I wrote the following in a daily beemail and got back 20 replies, approximately 18 of which were gently explaining that I was being an idiot and that there’s not an ethical issue here (other than 1 person who said advertising in general is ethically dubious, admitting that’s a pretty extreme view):
Holly and Bee and I were discussing an interesting question about advertising today. Platforms like Twitter make it easy to pick demographics like “males 25-34” which on first blush might be a good one for Beeminder. Embarrassingly far into the conversation it occurred to me how messed up that is and I decided we should refuse to use protected categories like gender for ad targeting. Using things like household income seems like a business necessity though.
Like there should be a fair business case for any targeting we do. Having money, sure. Speaking English, yes (maybe even targeting only native speakers even though that excludes many people it shouldn’t, because the better your English the more likely you are to make sense of Beeminder’s hypernerdy webcopy). But having a penis, no, there’s no way to justify that even if the correlation exists. Like we can pick all sorts of other things that might (unfortunately) be proxies for gender like being a software engineer – there are plenty of reasons to treat software engineers as a coveted demographic – but targeting men because that yields lots of software engineers or whatever is bad.
Then it occurred to me to wonder what we’re potentially paying for this ethical stance. Like if, hypothetically, software engineers are a gold mine and everyone else is dead weight but there’s just no way to target them specifically, it might be pretty tempting to just target males. It could even, probably hypothetically, be the case that ads are only profitable by targeting males, if there’s really no way to get at the attribute you really care about that happens to correlate a lot with gender.
So, yeah, interesting stuff. Super curious what your thoughts are and if there are aspects of this I’m not thinking of, or not thinking of the right way.
PS: Holly clarifies that the analytics suggest 70% of current users are male.
There were a lot of great insights in the responses that I’m encouraging everyone to repeat for posterity in this thread. My conclusion is that what I called an ethical stance is still a good idea but just because we expect it to actually work. If the numbers say otherwise, it’s ok to do what the numbers say to do. [UPDATE: with just a couple caveats as articulated by @adamwolf below]
…It wasn’t me, but advertising in general IS ethically dubious, in the way it’s currently performed. But currently, anything Beeminder seems likely to do is going to be so far above Taboola’s “One Simple Trick” ads that it won’t even register on an ethical scale, so yeah, don’t worry about it.
My reply was more interested in how we can target the seemingly under-adopting demographic:
I think targeting ads makes the underlying assumption into a self-fulfilling prophecy a lot of the time. Mostly men sign up -> target more men -> women continue not to hear about it. Also then you will tend to miss enbies and people who just prefer to withhold gender online as well. I suspect analytics just ignore us or assume we go in one or the other bucket.
I’d be more curious how to make it sound appealing to women. Clearly we have enthusiastic female users: what makes them tick? Do women tend to select particular first goal types we can put in ad copy, or conversely can we make those goal types more appealing in themselves? Do we need to stick some testimonials from women somewhere obvious? Make the fact that Beeminder has women in the team clearer to make it more obvious it’s not a site where gender matters much? Do women tend to value social interaction with their tools, in which case we might want to make the forum more prominent and active and seed it with more posts (genuine ones though, obviously) of our own?
For probably obvious reasons I am sceptical of gender-based whatever. It works because of socialisation, and perpetuates it. Meh.
I guess I’d say that something you might consider is how much of an untapped market women might be for a friendly, colorful, non-dickwaving commitment system like Beeminder. I am perhaps atypical of my demographic in that I’ve been interested in self quantification and akrasia resistance for decades now, but I wouldn’t be surprised if women, as the primary managers of household affairs and the performers of a disproportionate amount of record-keeping and emotional labor, might really be hungering for a habit management system that can keep track of all the million tiny pieces of their lives and (through the all-powerful Beeminder sting) actually keep them on track. I know that as I’ve turned from a freelancing data nerd into a parent, I’ve had to get a lot more efficient about matters domestic and financial, and my Beeminder goals have gone from “am I getting incrementally better along several axes of self-improvement?” to “what do I need to get done today so that my life will not be on fire?” It’s become all the more essential to my daily functioning. And while I’m the breadwinner of my household, I do a lot of the work that women are expected to do.
A large part of what I love about Beeminder is that it’s kind of whimsical and not ultramacho about how habit-building works. Women are the primary market for planners and bullet journals and other such analog habit trackers. Why should Beeminder necessarily be more appealing to men? I think there might be some bias at play in terms of who it’s already been successfully marketed to and therefore who it seems effective to market to. But it’s possible that that sort of thinking leads to leaving real money on the proverbial table.
I think @nepomuk has a great point. While we’re probably all equally undemographically representative, most of my female friends are also very interested in planners, tracking, organizing, and the like, largely out of necessity. Some are into complex systems (which beeminder currently falls under) and some aren’t, but definitely the whimsicality and friendliness (both in look and in actuality – yall are very friendly!) can be leveraged to make it more appealing to people who aren’t necessarily in it for the sweet, sweet data, which includes a large portion of women.
(And as another recent mother, I definitely feel like I, too, am largely beeminding fires to put out, which is almost certainly more valuable than most of my previous self-improvement beeminding.)
I don’t understand why using some categories to advertise would be unethical. Now if you were to argue that advertising in general is unethical, I’d have a hard time disputing that, especially with somewhat monopolistic platforms like twitter where users pay with their privacy. (if you’re not paying for the product, you are the product.)
But those concerns aside, why would targeting your ads to specific groups be unethical? The untargeted groups will get the same number of ads anyway, and will still be harmed the same amount - they just won’t get (or will be less likely to get) your ad. From an individual’s point of view, ads are annoying and irritating, and therefore advertising causes harm, but I just don’t see how targeting ads is more unethical than not targeting. If anything, it’s less unethical as more people who might benefit from the ad will receive it.
You mention “protected categories” - keep in mind that which categories are legally protected against some forms of discrimination in a given state or country is a purely political decision, and not a good basis for ethics. So setting aside politics, what is the difference between targets you think are ok and ones you think aren’t?
So yeah - what am I missing? Why is it messed up or bad to target males? I think it is a reasonable assumption that males are in fact more likely to use and benefit from your service. No one is actually being hurt or denied a service from your targeted ads - it’s difficult to see how any harm is being done from the targeting.
It might also be worth noting that what you’d be targeting with some ad platforms is the platform’s guess about the user’s demographics, which is imperfect. That may raise additional issues - for instance, if a platform codes more techy-seeming users as male, you might end up actually targeting techy-seeming users by targeting males.
If you’re really worried about ad ethics, I’d argue a much more productive direction would be to realize that ads are an annoyance and disruption, that they make people’s lives worse in all sorts of ways, and to try to minimize that harm. (remember the torture vs. 3^^^^3 dust speck thought experiment from Less Wrong? Ads are worth at least 100 dust specks.)
The issue is not if you’re targeting men in your ad filtering imho, but if you’re promoting gender stereotypes through your messaging. For example, a copy like “beeminder creates strong men” would be unethical.
That said, you have no business-related basis to filter by sex. The 70% male/female current users percentages are not due to any gender traits but just the result of other factors, e.g. the “nerdiness” setup/overall branding of Beeminder.
It might help to get some perspective on demographics on the Quantified self community, and/or other apps targeted towards personal/health tracking.
I think it’s silly to target men for Beeminder ads. I also do not think it is unethical to target men or women with Beeminder ads, but I think there are targets that can be unethical for some types of advertisements.
Facebook allowed folks to target real estate ads based on ethnicity, and target other ads to “$RELIGIOUSGROUP haters”. I do not think those are good ideas.
While I trust the Beeminder founders, I think it might be safer as an industry to have some base rules that might be “overzealously ethical” and disallow a few safe ads, rather than to have no real base rules, and just say “think about the ethics of the situation every time you approve an ad purchase” because that’s what got us to a world where one of the largest advertising groups recently allowed “whites only” real estate ads. Before you protest “they don’t allow that anymore”, please wait 2 weeks for yet another article that reveals a new thing Facebook lets you target.
So is Beeminder overthinking it here? Probably, but I wish more people thought about ads this way.
At dreev’s request, I will paste my reply here. (I’m doing so verbatim, with no thought to others’ replies or my own reconsidering, to avoid worrying too much over the words.)
I understand the position you’ve taken, and would respect your continuing to pursue it. A couple of thoughts that popped into my head, and may not be worthy of any attention:
Perhaps this is a chance to reconsider the way you market to Beeminder. So many of your best user posts have been written by women, and most of Beeminder’s team is women (I think?); I’d be surprised if there weren’t a way to take those facts and combine them with Beeminder’s natural appeal to create a campaign that would appeal roughly equally to both genders. Or a campaign that appeals mostly to women, and a different one that appeals mostly to men. This could seem sexist, but it also seems pretty natural that people are slightly more likely to use an app if they hear user stories from people similar to them, and I’m kind of down for “Beeminder should use the best marketing it can so that people find it and use it and improve their lives, even if it means the methods aren’t ideal”. Speaking of which…
If we assume that the ideal state of the world is “as many people as possible use Beeminder, assuming they’d benefit from it”, then adding as many people to Beeminder as possible is good, period. If targeting males adds more people to Beeminder than targeting women, or targeting both genders, there’s something utilitarian to that. (Especially since Beeminder is not a product where community management is priority #1; if a social network were only targeting men, I’d see that as fostering unhealthy dynamics, but given the anonymity and lack of photos on the Beeminder boards, a skewed gender ratio at least seems less problematic, albeit not unproblematic.)
My incentives are that I want you and Bee and the team to be happy, and I also want Beeminder to be popular and financially healthy, and I think the latter promotes the former, but I don’t know whether the bad feelings from targeted ads counteract happiness more than any slight financial benefits aid in happiness. It’s a gray area, for sure!
Reposting my email response to Danny, at his request.
I am all about the world becoming a more equitable place, like, this drives most of what I do. I am also in every category of privilege except male, so it's possible that what I'm about to say is still blinded by that. With that caveat:
I don’t see any oppression in catering ads to specific demographics. You aren’t saying “males only” or “25-34” only; everyone is still welcome, and on the same equal terms. I also think that an equitable world is one that necessarily recognizes diversity, rather than saying “we’re all alike so don’t discriminate!” And, ad-averse as I am (I will never see any of your ads because I adblock the f*ck out of my Internet experience ), I think crafting different ads to appeal to different interests recognizes that not everyone is alike.
I am just wondering why you’re so certain of this - it’s certainly possible that the “other factors” do relate to gender traits. Generally males tend to be more drawn to more quantitative-based tools for a variety of reasons, though I agree promoting stereotypes is not a good idea.
Wait, stop! This is a perfectly reasonable, factual statement (the whole premise of this thread was about that very statistical correlation) but no one is allowed to respond to it because it takes the thread off on a tangent that could be contentious if people get stuck in and isn’t really germane.
I feel like a cromagnon for asking this, but why is it messed up to target by gender in the first place?
I totally get it in the case of job applications, housing, etc, where it’s about who gets a scarce resource. But not sure I see the issue with advertising to buy a product.
To make it more personal, if some company wanted to focus their ad spending to a group I’m not a member of, I’m having trouble seeing why that should upset me, even being explicitly discriminated against. I can think of lots and lots of products where I know I’m definitely not the target audience, exactly because of my gender, and don’t find that personally offensive in the slightest.
This isn’t a hill I care to die on — if it’s important to you on principle, that’s your right of course. Or obviously if it’s a legal concern, although my high level understanding is it’s ok here.
I’m partly asking out of genuine curiosity. But also partly bc as you say, I do want Beeminder to flourish, and maybe this sort of thinking could leave a lot of potential revenue on the table? Ad budget misallocated away from happy customers?
But I guess the underlying issue is: Why would Beeminder be gender specific? It doesn’t seem baked in.
So then using this sort of discriminator would simply be bad for business reasons, as opposed to moral reasons, if more precise targets (like Software Engineer, as you suggest) will get more bang for your buck.
Thanks so much for setting me straight on this, everyone! To be clear, that’s 100% sincere. I was wrong and you all articulated brilliantly why, and I changed my mind! Ah, discourse (lowercase-d discourse in this case, though for this forum platform, Discourse, as well and if I’d put my misguided thinking here in the first place I’d have saved some duplicated effort in setting me straight!).
Anyway, also special thanks to @adamwolf for vindicating my vague intuition at least, and making me feel like it wasn’t a total waste of time to have at least thought about the question so painstakingly!
I’d been asked to post my reply and I’m a bit late to the game, but thought I would throw it out there given that I had what seemed like a much more aggressive stance than some others here:
I don’t find this an ethical dilemma at all. If you believe in your service you get it into as many hands as you possibly can. Period.
Beyond that, you’re treating this like it’s zero sum when it’s not. Just beause one person signs up, doesn’t mean they take a spot from another. In fact, you could easily make the case that by foregoing signing up profitable dudes there is less money for the company, less people hired, less velocity, and the product doesn’t touch as many people as a result.
I love how you truly care about this stuff. I think this is one spot where you’re making up issues that don’t exist.
Here’s the benchmark I use to avoid the slippery slope when I’m engaging in persuasion. Am I making the other person’s life better? If I can answer yes to that question honestly, then that’s all I need to know. Even aggressive persuasion tactics are in play at that point.
The only thing I’ll say, is that you need to leave some buffer to account for the fact that when you persuade you’re necessarily going to be influencing the other person to adopt more of your values. Just because they work for you, doesn’t mean they’re appropriate for them. But this stuff is still around the margins. If you can help people you do.