Ask HN: How do you set prices?
I'm helping a friend's startup figure out pricing for their products. What are some tools/services that you use to do this? Are there any pricing analytics platforms that you'd recommend?
I can't find anything that wasn't enterprise focused ("Call for pricing!"). Are there any services that your company uses that you're happy with? Ideally, it would help calculate common metrics & make suggestions for pricing experiments (& help run them!).
182 comments
[ 3.0 ms ] story [ 220 ms ] threadSome general considerations are your market strategy, your financial situation, your customer's finances, pain that you would solve, whether your solution results in growth or strategic advantage vs. back-office savings, ... and neither last or least, the emotional wins of the buyer.
Another resource with lots of actionable advice: A Gigantic List of Psychological Pricing Strategies: https://www.nickkolenda.com/psychological-pricing-strategies...
# The Sequoia Guide to Pricing https://www.sequoiacap.com/article/pricing-your-product/
https://web.archive.org/web/*/https://www.sequoiacap.com/art...
Of course, once you achieve monopoly status, you can set the price at the intersection of your marginal costs and marginal revenue to maximise profits.
In practice? Pick a random number that doesn't sound too stupid. Try raising your price. Keep doing that as long as it works.
Why this is in Theoryland is because you cannot adjust prices and get feedback so easily in real life. But hey, the logic checks out!
In a closed system. There are too many factors obfuscating the landscape in real markets (most notably consumer access).
The only tidbit is, in a new market you can overcharge.
On the one hand, yes, but on the other hand, machine learning gives us a powerful set of tools to analyse these factors, which we could use to come up with a model that learns from the market to predict consumer surplus (but you're right about the naive economist approach).
When I ran my consulting business, I was always afraid to raise the price. I discussed this fear with one of my customers, and he told me that I was too cheap, and that I should raise the price by 50%. He depended of the service, and wanted me to stay in business. I did raise the price 50%, and I don't think I lost a single customer.
To answer your original question, I don't know of any specific tools for pricing. How a company decides to price a service tends to be pretty confidential. It's also normal to base pricing on how much competitors charge. For something new, it is tough. Get customers interested, sell it for more than it costs to make (before you run out of money), figure why customers like it, and eventually you'll have a pretty good idea of what to charge.
*this is probably less than you might think; at low volumes you can run a for a week then b for another week.
You could call the lower price "on sale".
A "randomly selected sale participation" note.
If someone is unhappy with their purchase and lets me know, I nearly always give a refund (where possible; the App Store prevents this). Sometimes I've given my software away for free if someone asks nicely.
I have never had any backlash to running pricing A/B tests.
About once a year I will run prices at 0.5 and 1.5 the normal price for a week or two to make sure I am still on the right point of the demand curve. Sometimes it moves and I will stick with a new price.
Is random discrimination fine? I'm genuinely asking. I don't know the answer, but somehow it still feels morally shaky to me or unfair.
Yes, we do give refunds or discounts retrospectively in many cases, and try to be nice and fair and open. But there's just something a bit awkward with randomly charging people more (or less) for the exact same product, at the exact same time. I think most people will feel manipulated if they discovered it, even if they eventually paid the lower price.
It's a fool's errand at low volumes. Whatever you find out isn't statistically significant.
Depends on what volume you are measuring.
In my case I have an abysmally low conversion rate, so I can in fact have a low volume of sales relative to the number of visitors and still have statistically significant results.
First, set up a survey on Google Surveys (https://www.google.com/analytics/surveys/) and select the most appropriate audience for your concept. Set the first question as a screener of who you think this product is for. So if your business wants to sell to pet stores, you might set the audience to "Small Business / SMB owners" and have the screener question be "Do you own a pet store?" and screen out anyone who says "no."
Then, briefly describe your product and ask a straightforward question about how much they would pay for it. So, something like "What is the maximum you would pay for a service that handled the logistics of mailing pet food to your customers?" Then make the answers to that question your possible price points - "$9.99/month", "$19.99/month", etc. Make sure to include a "I would not pay for the service" or "$0" option - this is an excellent gauge of whether or not your service is actually something people will pay for. If you run a pricing survey like this and 95% of people say $0, that's pretty telling.
When you look at the results, you'll see a clear curve from the higher prices to the lower prices / not interested option, but you'll be able to see what a relatively targeted group would pay. So if 40% of people would pay $9.99/month, and 10% would pay $79.99/month, that tells a story you can interpret into a basic pricing strategy.
Depending on your budget, run 3-5 of these with different prices, different pricing anchors, different wordings, etc. - get as much data as you can.
I used this method for my last startup, when we were trying to figure out how much a specific niche would pay for our product. The pricing research we did through these surveys led us to a conclusion of about $29 per product, which was actually much higher than we had anticipated (we were going to sell it for $9), so we priced it 3x higher than we were going to. Very long story short, we made the right choice - people bought it and we had very few complaints about pricing too high. We even raised prices eventually after adding new features.
Now obviously, there is a delta between what people say they will pay and what they will actually pay, but this method might help get to a starting point, or add a layer to your existing research.
Admittedly, I do this sometimes when I have random dumb ideas for companies and I want to see if there's a market for it without really committing anything.
At the same time - never forget that a lot of the "usual recommendations" are wrong :)
Building an enterprise-oriented product, pricing is really a baffling mystery. Especially since I can't find a good single-factor usage control.
You can present different prices through A/B testing and see which one is more profitable.
Then, devise a way to make sure all features pull their weight. Get rid of the features nobody uses or improve them, but don't invest on development and maintenance of features nobody needs (feature creep).
Feature creep translates directly into unaccountable product people and software rot. Software rot translates into checked out engineers that either hate their job or don't care about the project, and heavy/inefficient organizations.
The closest thing to a "methodology" I've found is asking these four questions and having the users generate their own pricing curve [1]:
Here's a real set of curves this process generated for me recently: http://imgur.com/lPKLk53 ($ values redacted)
The four questions are:
1. At what price would you consider [the product/service] to be so expensive that you would not consider buying it?
2. At what price would you consider [the product/service] to be priced so low that you would feel the quality couldn’t be very good?
3. At what price would you consider [the product/service] starting to get expensive, so that it is not out of the question, but you would have to give some thought to buying it?
4. At what price would you consider [the product/service] to be a bargain—a great buy for the money?
Require a specific $ amount as the answer to each question.
Take ~100 users, ask them all four of these questions, and then compile the results. You really have to do this on a subset of your own qualified potential customers to get any meaningful data.
The neat thing about this is that this creates a price sensitivity curve without anchoring the interviewee with any prior numeric values.
At the end of the day, it's still a gut call about where to place your price point relative to the user's alternatives, and testing is encouraged, but these four questions are a decent start.
[1] https://en.wikipedia.org/wiki/Van_Westendorp%27s_Price_Sensi...
But anchor yourself. You should certainly know the prices of your competitors (or the closest thing thereto).
Rather than asking people, if you can, try just changing the prices and see how customer acquisition changes.
I sell a program online to consumers, so for a while I tried several experiments in pricing, including a "pay what you want". Structure. Eventually I settled on a price that's about half what the highest accepted price was ($10) and the typically "pay what you want" price ($1). Right now I sell for $4 (perpetually listed as 50% off of $8), which brings in about as much revenue as the $10 price. The advantage, though, is that it gets in the hands of more people, so there's a higher potential for word-of-mouth advertising.
So, I would say, if you're in the position to do so, experiment with the prices in the real world and look at how purchasing behavior changes.
This is a very good point and something I haven't considered.
On the flip side, depending on your type of product, I think more customers could lead to an increase in support burden to the point where it is costing you more than the amount of extra dollars you get from the additional amount of customers.
Eh? So if I told you I had two products A and B and I showed them to potential customers who on average said they would pay $100 for A and $10,000 for B, you wouldn't be willing to bet that when I actually take them to market B would end up retailing for more?
You just compared two (purely hypothetical) things someone said they'd do, you didn't address what @Osiris was talking about, which is the discrepancy between talk and action. The comment is correct, there is a lot of evidence out there that people do not pay what they say, that actions and talk do not correlate very well. So it would be interesting if you had some counter evidence to back your rebuttal, because theories based on logic alone rarely survive contact with actual human behavior.
This is essentially the best advice you can get.
IF you have to be safe because you're e.g. in Germany and you can't just change other peoples contracts after they signed up. Just keep the old contract/subscription types and then offer them a migration path sometime in the future.
Isn't it the very nature of contracts that you cannot change them unilaterally? I'd be very surprised if there are places where you can legally do that, and I wouldn't call such a thing "contract".
You likely can lose chunk of customers because of big price fluctuations..
Illegal in the UK IIRC
Also, those are customers that will not buy your competitors' product.
I have purchased both TextMate and Sublime, for instance, and now use Atom.
Isn't that quite dishonest? I'm fairly sure this is illegal in at least the UK, so I'd assume in plenty other EU countries too. Are you in the US? Is this practice legal there?
Of course, without deep pockets, a startup is also unlikely to be targeted for this.
1: http://business.time.com/2012/06/11/can-you-sue-a-store-for-...
>Isn't that quite dishonest? I'm fairly sure this is illegal in at least the UK, so I'd assume in plenty other EU countries too. Are you in the US? Is this practice legal there?
That's weird, are people in the UK not able to see through marketing gimmicks?
Really you could consider it a subset of truth in advertising law though. How can it be a 50% discount if it's never been sold at full price for a significant amount of time?
After I get ~50 - ~300 users, I tell them I'm going to set pricing. I usually do this with a pop up or email. They then have a drop down with "how much would you be willing to pay monthly for this service"
Usually you get some range, say $5/month to $100/month. I usually pick one standard deviation above the average. So in this case, let's call that $60/month.
Then, I usually will A/B test pricing by changing it month by month with "sales" so my current customers don't get too angry.
I've done this three times on my products and several times for others. Thus far, it's worked very well. You don't always want to capture the most customers to make the most profit. This method has you start at the top of what people would be willing to pay and work down.
Plus, you start with a user base for free that share you with their friends. Seriously, this has been the best way to jump start the business.
i think someone else in the thread made a comment about 'pissing off' existing customers, but a 'sale' is the perfect solution.
1. In a way they address a different psychological need for people. It's about getting a good deal or a discount over the "normal" price, rather than considering the price itself... So you're testing different things that in some way conflict.
2. Over time, if people see there's always a sale, they might hold-off from buying?
3. It just gives a slightly "cheap" reputation (think shared hosting as a rather extreme example that comes to mind, or any site that always has tons of coupons on coupon-hunting sites).
For your scenario, you need to start at a higher price point, no?
I just released a product recently which HUGELY undercuts the competition. I was able to build it cheaply, over the course of a couple of years, and as a result I never took investment or hired anyone, so I can offer the product at incredibly low rates (compared to the competition).
My worry is that I may be pricing it so low, that I'm actually scaring off potential customers. This is an enterprise product, so they're used to seeing massive licensing fees. I initially thought that offering it at the lowest price I could afford, would mean I would garner the most customers, but now I'm starting to wonder if that's true... I have no experience pricing things, and I'm really considering putting out a survey to existing customers.
So yes, worry about scaring off potential customers with a price too low, but don't worry about scaring customers with a price that's too high, unless you scare off proportionally more than you increase the price. The way you describe it sounds like you could make way more money by raising your prices.
Now there is a memo that the Silicon Valley web startup economy didn't get...
Having run and advised numerous companies over years, increasing prices is one of the best decisions a startup can do. In 90% of cases there are no ill effects and just increased income which helps buld a better product. Profitability FTW.
[1] http://venturehacks.com/articles/pricing
There's a school of thought that says that you should only achieve a certain percentage of sales. So if you're getting 90% of deals then you might be pricing too low, and if you're not getting anything then obviously too high. But if you're making 10% of sales and your selling for 100x more than your guess at the price then you're likely doing well.
There probably isn't one method that works in all cases.
5. How willing are you to undercut yourself, due to financial desperation?
Out of that question will pop out a selling price.
Start with two price points: one targeted at price sensitive customers, call it price A, and one targeted at price-insensitive customers (think enterprise), call it price C.
Set price A lower than you think - you can always add features and raise later to the point that people start to scream.
Set price C at much higher than you think - use volume discounting to offer discounting where needed, but this is your anchor point for conversations with big customers.
Consider whether price A is your acquisition channel or if you will have a "free" tier. If you don't have a free tier, you can feel fine with a lower than optimal price for tier A, as this is your conversion channel to price C.
Eventually you'll want a tier between A & C (call it B) to anchor pricing and encourage people to choose A or C. Refer to research on movie popcorn prices.
Set price A higher than you think. It is easy to lower the price later or to give -30% to customers and have them think it's a bargain. A higher price tag will confirm that your product is valuable and useful and you've got a market.
If they don't want to pay, you need to make a better product, or sell harder, or market it better. You don't want to fight for cheapness, cheap customers are cheap, annoying and they won't cover the costs to run the service.
Then again, we didn't actually try raising it and face backlash. It's possible we were just too cowardly to try and fail.
You have to remember that most industries are very incestuous with people moving between companies within that industry. So what happens if someone from a price-sensitve customer (price A) moves jobs to a price C company and tells the new company the low price that their old price-sensitve company was getting? Price C company will demand that price (as I would) and then you're in trouble.
I make things simple. I charge the same price for everyone with discounts for volume. That's it. Nice and clean.
Even if not, some great advice here! Thanks! ...but I think most of it is B2B focused primarily. I'm curios (and can't find much info) about subscription pricing for consumers in particular... And especially for products that aren't as wide spread as Spotify or Netflix.
They recommend Amplitude if you can't build everything yourself. They say it lets you have cohorts and compare behavior between cohorts, so it probably helps compare common metrics. I don't know if it makes suggestions for experiments or helps to run them.
https://www.amazon.com/Strategy-Tactics-Pricing-Growing-Prof...
What I learned was invaluable. I do have to warn you, the book lays it all out. This is excellent, of course. However, what happens is your level of confusion as to how and why to price using a certain approach will grow as you progress through the book.
Somewhere past the middle things start to coalesce and your choices become clearer. Again, this depends on the nature of the product.
Please note this is not a critique of the book at all, it's excellent, this is a complex topic and it is only natural to be confused before reaching clarity.
Example: How much money would you be willing to pay for a lead list that could make you $100 in sales? Assuming you have no internal cost to acquire (for sake of simplicity), if you paid $90 for that list, you’d make a 11% ROI. What sort of ROI do your customers need in order to bite?
Call all your competitors and pretend to be a customer. Get quotes from all of them.
Position yourself based on their pricing. If you're trying to be high end, charge more. If you want to be seen as affordable charge slightly less.
customers make a lot of conclusions (right or wrong) based on price.
same reason 'healthy' things are priced higher, even if their cost would be lower than the comparable 'unhealthy' thing.
No lies necessary.
to find out what they are willing to pay, ask for budget.
If you are experienced and have a strong vision on the service you are offering and how it should be priced, you can think about premium or non-traditional pricing schemes.
The list of things you can do to repackage your product to go after different market segments is very long and very few of the bullets on the list involve changing code. I think the most important thing to do is to select a group of customer prospects deliberately and then make something those people want, rather than making something for the whole world and trying to optimize the number of people who buy through prices.
You can cost yourself customers/clients by pricing too low.
We've been looking at pricing services suggested in this thread, and we've been self-selecting away from the extremely expensive services (e.g. Price Intelligently [1]), as that indicates that the service is targeted towards large enterprises, and not small SaaS companies.
Conversely, if the services were priced lower, that would probably scare off large enterprises.
[1]: http://www.priceintelligently.com/
There are lots and lots of cases where competitors have been in a niche for a long time and they now use aggressive discounting to experiment with (what's left of) customer acquisition. You price below them, you will die starving. Also, it is dramatically harder to raise prices than to lower them, so starting low and then increasing is a very painful thing to do.
It depends on your perspective. When you're small and growing it's actually the opposite.
If you grandfather your existing customers (ideally, indefinitely), then it's quite easy to raise prices. Assuming there is still demand at a higher price range, which in many cases there will be, since your product gains traction and improves over time (so you can also justify charging more).
As your customer base grows, it's also harder to lower prices. Would you cut the cost of all your existing customer base? That immediately cuts into your recurring revenue, and you don't know if it's going to bring more customers to balance out over time.
In my opinion, the hardest part of the problem is getting away from the psychology that pricing "uncomfortably high" is somehow cheating people. It isn't - your product is more often than not more valuable than the sum of its parts.
1. Set up Google Ads for your product to drive some traffic to an hidden web page
2. Set up A/B testing for your product with a range of different prices. I set up six web pages with prices ranging between $5 and $100.
3. When users click 'Buy now', have a page saying that the product is not available but record the number of people who clicked it.
4. Calculate which setup generated the most income.
Are you collecting their email's, sending them to a page or social page to check back when your product is available?
Someone picking a higher price, and you end up pricing it lower that sounds like a plus. If someone picked lower then you actually price it, wouldn't they have been false advertised to? Unless they actually knew it was a survey up front then that seems more fair.
This actually totally broke one startup I knew (that was trying to build an API for ordering products from loads of ecommerce sites).
People like to think they're being clever by A/B testing their prices. The fact is enough people are smart enough to figure out that their "A" slot is worse than the "B" slot you've given other people.