Comment by bartkappenburg | Hacker News Reader

Comment by bartkappenburg | Hacker News Reader

Parent Full thread

We're actively using 3 principles of influence from Cialdini in our SaaS-product[1] (we also refer to it as 'CaaS', Conversion as a Service).

Basically we make persuasion available for the small- to mid-size online shops by providing a tool that integrates tightly with your (shop)backend. Based on data-analysis from your visitors (we ask for a small JS-snippet on each of your pages) we try to predict, by using machine learning, which visitor is sensitive to what kind of messaging at what particular time.

Examples:

Scarcity: "3 products left in stock", "order within 2 hours and 13 minutes and we'll ship your order today"

Social proof: "23 people bought this product today", "the last order of this product was 3 minutes ago"

Authority: "Recommended by institute x"

We're seeing uplifts in conversions rates from 5% to sometimes 200% (on average 10-15%). Key is timing and finding out what works best for a certain type of visitor. So, for example, shopping at your lunch break often results in a certain behaviour on a shop (ie. going directly or through the shortest route to your desired product) which implicates a sensitivity for a scarcity message. On the other hand, shopping in the evening (more time on your hands) tells us (but not always) that you're maybe more inclined to make a good comparison between products. A social proof message helps the most in that case. It's a simplified version of the reality but you get the idea :-)

The tool is easily integrated in shop software (Magento et al) and usable for the less tech savvy users. The messaging is delivered through non-intrusive pop-ups (maybe you've seen them) and are fully configurable on for example: referer, time of day, position, layout, animation, type of visitor, country, etc etc. We also have an integration with Google Analytics to track performance.

You can try us for 30 days free, shoot me an email at bart[at]conversify.com, mention HN, and I'll give you 15 extra days for free.

[1] https://www.conversify.com

(EDIT: I understood the messages to be misleading, reply says they are true)

This walks a line between helping the customer to make decisions manipulation (I'm being generous here). If I saw an online store doing this I would think less of them (again, being generous in my language) and go somewhere else. Also I can tell the difference between real scarcity (like an eBay auction timer out) and simulated scarcity...

... is what I like to think happens. Obviously it works and the chances of me, or anyone else who thinks the same way, being exceptional is probably slim.

We are not simulating scarcity, we are integrated with the shop's backend and show real stock. If a shop owner wants to lie about his stock that's his responsibility, but hey, he also can do that without our tool.

The tool only enforces certain usps or charateristics of the shop at the right time.

Good to know. I assumed they were lies: there's a lot of scumbaggery in sales, so it's perhaps a natural assumption.

No problem. Also, the shop wouldn't gain anything in the long run if they were telling lies. We measure conversion over multiple sales and that shop would see poorer performance for that message compared to our control group.

Some website are playing blurry lines games so much. They do use time / stock limited "flash sales" to trigger buying. Then relaunch a new flash sale a few minutes after. Most of the time you won't notice it, you're either busy looking at other stuff or just went off the web because you spent enough money. I even thought it was a page refresh bug until I understood the pattern.

Here is another one you can add that I have found works with sales that I make (although those are big ticket sales).

Attach some proof of what others payed in higher prices. As in "10 people paid $x for this last week".

Thanks! Will try that with one or two of our customers, shall I let you know our findings?

Definitely - startupdomainadvice@gmail.com

Pricing strategies are one of my specialties and I use them in negotiation almost every day.

It'd be interesting to know how Booking.com, a good example of these tactics, uses "scarcity" messages, as every time I try to book something on that website I always get messages like "only 2 left" and "5 people looking at this right now".

I think Booking.com was actually a client of Persuasion API; a competitor of Conversify (coincidentally all three Dutch companies.)

I think they actually ran this experiment only through their mailing list though but either way Booking.com was already using all these conversion tactics before though so at best they actually used this to occasionally turn off some of the less effective social proofs.

I have mixed feelings. On the one hand, I applaud how you turned those principles into a well-packaged, professional product with metrics, studies and whatnot. On the other hand, when I get subjected to those I feel like being... manipulated, to put it generously.

Regardless, your comment and the subthread it started gave me a product idea - an adblock-like plugin that would remove all such manipulative elements from online shops. Basically, it would cancel out all your hard work ;). But hey, it's an arms race between salesmen and consumers.

Thanks for the compliment ;-), regarding your feeling of manipulation: I think it's impossible to block all persuasive tactics in general. Let's take for example social proof: a set of positive reviews and their respective number of stars mentioned at let's say a product listing is a form of persuasion that, if taken out, makes the listing far less valuable. Is that something you would see that the persuasion-blocker would remove from sites? ;-)

Fair enough :). The problem of manipulation is something I can't really wrap my head around completely. To add another example supporting your point, I do at times care about scarcity information (especially when I need something within next week or two and I can't risk shop running out of a product and not restocking it fast enough) - but your comment reminded me that those can easily be gamed (especially if seller is less than honest).

I honestly don't know how to reconcile my own ideals with day-to-day capitalism. Brick&mortar stores all around not only manipulate, but sometimes blatantly cheat people. I realized recently that I probably couldn't open a store myself - my own conscience would run me bankrupt pretty quickly. I'm not made for this age ;).

If a persuasion blocker removed reviews from sites nothing of value would be lost, at least to first order.

For any book on Amazon, for example, 90%+ of the reviews are zero-information, because they either are entirely about the reviewer ("I thought this book was great/sucked/whatever" in various long-winded guises), are paid shills, compare the book unfavourably to another one by the same author (will will also have reviews comparing it unfavourably to another one by the same author) or are incoherent.

I've long thought that a review-sanitizer wouldn't be that hard to build that filtered out large numbers of reviews using what amount to anti-span techniques. Con-artists like you suggest that a more general counter-measures approach would be worthwhile.

You may object to being characterized as a con-artist, but you are precisely using algorithmic techniques to gain the mark's confidence so you can manipulate them to your own ends. That is the definition of a con-artist, and it's quite different from ethical salespeople do, based on my own experience in sales.

So here's a question: how would you deploy your algorithmic knowledge in a way that focused on qualifying rather than closing? The goal of an ethical sales-person is to qualify people in the channel, because a fully qualified lead is a closed sale. It may be a quite similar process, but you're asking a fundamentally different question. Not "how can I manipulate the lead to close" but "how can I identify the lead's needs and ability/willingness to pay?" Maybe it's not possible, but if websites focused on this rather than confidence tricks (which is what you are using) people would probably be much more willing to part with their money (the book "Getting Into Your Customer's Head", which is probably somewhat dated now, gives a fair idea of the approach I'm talking about.)

Thanks for your comment, and yes, I object to your qualification of a con-artist.

First, manipulation in my view is an unethical forced change of behaviour (buying something you don't want). We don't manipulate people with our tool (I'll come back to that later).

Secondly, con-artists are people that mislead and manipulate other people to have a (personal) gain, either financially or otherwise.

So, imagine an online shop. That shop has some usps (best widgets in town, fast delivery) and some information on each product (number in stock, number of buys etc etc). That information is usually 'embedded' on the site (frontpage, product details, checkout page). To communicate all those pros you would have to guess where in the user journey the customer has the need to read one of those perks (or you could just repeat them over and over again which hurts the user experience).

Our tool takes those usps and information and targets those at the right customer at the right time. You hate social proof? Fine, we won't bother you with that. You need a confirmation about the trustworthiness of the store? Fine, here's why the shop is to be trusted (reputation, expert in field etc etc)

We know what is important to you (qualifying) and communicate that. That's not manipulation or a con-artist's MO because it's true and we don't mislead you. We know, based on your data, what your needs are. (if a shop owner wants to communicate false claims, that's fine, but in the long run people are not buying there anymore. See my other comment on this subject).

Persuasion is everywhere, whether you like it or not. We just try to make it more effective by removing all the random persuasion messages (which are sometimes screaming at you) and finding out what works for whom and when.

> Secondly, con-artists are people that mislead and manipulate other people to have a (personal) gain, either financially or otherwise.

That appears to be exactly what you're doing. Unless, of course, you're trying to claim that your company is some kind of altruistic non-profit that manipulates people for the greater good.

I fail to see how he is misleading as his tool is meant to provide facts that "persuade". Of course scumbag sellers could use it with falsified data. However, that doesn't make him the con-artist.

Did you even read my comment?

I did, and I agree with the criticism. The fact that you're offering factual information doesn't change the fact that you're:

1. Deliberately manipulating buying decisions for profit.

2. Concealing the fact that the information you present is intended to manipulate, not to inform.

3. Ignoring whether or not a buying decision will genuinely benefit the customer and not just the seller.

Put simply, you seem to have zero interest in the customer except as financial prey.

Do you really not understand how shady this is?

It's true that persuasion is everywhere - but so what? Germs, pollution, poisons, politicians, and other bad things are everywhere too. That doesn't mean we shouldn't try to mitigate their damaging effects.

If you don't understand why this is a problem, consider how customers would feel if you explained your manipulation up front.

How much damage would the brand suffer?

Now, if you had customers who were happy with your presentation after an open explanation ("Yes - tell me when you're about to run out of stock so I don't have to wait until you restock") that would be different.

But instead of giving them that choice you're using textbook psychological exploits to push their behaviour around in ways that benefit you, potentially at their expense - and IMO that's certainly not a cool thing to be doing.

Let me touch on each individual point of your criticism:

1) Doesn't every single shop/business try to convince (I'm not using manipulate because of it's incorrectness and the negative association some people may have, see other comment) you as a potential buyer through the same principles? Is that a bad thing? Why isn't it 'allowed' in your eyes to emphasize your strong points (usps) or factual info about a product? What's the difference between our tool and mentioning these facts on the site it self?

2) We inform people about a product shop with factual information so that they can make a better decision. Example: A customer is worried if this shop is legit. Letting them see a message with the number of positive reviews the shop has got can make that worry go away. Social proof also helps in this process (just like we do in the real world when seeing an empty restaurant versus a crowded restaurant).

3) See (2). Our main goal is trying to convince customers of buying at a shop. Helping the shop with communicating their sales pitch to customers isn't about ignoring the interest of the customer. It's in the interest of the shop owner that he guards the interest of the shopper or else his retention or returning rates are going to be very bad.

Your comparison of germs (etc) with persuasion misses every point...

If I would tell customers that if they feel insecure about something (is this shop legit?, is this a popular product?, how long do I have to get it tomorrow?) our tool would provide them with the answers to that at the right time (using our algorithms and data-analysis), then I'm sure no one would object.

Lastly: people are intelligent creatures. If stock info on a product doesn't have your interest but you see it anyway, you'll ignore it (like we do all day with other inputs/noise). But if you are interested in a product and need it fast, than a scarcity message at the right time will certainly help you.

By your logic, 419 scammers aren't trying to manipulate people... just "convince" them.

You should read Cadlini's book Persuasion. I am sure it discussed what to watch out for in general life at the end of each chapter.

Thanks for the tip. I read it years ago, maybe it's time for another pass :).

It was a while ago I read it so I hope it was that one I am thinking about.

very interesting. i know for a fact i am influenced by scarcity messages. i can feel my pulse quicken a bit when i'm shopping and it says there's 1 or 2 products left in stock.

i can't remember if i cultivated this consciously or not, but when i see a scarcity message i think "well, if i'd hit this site a few hours later it would have been gone; might as well assume i did that and save my money". anything worth having will still be there later.

Are the messages true?

I'm guessing that the messages are tailored to whatever pushes the buttons the best. For example, if only one person has ordered today but it was 5 minutes ago, then show the "this product was last ordered 5 minutes ago" message instead of "1 person has ordered this today". Or, if there's lots in stock, then show the "you'll get this tomorrow" message.

I don't think it would be hard to make the messages selectively true.

Yes, that's basically how it works ;-)

If you can hook in FB data then you might be able to figure out so much more about the users/visitors (what's more likely to work for a given visitor).

I'm impressed by the product/service but I'm so jaded about 'being gamed' that I'm always cynical when I buy online.

I'm hearing time-of-day dependencies from a number of companies. It seems to be a major thing.

I wonder whether for some merchants the best message will always be "1 product left in stock" and therefore it is in the merchant's interest to restock only one item immediately after every sale...

Funny you mention it, it is actually a little more complex in practice. The 'soft spot' for number of items left in stock is somewhere between 2 and 5 (135 left in stock makes one postpone the buying decision obviously ;-)).

The problem with 1 left in stock is that people are responding well to scarcity in general but often don't like to be pushed toward a decision right now.... this is counter effective.

The other problem with "1 left in stock" is that I might want to buy two, and if you claim to have only one, I'll go to another site rather than order from you and wait an unknown amount of time for you to restock

Yes, that also plays a part in the effectiveness of the message.

Yeah, I usually distrust any stocking information - unless my need is urgent or I'm after some unusual item (like something handmade rather than manufactured) I assume a) that it's BS and b) that if it's that popular, it'll be back in stock soon enough, or available elsewhere.

The social proof of seeing other people order, oddly enough, I find much more persuasive even though it's just as subject to manipulation. Maybe I'm just jaded from the ubiquity of 'hurry in, this offer can't last' advertisements on TV, on the weekly coupon junk mail that shows up in the mailbox and so on. I always wonder why car dealerships make such a huge thing out of their sales, given that they seem to occur on almost a monthly basis.

Having owned a lot of uncommon, unusual, and vintage products over the years, there is intrinsic value to many other people owning the product beyond simple social proof.

Quite so. I'm into synthesizers and like any other specialist hobby, the more rarefied examples have significant costs of ownership besides scarcity, eg the difficulty of maintenance or finding people competent to do repairs if they become necessary.

The goal of our algorithms is to predict that you are more sensitive to social proof than scarcity.

In what case(s) did you get a 200% increase in conversion, and what's your explanation of that effect?

I messaged you, let me know.

so you are a digital used car salesman scum, nice

Reply on news.ycombinator.com