Hacker News new | comments | show | ask | jobs | submit login
Double your price (and no, I'm not kidding) (jacquesmattheij.com)
180 points by jacquesm on Aug 27, 2010 | hide | past | web | favorite | 101 comments

From my experience doing price testing, there are certain natural grooves for prices. Generally the fit the 1,2,5 model: $1, $2, $5, $10, $20, $50, $100, $200, $500, $1000.

These prices have a Goldilocks structure where there is a natural market price bounded by a high and low price point.

For instance, for small business software as a service, cheap end is $10/mo, high end is $50/mo, and just right is $20/mo.

In another case, price per user tends to be $5/mo./user at the low end, $10/mo./user is the fat middle, and $20/mo./user is the high end.

You can see how the OP fit this model as he went from $5/mo. to $10/mo. to $20/mo. Without knowing his product very well, I would surmise that each price increase signaled a higher level of quality to the market while still remaining within the reasonable and expected bounds.

You're picking good range pricing, I'd add to that in our experience odd numbered pricing - eg 79, 95, & 99 - outperform even numbers.

This seems to fit accross the board - meaning if you selected a price group to test of $10 / $20 / $50 as your first test, run a secondary test with your winner and even derivatives.

EG if you sell the most @ $20, try $19.95 / 19.97 / etc - you can squeeze suprisingly large gains out from that fine tooth testing.

We covered some of this a while back in our blog post on price elasticity testing w/some of our results too -


Exactly the same applies for the amount you charge in consultancy fees. A couple of years back I charged X per month in web development (spare time work) but have slowly increased to where I am now charging 4.5X; and still getting more offers than I can handle.

Always try upping your prices, that should be the first rule of business :)

One really excellent price hack if you're doing consultancy is to use price increases as a way to get rid of problem customers. Once you decide a customer is not worth your time & energy to maintain, every month or so, send them an email saying "Unfortunately, due to high demands on my time, I have to raise my rates by 20%. I'd be happy to refer you to my colleague <your biggest enemy> if this increase puts you over your allocated budget".

This is an essentially 0 risk way of doing price discovery and you will usually be surprised at how high you can go before people finally quit.

That's really true, but it can backfire spectacularly.

I once had a customer that was really in the 'nothing but headaches' category. They came with a new job and I really should have said 'no', but instead I figured I'd just quote a very high price and they'd go away.

They accepted :(

The problem isn't that they accepted, it's that you didn't quote a price high enough to offset the headaches they caused. If you had, you would have been happy!

I misjudged the depth of their pockets.

That's what I'm saying, if you had gone with a price that would make it worth your time instead of a price you thought they'd decline, it would be worthwhile. I see how you can fall into that trap, though.

Trust me, it was enough that I should have been happy and it still wasn't enough.

Then it wasn't enough :p We'll never stop arguing, but consider this: Wouldn't you be happy doing it for a billion dollars? Somewhere between that and what you got is your "happiness threshold", it's just that it's more than what you got now.

Unless you wouldn't have been happy with any amount of money, in which case I assume your client was the third Reich!

> Unless you wouldn't have been happy with any amount of money, in which case I assume your client was the third Reich!

Not quite, but close :)

Ah, I see now :P

The trick is to quote a price high enough that even if they accepted, you would consider the compensation worth the aggravation. You win either way!

There are not problem customers, just problems.

I think it's best to avoid beginning working with a client if you don't think you'll have an enjoyable, productive relationship. If you start having problems with an existing client, the best thing is to communicate openly and clearly about the problems and try to solve them together. It's important not to write them off as a "problem customer" but to give them a chance to know what isn't working and to solve it. If you're not able to work things out together, then it's time to part ways.

I had a phonecall from my broadband provider this week, saying "Since we took over this company, we've been reviewing accounts and you've had this for a long time - we can offer you the same service for less money".

I nosed around their site, and found I can regrade my account with a "just a price cut" button.

I'm mystified by why they would pay someone to seek out and cut their profit from previously happy customers. If they'd told me they had taken over the company and needed to increase the price by that amount, I would have agreed. Is this a sign that the new owners aren't very good?

> Is this a sign that the new owners aren't very good?

Upvoted you, but the answer is no, this might be a very smart move by them depending on their strategy.

If they want to lock down market share, this is a great play to build loyalty, especially if there's new competitors directly in the space (other broadband) or substitution products (3G, whatever). They might have run some numbers and figured X% of their customers would get poached if they remained at high prices, and it was better to preemptively lower prices and get a lot of goodwill in the process. Market share + loyalty. Max short term cash isn't always the best goal.

Edit: Especially in a business that's high capital intensive with low marginal costs. Furthermore, doing this might dissuade other companies from coming into the space to compete. Strategy can get complex, especially when you've got millions/billions already invested infrastructure.

Turnover is also a metric most service providers pay close attention to.

Simplicity is a sign of a well-run organization. You ever heard the line, "I don't lie because I have a terrible memory"? If you have 15 different price plans, you now need to maintain documentation, train customer service reps, detail tiered specs in a computer system, update all of the information, etc. You get the idea. Complexity leads to higher costs.

If they offer one (or a few) options with consistent service and pricing, they will do them better with higher consistency and quality.

Makes me think of In 'N Out Burger...getting hungry. :)

Churn rates and cost per new customer are two very highly watched metrics in telcos. Keeping you happy keeps up their numbers. That said, I wish they also bought my provider.

Reminds me of Roller Coaster Tycoon, the customers say stuff like "This park is a really good value" and you bump your entrance fee until they stop saying it. Except in the real world people rarely tell you your your product is too cheap.

I only somewhat agree. There are ways your users tell you your product is too cheap, without actually having to tell you. One good indicator is when you have a lot of people signing up, and then not using it, in which case your product is so underpriced that it attracts indifference.

I've seen companies sign up for a $3K/month product, then not use it for 6 months.

Large companies do silly things.

I've seen multi-$100k investments go unused for so long as to become worthless, most notably in the form of enterprise-scale battery backups which needed to be used within X months and were still on the original shipping pallets 2 years later.

I think the point of this article can only really be taken as applying to the consumer web; while its point might ring true for enterprise customers, they're a much more murky environment in general.

Sure I've seen it too, but I'm referring to trends here.

In the real world people will rarely tell you that they are going to a competitor if you don't stop raising your prices.

This relates to an economic principle called Price Elasticity. The Wikipedia article is pretty informative on the subject http://en.wikipedia.org/wiki/Price_elasticity_of_demand

What you've done is some basic "elasticity testing", which, in the "real world" can often be difficult or impractical, but lends itself particularly well to online transactions of relatively small dollar amounts. More people selling small services online should be doing this - traditional businesses would love the luxury of doing it so easily.

Yes, that's exactly what it is about. And about your own 'perceived value' versus what your customer thinks it's worth.

I've read some anecdotal evidence supporting this as well (atleast for goods). A store owner can't sell a line of jewelery, so they write a note to an employee to cut the price in half. The employee misreads the note and doubles the price of the jewelery instead. All of a sudden the line sells out.

Part of the reason is that if people notice something's cheap, they'll think the quality is cheap as well. And vice versa.

A/B testing prices would be ideal when trying to find out how much people are willing to pay for your product. Unfortunately it's a rather risky practice, so doing it over time as suggested by Jacques may be a better approach.

If it works as well for any of you as it did for me you will not regret trying this.

It's amazing how much you can be 'off' about what your product is worth to your visitors if you do not do drastic experiments like this to figure out the range of prices that are acceptable.

I think we all subconsciously price based on 'cost+', we price based on what it costs us to provide a service, tack on a profit margin and then we round that to the nearest reasonable sounding figure.

But you should be pricing based on perceived value. Not on your costs...

And the perceived value may be a lot higher than is apparent to you as the creator of the product.

Wouldn't that only work until a competitor with a sufficient product crops up with a cost+ price?

You'd be surprised how big the world really is. Plenty of people will have never even heard of you or your competitor. Good luck marketing your product to the point where you've gotten 100% saturation in terms of reaching your target audience. Even coca cola and google can't claim that.

I'm pretty sure Coca Cola can actually claim 100% saturation.

In fact, they probably can, after all, they're the #1 global brand. And those that they haven't reached yet they probably can't sell to. But it can be quite surprising to hear people say 'google? what's that?'.

It still happens to me occasionally (but rarely), and it never ceases to amaze me that there are still people living under rocks thick enough to block out information like that.

Same with wikipedia, you'd think that the news is out by now. But there are still people that have never heard of it.

Random anecdote: Back in 2002 I spent a week in a village in rural Haiti trying to get an electric generator set up. We had car trouble between villages and our resident mechanic was fixing it on the side of the road, when a woman from a small hut came out and asked us if we wanted Coca Cola. She sold us bottles of Coke kept cool in a hole with a tarp over it. I've never looked at the Coke brand the same way since.

How is that for saturation? Amazing story. I heard that Coca Cola was pretty much the first western brand in to the most unlikely places.

In rural Kenya, Coca Cola is the only thing visitors are advised to drink.

There are also a _lot_ of people who have heard of Wikipedia, and used it, but have _absolutely no idea_ that it's user generated and that you can edit it. They are generally pretty stunned when I demo it.

Hm. For truly niche software, I suppose I can see that, but for anything with more than even a few hundred customers, I don't think it'd last. Someone (like me, for example) would smell an opportunity and sell for less.

I suppose it's not a bad thing for you in the short term, but I'd caution against getting comfortable on the profit cushion, and reinvest in revenue sources that can't be taken away by the same opportunistic competitor.

I'm not sure how long 'it would last' would need to be for you to be convinced, but for me it has been going for about 6 years now since that conversation and it still works.

The interesting bit is that I was just as skeptical as you are (and probably even more so) and in the end decided that trying it was the easiest way to see if it was true or not.

The second time I was even more skeptical but then I figured 'in for a penny, in for a pound', just to be able to tell Richard that he was wrong the second time. But he wasn't.

It doesn't hurt to try though, and if it does work for you then it's free money, which is nothing to sneeze at.

Keep in mind that if your product costs you $2 to put out there per paying member and you manage to increase your sales price to $19.95 you've just gone from a 50% gross profit margin to 90%.

Understood, and I think I agree with you. I was thinking in terms of larger prices (i.e. $500 on up; I tend to work with custom and specialized software in my day job.).

I can almost guarantee you wouldn't build an entire product/website/SaaS just because you 'smell an opportunity' in a market 'with more than even a few hundred customers', would you?

Selling for less doesn't always work, because the more expensive guy can spend more on marketing.

Also, for consumers, $19.99 is pretty much a fixed limit. Above that, you loose people, but below, you don't necessarily gain more signups.

I suppose that's true; below a certain point, the effect of a price is pretty inelastic.

How about pulling off what OXO did? They realized their products cut be copied and sold at a cheaper price. So they created their own competition. Just made a worse design, used cheaper materials, slapped on a new brand and put both lines on the same shelf.

That way you can market to two demographics and cover the market. Extra bonus is that you can use essentially the same underlying tech.

That is useful, but it is actually just one example of price discrimination - and you might want to create more than one competitor to your product.

>How about pulling off what OXO did?

Made gravy?

OK, what did OXO do, they're a gravy brand in my country ...

Good point. I think Microsoft calls those SKUs. (They just didn't put the same brand name on it, I suppose.)

As long as your product is not actually identical to the competition, you'll be amazed at what little things count as differentiators for customers.

An anecdote:

My company produces a PDF reader for iOS+. There are, give or take, 20-30 competing products. One of them is iBooks, which is heavily marketed by Apple, a high quality product, and free. Most of the others are cheeper than us, but we're out-grossing and out-unit-selling many of them. There are at least a dozen products with non-zero day to day unit sales. One day before we went live in the App store I discovered that there were not one but two products with basically the same name as ours++.

We can't credit marketing for this either. To be frank, our marketing has been terrible. We've tried to build a better widget, but mostly I think it's that we have a slightly different widget.

From the perspective of a customer: I pay for Lighthouse and Github every month. Github offers issue tracking, and both Github and Lighthouse try to keep things pretty minimal, but I prefer Lighthouse. It just feels more like the bug tracker I would build if I decided to go down that road, and that's worth a few bucks to me every month.

+ Plug: http://ballisticpigeon.com/folio

++ We did our due diligence when we started writing the app... about a month in... before we started writing our user manual, but at about two weeks from launch we were in crunch mode and stopped checking anything. They were both released in this window.

Proposal for an ethical A/B testing process for price hikes:

A: The existing, lower rate. B: Advertise a higher rate. After signup is completed and the customer is fully expecting to pay a higher rate, inform them that the newer rates haven't "taken effect yet" and they will instead be billed at the previous, lower rate. This preserves both experimental validity and ethical pricing.

That won't work because you can't test retention at the higher rate of payment, and that's a very big factor in the subscription business.

1 months @ $40 vs 3 months at $20 ...

What do you mean by "risky"?

Concurrently testing 2 pricing schemes for the exact same service feels sketchy or unfair (to the customer) to me–especially if its a 2:1 ratio ($5 vs. $4.99 I can understand).

I'm sure it doesn't fall under "price discrimination" but it still doesn't feel right ... should I not feel that way?

It's understandable that you feel that way. But really, it isn't sketchy. Think of the lower price as a discount that wasn't offered to everyone. The people who paid the higher price thought the value of your service was worth it...no one twisted their arm.

If you're really against it, don't test two at the same time. Just double the price for new customers and see how that affects signups.

It would certainly be considered illegal in some countries (France, I'm pretty sure) to have two prices at the very same moment, based on random, for the exact same product.

Do you have any idea as to why it would be illegal? In my country it's not at all unusual to see a product in a supermarket being sold at a discount price in a display at the front of the store, but at the usual price on the usual shelf.

In France if two prices appear somewhere, you'd pay the lowest on request I believe.

My gut feeling is that A/B pricing would be illegal because it's unfair: a part of the buyers have a discount the others don't even know about.

IANAL though :)

How about gas stations? They routinely charge different prices for the same product from the same store?

Do you mean same brand or same physical store ?

I believe as long as it's a different physical store, it's probably possible to set a different price.

My impression (that would need more research) is that having two different prices for the same product at the same time inside the same physical store or on a website would be problematic; otherwise it's probably ok.

I think I really need to make more research - A/B pricing would definitely be interesting.

Not in my country they don't.

I'm not sure which is more common (fixed price across the country like here in Uruguay or Argentina, or store-by-store variation like the US/Canada)

That's not the risk. The risk is your users comparing notes and getting mad at you.

There is risk because users may find out and be less than impressed by it.

I'm currently in the beginning phase of A/B testing prices on ForeverList.com right now.

Here's how it works: The "general" category is assigned a base price of $1. All other categories are assigned "price multipliers". For example, the price multiplier for cars is 15. So, to list a car would be $15.

I have a table of ad codes which modify the base price by a factor of 1 to 5. When the site recognizes an ad code passed as a url parameter, it uses this multiplier, plus the category multiplier to determine the price shown to the user. So, for example, using an ad code with a price multiplier of 5 would make "general" listings $5 and a "cars" listing $75.

In addition, each ad code can have an associated promotion. Each promotion has a flat discount or is percentage based. So, an ad with a 5x multiplier and an associated promotion of 50% off would be $37.50.

When a user signs up, the ad code is associated with their account and they are always show the same prices. Likewise, returning users that haven't signed up are shown the same price as long as they haven't deleted their cookies. I get an email when a user is shown a price change just to be aware of how often this is actually happening and if it's becoming an issue.

If the user clicks on ad 1 and then ad 2, which have both different price multipliers and promotion percentages, the multiplier stays the same but the percentage off changes if the new percentage is higher.

I haven't executed the full test yet, but the idea is to find the optimum price at which users will buy a listing. If you want to see it in action, try this:

1) Go to http://foreverlist.com/why

2) Open a new tab and go to http://foreverlist.com/why?ad_code=FBSV45k (multiplier 2x and 25% promotion)

3) Clear your ForeverList cookies. In firefox this would be edit->preferences->privacy->show cookies, put foreverlist in the search field and click remove for each one.

4) Open a new tab and go to http://foreverlist.com/why?ad_code=FBLuzq6 (multiplier 5x and 50% promotion)

5) Compare the example price at the bottom of the comparison chart in the "foreverlist" column in tabs 1, 2 and 3.

I'm writing up a detailed blog post about the experiment, so if you're interested in the results I'll probably post it here or you can just check back on the site directly (There will be a "blog" link). I'm also working on improving the site design, so the full-blown test may be delayed for that.

After I've determined the optimum price and implement that permanently on the site, I plan to issue a refund to anyone who paid a higher price and thank them for their patronage, etc...

I'm hoping there is a market for ForeverList, jacquesm/tptacek are right about pricing and the winning ad isn't the one with a price multiplier of 0. :)

That's a really nice way to go about it, the 'plan' I gave is much more coarse (and assumes a single product). I'm really curious how this whole thing works out for you.

And I'm also really curious how many people will try the price doubling (once or more time) and will write about it.

How are you planning to do the statistics to come to the right answer?

I have one answer at http://elem.com/~btilly/effective-ab-testing/#slide94 but I'm interested in seeing how others tackle this problem.

I downloaded your scripts and will check them out after I've got all the data.

another mistake a lot of people make is try to base their price on competitors.

sure you can undercut their $3/mo...or you can bump your price up to $15/mo and use the difference on advertising.

Chances are that 99% of your customers will never even see your competitor...so you won't hurt your business by charging more...and you can use the extra revenue to promote your product in places where your competitor just can't afford to promote in.

> Chances are that 99% of your customers will never even see your competitor...

That seems to be the hardest thing to convince people of. 'But my competition'... Yes, you are aware of your competition but your users are most likely not. They found you instead.

This will stop working fast if you charge enough so that people start looking for alternatives though.

Exactly. There's a price point always where people will stop and say to themselves; "maybe I should shop around" in which case they'll find a competitor of yours and complicate things.

Is this kind of experimentation feasible for expensive products targeted at businesses? A product that costs $300 a month. It would seem dangerous to try this and get bad press and angry customers. The problem is (1) people know how much it cost before, it's not impulse buying (2) one business might decide to buy a second license and notice that the price has doubled.

So, is there a way out?

The best way to do this experimentation (based on my experience when I started selling RateMyStudentRental to colleges and universities), is to start directly selling to businesses before posting official prices. Put up a general "Call for pricing" information page (no, this doesn't necessarily mean anyone will actually call you), and hit the streets.

One thing we found early on selling to universities was that if we priced ourselves to low, some schools wouldn't give us the time of day, other schools would start comparing us to "competitors" who weren't our competitors at all and did nothing like we did, but who shared a similarly low price.

I still remember when we sold our first (small 3000 student) school for $6000/yr. When they gave us the check they said, "Eh, that's a drop in the bucket." Talk about leaving money on the table.

Anyone who does custom proposal/sales work has had those moments. The moment after they say yes when you suddenly realize you charged drastically too little. Live and learn.

Create a premium product with a number of features that you can defend as costing lots of $ to develop, see if you can interest people in the premium package.

If enough people do drop the original one and make the premium package the 'normal' one, offer a discount to your old users.

I don't think that will work. If you have a choice between "cheap but does what I need" and "expensive and does what I need", which will you pick?

People aren't going to spend more for the hell of it. I don't think this is a useful technique for determining the minimum viable cost for a service.

Every car ever sold had to compete with other cars that were 'cheap but does what you need', and yet, more expensive cars still get sold.

Even the same model car but a more expensive trim will sell for a higher price than the same model without the trim.

Features can be used to substantially increase the price of a product while only marginally increasing the cost.

You're assuming that people always act rationally when spending money. Any salesperson will tell you that they don't.

Start (too) high, go down. When you stop getting more-than-double the users for half the price, stop.

I don't think you piss off too many existing customers if they find out it's more expensive later on. They will feel like they got a deal. Now if it was half the price you could easily deal with some anger.

Very interesting.

The concern I could see with this would be that you raise your price for a few days - you get a few takers but not as many as you hoped, so you lower the price back down to the original levels. I'd be a little concerned that those who subscribed at the higher levels would not be happy paying a higher recurring monthly fee when they could now subscribe to the product at a cheaper rate.

Would you / did you just manage those users back down to the original payment levels for the future months payments ? Or leave them at the higher payment levels and manage on an adhoc basis should they raise issue with pricing levels.

I refunded them the difference and apologized for the 'mistake'.

It doesn't sit 100% comfortable suggesting this was a 'mistake' - not saying this doesn't work - just that I don't feel entirely comfortable with it.

To follow up on this - with A/B price testing how does this work when you are lowering and raising your prices all the time - what's the general approach - to admit 'mistakes' or refund the difference or just manage on an adhoc per issue raised basis ?

Whether or not you choose to call it a "mistake" and how you handle it depends entirely on the business model, image, product, customers, etc....

A recurring service that is purchased, where you priced things too high and saw prices drop off, and then quickly moved it back - telling the customers who paid too much, and who would likely get pissed and leave when they find out you were unlucky represent an opportunity to market yourself - saying "Hey, we over did it a bit - we're crediting what you already paid forward at our new reduced rate, and keep charging you at that rate going forward, we appreciate your business!". If anything, that's going to make your customer who was willing to pay double what you are now charging him happy, and spread the word. Who doesn't like to find out they just got a 50% refund?

Agreed, but I would feel even more lousy continuing to charge them the higher rate while charging less to others.

You don't lower and raise your prices all the time, you do it once to determine the sweet spot and then you stick with that for a much longer time.

For me it was exactly as described: 2, evaluate, 2, evaluate, *2 evaluate -> go back one step.

At each step we had at most a few hundred signups processed before we made the decision.

The fact that it took as long as it did is only because I was skeptical, I should have realized much earlier that it can't cost too much to try, after all your signup rate would have to halve before you start to lose money.

Makes sense. Did you refund all users or just those that raised an issue ?

All of the ones that were paying the higher rate.

It seemed to be the most fair way to deal with it.

It's also interesting to note how different the iOS AppStore is from most of other stores. Some apps keep changing the price from 0.79 eur to something higher and then back again.

This is easier to see with one of those apps that keep track of apps that lowered their price, some of them keep appearing on that list.

So what was the website/product in question for this post?


Fair warning, not always safe for work, though we do our best to get people to label their cams properly.

Btw, this is awesome: http://ww.com/dilliecam

Yes, it is. That's my favorite cam actually, the guy that set it up is super friendly.

ww.com is a huge grab bag of both the good, the bad and unfortunately the ugly. People got married there, it has caused divorces, I've seen babies be born and I've seen people die... It's been the project of a lifetime.

I tried to read your story, but the firewall over here thinks your site has pornography, thus is blocking your site. Anybody else seeing this?

> I tried to read your story, but the firewall over here thinks your site has pornography

Not that I'm aware of!

In that case, you can merely bask in the notion that perhaps your content is so good, filters consider it to be pornographic!

Maybe it's those sexy semicolons...

Best not use them here; or HN might get banned too... oops!

What about for iPad apps?

No idea. The only way to find out is to try. I do not develop for the iphone/ipad, if I did I would try it myself.

I think that the app store makes it so easy to compare stuff side by side that I would expect the difference to be smaller, but that's just a gut feeling.

Here's an example. Draft from 37signals. As soon as they launched there was noise about their pricing, after a week or so, it got quit even though sales were still going up. Not sure what they're up now with the pricing.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact