Hacker News new | past | comments | ask | show | jobs | submit login
Pricing Experiments (conversionxl.com)
202 points by DanielRibeiro on Aug 11, 2012 | hide | past | web | favorite | 29 comments



I'm dealing with a rather strange pricing dilemma now, actually. My startup sells a somewhat-arcane electronic test instrument which is generally used by larger corporations, government agencies and defense contractors, and universities/national laboratories. It's advertised for sale at $4995.00 in 1-4 unit quantities, which is an insane bargain compared to what these users are accustomed to paying for similar equipment.

I sold one to a particular Fortune 500 company a few months ago. Recently someone else from the same company -- a fairly highly-placed manager, going by his LinkedIn profile -- emailed to ask for a quote on a second unit. Except he insists that his budget is in the $4800 range, and he needs the quote to come in under $4995 in order to make the purchase.

I can't begin to guess how to interpret this, or how to answer it. Lowball offers are nothing new, and generally easy to turn down since demand for this piece of gear is in line with projections. And there's plenty of room for negotiation on purchases of >4 units. But a lowball offer for one unit, from a company with a $15B market cap, asking for a discount that's negligible to the point of wasting both parties' time? WTF?

I'd like to maintain a good relationship with this company and do further business with them, but I don't understand the signals they're sending with regard to price tolerance. Obviously if I were asking $5995.00 instead of $4995.00, they would have tried to drill down to $5895.00. I have to believe that the same would be true if the advertised price were $3995.00. Could I have avoided this bizarre negotiation by picking a price of $X495.00 instead of $X995.00?


The game theory for large corporations asking for discounts: If we don't ask for a discount, we don't get a discount. If we ask for a discount, sometimes we get a discount. Thus, we write in our Big Book of Purchasing Rules to always ask for a discount. (Or you just ran into one guy who likes the psychic charge out of winning a concession.)

P.S. You're only an insane bargain for one sale. After that, your price is the reference price.


this P.S. is brilliant. If it's a bargain then you may think about increasing your prices?


and it applies to every market, particularly labor.


Adding to what Patrick said:

Your counterparty is usually evaluated on the delta he can generate from your published list prices, and virtually nothing else.

So, if you want to, you can game the system by extracting other concessions. For instance, you can create a SKU with a reduced price but for which support or updates are expensive; your counterparty gets to record achieving a discount, but you ultimately make more money out of the deal. Another common concession to wring out of these deals is commitments to future purchases.

You would only do this if you want to; that is, if you think the opportunity to come out ahead financially is real enough to play the game. If you just want to make the sale, say "no, in low volumes, our price is $4995; sorry". They'll buy.


Maintaining a good relationship means a relationship that is good for both of you, and succumbing to tactics like that is signalling to the customer that you are willing to be taken advantage of. As a customer there is no downside to asking for a discount.

this is a perfect opportunity to upsell. let the customer know that you're unable to negotiate on small purchases, but for high-volume orders you can absolutely work down to a $4800 unit cost. If the company has already ordered one unit and is coming back for more, you know they are happy with the performance of your product. You hold all the cards here.


Early in my career (mid 20s) I was the type to try to beat down prices no matter what. I worked for a small company and the dollars certainly meant more to us than to a $15B market cap corp. But, in reality, I was just looking for something to do and trying to distinguish myself however I could. When it came time to talk about a salary increase, the thousands I had saved the company was one of the things I could talk about.

This is a weird situation. My guess is that the guy just gets a small thrill out of it or feels like he wouldn't be doing his job if he didn't at least try. So, if the price is $4,800 for 1, what's it going to be for 5 or 10? Personally, I'd just politely tell him that you're already undercutting the competition. It's a fair price and best you can do for 1 unit.


Perhaps, agree to their price if they commit to buying several units?


sometimes there are prices that are set by the bosses for what can be bought without putting in for budget approval.

For instance they have a credit card that they are authorised for charges up to $4800, but anything over $4800 has to go through budget approval, that could take up to a year to get.

Ask questions why they need it and then you can negotiate. If this is the case then ask them if you can charge them $4800 for the device and invoice them $300 separately for delivery or something like that.


Exactly this. Go sign up for patio11's mailing list [1]. He just sent out an article to all his subscribers that talks about selling into Fortune 500 companies.

In all likelihood the manager doesn't want to have to go through corporate purchasing (or sole source). He wants to use his p-card, and that's his limit.

[1] http://www.kalzumeus.com/2012/05/31/can-i-get-your-email/


Really liked this article. It pulls together a number of things I had heard before with some new ones. We are all quite a long way from the economist's idealised rational actor.

I released my first real experiment http://itunes.com/apps/fastlists (yes it's another lists app but it is currently free, with no ads and has a couple of useful features making it especially useful for reusable lists for shopping, packing etc. The design is functional rather than beautiful).

I'm currently adding some in-app purchase options to test the reactions of the users.

My current idea is to offer the following options: (price tiers equate to $1 or local equivalents I think)

1) No-ads (tier 1 price)

2) Privacy - No anonymous usage collection.(tier 1 price)

3) Raise item limit (by 100 to 200 items) (tier 1 price)

4) Unlimited items (tier 3 price)

5) Everything including future in-app features (tier 5)

For me option 5 is important because I hate feeling I will need to keep paying bit by bit for something. It will include everything that doesn't need server side support and expense. I actually hope most customers choose this, the other items are there largely to justify the pricing of this item.

Ads and anonymous usage monitoring will come in releases after the purchases have become available as I want to give people the option to opt out first and for ads I will probably make the first month of use add free to try and get the user stuck in. I may also add some alternate skins as in-app purchases too later although that is lower priority than using the iOS 6 social framework as a way to try to get users promoting the app for me by sharing lists.

There is currently no item limit so I will allow existing users who installed before the limits were introduced to have unlimited items on the devices they have already installed on.

At the moment this is just a small scale test because I haven't done any promotion for the app yet so the current numbers of users is small although the reviews are currently very good (4.5* average in US and UK).

Any thoughts on the pricing model? Or the app?


Re-read the article and pay attention to the economist & decoy pricing example. If you want people to buy option #5, then have less options, and make choosing 5 a no-brainer.


My first thought is that there are too many options, and the privacy option should really be a user-controllable setting to begin with.


Even with privacy off I won't get or keep any personal information just basic usage stats that will help me understand how users are using the application. This is something that I value the ability to do, it should be beneficial to users as I can improve the product so I would prefer not to give it up for free although I do want to offer people the option of not having any data sent back.

Things that I plan to monitor when I have time to set it up include things like frequency of use, number of lists, number of items, maximum depths of lists. Whether it is an iPad or iPhone it is being used on, what the default language is, whether they are using the email features and what items they have purchased.

I will have a persistent ID for each Install so that I can track usage overtime and try to identify if people stop using the app but it won't tie directly to any person or even device that could be identied by any means outside the app.


I think it gives the wrong impression though. People don't expect to have to pay for their privacy (the level is regardless - the default expectation is that data is not being kept), so it will get people's noses out of joint.


I would get some usage information to check the assumption anyone is hitting the 100 item limit to begin with.


And if you want your product/service to be perceived as very cheap, particularly if it's a fairly commoditised product: make your website look as cheap and awful as possible, à la http://ryanair.com


The typeface used in their logo...

I would love to read more about this method. Has it been thoroughly studied?


That's actually a big improvement on what their website used to look like: http://wiki.darkpatterns.org/Trick_Questions#Example:_Ryan_A...


Really good article, with a lot of actionable tips.

With that said... I can't be the only one who is tiring a bit of the crediting of Steve Jobs with nearly-mythological feats. For example:

>People used to download music for free, then Steve Jobs convinced them to pay. How? By charging 99 cents.

Even ignoring the "people don't download music for free anymore" implication, it seems pretty obvious that charging $.99 for a download isn't what led to the success of iTunes.


Agreed. You could buy music for as little as $0.50 a song through RealPlayer.


But what if it were $0.49?


MIT's OpenCourseware makes available a set of very good course lectures on pricing here: http://ocw.mit.edu/courses/sloan-school-of-management/15-818... Although unlike the article, it focuses on economics rather than psychology.


The psych is covered in the 3rd set of slides (Pricing Under Consumer Uncertainty). As someone who took the course, it was excellent and the outlines actually do cover the majority of what is offered in the class. I would highly recommend reading through them, though they are a bit dense and lack the color and charm of her slides (not on OCW).


A/B testing your prices to find your most profitable price point over time can help you remove guessing from your pricing strategy.

The author mentions that A/B testing among concurrent visitors can get you bad PR but forgets to mention that A/B testing prices over time is the only way to confirm your price is the best. Anything less is just speculation.


dead right. whats works for someone won't necessary work for anyone else - you MUST A/B test. i increased my ROI week by week with a/b testing and in my mind now if you not doing it then it's a crapshoot.

But there are things that are (asymptotically) absolute truths in marketing and i loved the article. i bookmarked the site which i don't do very often


"Under the simple pay-what-you-wish variation, 8.39% of people purchased a photo (almost 17 times more than before), but customers paid only $.92 on average.

The final option — pay what you wish, with half the purchase price going to charity — generated big results: purchase rates of 4.49% and an average purchase price of $5.33"

There is a significant peculiarity here that the author ignores. Yes, pay-what-you-wish generated more money when the purchaser was told that half of the money would go to charity, but the percentage of people purchasing was cut NEARLY IN HALF when they were told that half of the money would go to charity. Doesn't that seem odd? If I tell you you can purchase a photo for whatever price you want, and you are willing to pay $X, and I then tell you that half of the money will go to charity, would you then decide that you are no longer willing to purchase at any price at all?


This brought to mind something from Cialdini's book, Influence. In one of the case studies he argues that in the absence of a reliable way to assess the quality of something, consumers tend to use price as a proxy for quality. In his example, a jewelry store has been unable to sell one of their product lines, so the store manager leaves a note to an employee to mark down the jewelry by 1/2. The employee misreads the note and instead marks up the jewelry x2, which leads to the jewelry quickly being sold.


Lots of good examples here. Ariely's first 2 TED talks are great, but be careful with behavioral economics. For most start-ups, thinking about pricing in terms of hardcore microeconomic theory is probably a better route. The fact that so many digital goods truly have a marginal cost of 0 for the producer brings life back to the "Let's assume we have a constant marginal cost" phrase so many bright eyed Principles of Microeconomics students (and MBA students) have heard coming out of their economics professors' mouths. My point is that if something is a fact and not an assumption, then that's one less assumption you have to make.

In most markets, I'd say that setting the price for a product almost always boils down to 2 general concepts: market power and price elasticity.

Market power comes from producer theory and, generally speaking, refers to how much control you have over the price of your product. One way to think about market power is to start by thinking about the monopoly case. A monopolist, by definition, is the only firm in the market. A monopolist maximizes their profit at the quantity, call it Qx, where [MR = MC] OR [dTR/dQ = 0]. However, unlike a firm in a competitive market, a monopolist sells their good at the price associated with Qx in their own demand function (which for a monopolist is the market demand). Geometrically speaking, in a less-than-competitive market the "spread" that exists in the space below the demand function but above the marginal revenue function, and over quantities up to Qx, is semi-technically where market power can exist.

There are a couple different types of price elasticity. The most commonly discussed one usually refers to a product's own-price elasticity of demand (EoD), which is a measure of how much your quantity sold will change if you change your price by a little bit. Or in other words, it refers to how sensitive the consumers if your market are to changes in the price. For start-ups who have competitors, knowing your cross-price EoD would also be useful. If changes in your customers income in some way affects your business, then you could even think about the income EoD. Their are also types of elasticities which are meant to measure how sensitive consumers are to prices as time changes. Microeconomists and industrial organization economists have done a lot in this space. On another note, Bing Travel and their price predictor is a good example of some really sweet data on consumers temporal price elasticity of demand that I would love to get my hands on.

At the end of the day, a model is only so good. But models aren't meant to be absolute truths, their meant to guide your analysis. If you're product is something that no market currently exists for, then read up on the topic of market design. Al Roth at Harvard has some really great material and is perhaps the most influential market design economist to date. In my opinion, the most interesting ideas, concepts and theories from economics all stem from microeconomics. Current macroeconomic theory is a scary, scary place somewhere in a deep, dark hole. For a funny video on the matter: http://gregmankiw.blogspot.com/2012/01/home-for-holidays.htm...




Applications are open for YC Summer 2019

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

Search: