbers move quite a bit.

[00:16:30] I'm delighted you were able to share it with us.

[00:16:32] Thank you for your time.

[00:16:33] Oh yeah, thank you so much.

[00:16:34] Really appreciate it.

[00:16:37] The Dr.

[00:16:37] Kat Armstrong Sewell, the paper she co-authored is only

[00:16:41] a Google search away.

[00:16:42] You're looking for a paper called Manipulating

[00:16:44] Consumers with the Truth.

[00:16:46] Relative difference claims in advertising and

[00:16:49] inferences of manipulative intent.

[00:16:52] Again, these Monday deep dive marketing science

[00:16:54] interviews are a bit of an experiment.

[00:16:56] If you like getting these on Mondays instead of the

[00:16:58] daily update, please visit b.link slash option one.

[00:17:02] That's the number one or if you prefer the old

[00:17:05] school daily updates instead on Mondays like we

[00:17:07] usually do visit b.link slash option two or just

[00:17:12] swipe over to the show notes right now in your

[00:17:13] podcast app and click either of the two links

[00:17:16] there.

[00:17:17] I'm Todd Maffin.

[00:17:17] Thanks for listening.

[00:17:18] See you tomorrow.