You can spend as much as you want on developing a product, but no one cares how much you spend – they only care about it’s value. You can’t sell a product until it’s built and usable and the developers need paid to create the product, so how do software business make money if they are out of pocket before they even start? In this talk Rich gives you the key points to remember throughout your journey from startup to success.
Rich Mironov – a BoS conference regular having spoke about Product Roadmapping in 2018 and Software Economics in 2015 – is a seasoned executive and serial entrepreneur; having been CEO/VP Product at six startups. With roots in B2B infrastructure, SaaS and consumer online, Rich combines ‘what-we-can-build’ with ‘what-markets-want’.
Rich is returning to BoS in 2020!
Get Weekly Advice From Software Experts
Fresh talks on entrepreneurship, product, marketing, leadership, hiring, and more dropping each week. Upcoming talk releases include:
- Jason Fried (CEO/Founder, Basecamp & Hey) on launching Hey
- April Dunford (Author, Obviously Awesome) on why you should stop selling your product
- Dharmesh Shah (CTO/Founder, Hubspot) on facing fears as an entrepreneur
Don't miss out - sign up now:
Rich Mironov: Thanks very much. Let me do a quick introduction of who I am and why I don’t normally spend my time on pricing – I guess that makes me smarter than some.
I’m a 40 year Silicon Valley veteran of enterprise and B2B software; last 30 years of that on the product management side. Only reason to say that is that product management isn’t something new, but it hasn’t arrived everywhere at the same rate. So for those of you who don’t actually have a product team and just have engineering and sales you’ll know where to call me later. And much of my work I coach heads of products around the world and I drop into San Francisco area companies as the interim or temporary head of product, VP of Product, particularly when they’ve forgotten to have one! So of which there’s a lot lots of startups that didn’t really do this before and it’s their first time. One of the things I notice often when I show up to these companies that have forgotten to do a lot of planning work and the product work is that pricing is something that happens about a week before we ship. Right. So we can’t actually take orders for something unless we assign a number to it. And somewhere that didn’t make the checklist. But before we can print our price sheets or put it on the website we need a number. And for me this is a fundamental failure of strategy if we think back to Derek yesterday was saying “we have to know how the game is played”. Pricing is the long game. Pricing is slow and complicated and ugly but it’s all about how we get our customers to buy the right thing and understand and see value. So we’re going to dive into that – I’m hoping there’s (finds what he’s looking for) ah good.
All right. So here we are we can either take the psychic approach and the ancient Greeks would have sacrificed a small animal and read the entrails to figure out what your pricing strategy was. Good. So here’s the four points we’ll make in case anybody can stay till the end. I’m going to talk very briefly about.
How software companies make money and that’s really important because before we start to assign a pricing number to something we have to understand which of two business models we’re in so we’re going to put up the professional services model where we charge for our time and this business of software model because that’s where we are where we charge for value based on the software we built. And honestly nobody cares how hard it was to build it or how many hours or how much money we spent because we’re pricing on value.
So we’re going to take one of those two choices and then if you if you miss everything else we’re going to talk about how software should be priced based on how much it’s worth to the people buying it, not how much we want them to think it’s worth it nobody cares about your business plan but you
Units matter everybody’s focused on the number. In general the number is much less interesting than the choice of what we count. So we’ll do a couple of exercises there
Then we’ll put up a couple of charts that you’ve all seen many times before because it’s how all good online or SaaS software is sold in these small/medium/large or bronze/silver/gold whatever your favorite thing is – because it makes life easy.
Ready to go? Let’s go.
How we make money in the software business
Okay. So the first thing for me is about how we make money in the software business. This is I don’t know if this is still valid here – we’ll see after Brexit what kind of currency we’ve got, right! What’s important here is going to be we’re going to separate the problem of what it costs us to build software from how much we charge.
So if you take your typical team and I’m going to use silicon valley prices because now we know that remote teams get paid based on what engineers in San Francisco get. Good for you. But a team you know call it a scrum team you’ve got six people you’ve got eight people kind of regardless of where they are you’re going to put that team together developers, test engineers, dev ops, docs, whatever it is. And your typical team costs you per year? Anybody? Roughly a million pounds a year. Now maybe a little less maybe a little more. We’re going to spend that money on the team whether they build something useful or not because we have to pay them. So, on the cost side every team that you add to your company is going put whatever a million pounds I’m told today the dollars and pounds are now one for one – is that true?
So you’re going to spend your million dollars a year on a team. It would be really nice if they were building something useful. More than that if you look at any medium to large software company what you discover is most of the money is not spent on building product. It’s not on the development product and dev ops and documentation side. The majority of the money in medium to large software companies is on selling and marketing and other things: legal, real estate, fast cars for the executives, whatever it is. And in fact a typical large company if we looked at Microsoft or Oracle or one of the really big players you’ll find they spend about one dollar in six on actually building this stuff and five dollars in six on selling it, marketing it, covering their legal costs, and paying something to their shareholders. So therefore if we spend one million pounds on building it we’re really making a promise to our shareholders and our customers and our financial folks that we’re going to bring in roughly six million this year for having spent one can you get to scale this anywhere you want maybe in your company it’s five but we’ll take six because it makes the math easy. And that’s really important because the idea that you’re going to spend a million building it and you only need a million back is fundamentally wrong if you’re in the product business.
So we’re going to take six because that’s a good number as any and say Don’t spend your million unless you have some plan to bring six million in or you’re diluting your shareholders and your CFO is going to take you out in the back and there’s a whipping to be had. Okay. But the great thing about software unlike physical hardware especially big iron stuff is the incremental cost per unit. If we’ve done a really good job of engineering away all of the friction is nearly zero. We can add one more user, one more transaction, one more customer, one more online commerce thing, whatever it is that we’re selling we’ll get to units in a minute for not very much money because the bits are built; because they work because they’re in the cloud. We can add a little more AWS and it’s all just fine if we have a sales force and I hope we do we have to pay them some commission that’s probably most of this 3% but we should be engineering away all of the other costs to adding another user to adding another transaction and that’s really really really important we’ll see in a moment because in the software business we’re not trying to make each unit as pricey as we can. We’re trying to get a lot more units right.
And so the goal here isn’t to minimize the cost of developing and I’m going to come back to armor in a second here which is if you think of building product as a cost center you want to have cheaper people building stuff – which is, by the way, a great strategy if you don’t actually care what you’re building you’re not trying to make money on it – but if what you’re trying to do is build something that the world will love and buy a lot of. The goal is not to have the cheapest designers. The goal is to have the best designers so everybody gets in the workflow and get signed up and uses the thing and it’s wonderful and it’s great. The chief financial officer who wants us to outsource and send it to someplace less expensive because we’re going to save on engineers doesn’t understand the software business. Because the way we make money is we sell a whole lot more of the bits that we’ve now built and we collect almost 100% almost 100% margin on each incremental one. Our goal is to build something great that the world loves not to cut costs.
Good. So here’s here’s the first takeaway really important right:
in the software business nobody cares how much you spent on the first one and we don’t expect to make money on the first one.
We don’t charge to our very first customer 140% of the cost of building it, if we expect to sell a lot of this stuff. We’re going to sell in value we’ll see how to compute value in a minute but we’re going to sell the first copy for a lot less than it cost to build. Then the second copy for a lot less than the third copy for a lot less. But at some point we hit break even. And once we’ve hit break even. It’s all cash and it’s all profit and it’s all money and that’s the goal of the software business. Here we are at the Business of Software.
So let’s do the quick contrast here because there’s two business models I mentioned and I had a chance to talk with a lot of you yesterday who are in what I would describe as the professional services business, right. In the professional services business. You build software for one client or customer you build it to the specs they give you they pay you cost plus a bunch of margin and it’s theirs. And by the way if they ask you to build something that’s useless and worthless as long as they pay their invoice. Honestly you don’t care. Because in the professional services business you’re building one at a time custom bespoke whatever your favorite word is. And when we draw that business model it looks something like this which is you’re always profitable as long as you keep your folks busy. The slope of that curve is how much markup you can make on each hour or a day or week of the folks you are basically renting out. So the professional services business is a technology rental business where you have really smart folks and you charge your client more than you pay them per hour. And so what we get in the customer development business is we are rewarded, we make more money we are happy when we keep our folks busy and we bill lots of hours. It looks also a lot like the legal profession and some other ones that may or may not be legal here. We get paid in the professional services business for billing more hours and doing more work. If we’re going to get paid for them. And so one of the things we love in life more than all other is the client who changes his mind a lot because we get to respond with a thing called a change order that says “Yes. Okay you can have whatever you like that’s going to be another £80,000”. We’re pricing for effort. And so all of our clients want to compare our cost of effort against all the other competitors and they would like to have cheaper developers because many of them actually don’t know what good looks like. They only know what cheap looks like. And we’re in price competition for the hours that we bill. It’s a perfectly good business model if that’s what you’re doing and you should know.
Let’s draw the product business model the software product model. First of all you’ll notice that we have to spend a bunch of money and we are in the negative at the start. We can’t charge for anything until it works. We have to build it before we collect money that may or may not seem obvious. This slope though has nothing to do with ours or time or people it has everything to do with what we’re selling: subscriptions, transactions, units (we’re going to get the units in second) but we make more money by selling more seats, by selling more brokerage transactions, by selling more e-commerce. Nobody cares how hard it was to build. Nobody pays us for our work. They pay us for using the things we’ve built and that slope turns out to be a lot faster/higher/steeper because once we’ve built the bits as we remember it once we cross pass a pass break even here it’s all money because we built it and we’re ready to go. So in the product business we get paid for large numbers of customers or users or transactions making them as frictionless as possible so we can get lots of folks in. And then having them like what we did. Again having cheaper engineers here is not a strategy for winning. Having better engineers is a strategy for winning wherever it is that they happen to be good.
So this is really important because I’ve worked with clients who are in the product business and are very successful and clients who are in the professional service business who are very successful. I have been unable in my half a dozen tries to help the clients who are half and half. This is a really really really terrible place to be. And I do a lot of marriage counselling with the executive team because half of the team wants to sell whatever this next customer wants and half the team wants to sell the thing we’ve built. And that’s a lot of shouting That’s a lot of extroverted shouting in the executive room.
Pricing starts by computing Customer Value
Okay let’s keep going. So now let’s talk about value because now that we’ve kicked out from under you the idea that you’re going to price based on how much you’ve spent and how hard it was to build because nobody cares. We’ve got to figure out how to compute what customers think something is worth. Here’s your £50 note from yesterday. Thank you Derek.
And I’m mostly B2B so we’re going to mostly answer the B2B question which is why do people buy our stuff. They either buy our stuff because it’s going to help them make more money, or it’s going to help them save some more money. And by the way it would be handy to know which one because apologies in advance for all those liberal arts folks we can do some maths in the next few slides. Simple math. Good. But most importantly we need to understand not how we think our customers value something. And Randy took us through this yesterday a great extent. Thank you – spot on. If we haven’t and if we haven’t figured out how our customers assign value to the thing we’re doing let’s do our homework before we start building it because if we spend that million pounds and we don’t actually know how customers value what we’re doing we’ve done things out of order. It would be really handy to have a thumbnail business case or the back of a serviette with some numbers on it than to spend the million pounds building but not have figured out yet how we extract value. As a product guy, I’m trying to be out way ahead of the development curve in case this is a bad idea and I’ve had many of those bad ideas. We like to cancel that bad idea before we spend the million pounds, not after we build it.
So how did they describe value which we can only get by having listened to them really carefully. And I believe in at least in the business to business space we should be able to capture something between 15 and 20% of the value once we can figure out what it is from them. Why 15 to 25%? First we’re going to describe value to them and they’re going to knock it in half because they know we’re a vendor and we’ve overstated it. And then they still need to capture at least half the value or it’s not worth investing in my product. Good. Let’s keep going. Right. Consumers is a whole different thing we’ve got comparisons with similar products, we’ve got high end value where people want to show that they’ve spent too much for it. Consumers are a lot less rational. And in general there’s a pretty small decision cycle here. So I’m going to stick with the more rational more hard bitten B2B enterprise side where I live. Okay let’s keep going. So here’s our Barclaycard. Let me I’m going to tell you two value stories just to set up the framework and it’ll turn out not to matter what the technology is so I won’t tell you what it is.
But imagine a situation where I have some software company and what we do is we sell something to banks, we still get paper mail here with invitations for credit cards? Endlessly! Good. OKAY. So endlessly the Barclays of the world are going to be sending something into your mailbox which is pre-approved for a credit card for only those of you who they think should have one. And so we know that they do a bunch of credit checks in the background before they choose to mail this to you to find out whether you are worthy or not very good ones right. Good because I’m still getting these and so we’re going to have some mythical application here and it actually doesn’t matter how it works we’re just going to run the numbers that says somehow my magical piece of software for the Barclays of the world will reduce the number of credit checks they need to do before they flood your mailbox with pieces of paper that you throw away to not take credit cards. So, our customer here is the bank. So, of course we need to have a story but the story is a numerical story because most banks only care about making money and so we’re going to have to talk about making or saving money. And so the generic formulation here looks something like this.
Your bank Lloyds whoever it is does some number of these credit checks every year. Insert number. (You know this is a spreadsheet. But right now it’s a slide.) And they pay some amount for that. And if we’re clever we actually know what that amount is because we went and figured out who the vendors were. We’re going to reduce that by some amount. Otherwise we haven’t earned our keep. And if you multiply those three things together I can tell you how much money your bank is going to save because we’re going to reduce the number of credit checks by 14% and you pay 51 pence a piece times a million a year. This is the basic piece of work that we must always have as product folks to give our sales team or we can’t explain to our customers how they’re going to save money.
Totally generic there’s one piece missing. What are we missing? Price! That’s right. Because we’re in the business of software, because we’re in the business of software we would like to be paid and we have put that yes so the paid looks something like this we’re going to only charge you a certain amount which ideally is substantially less than how much they’re saving or we don’t have a game here. And then if you divide those two things the chief financial officer can see what the return is on that investment. Every business to business piece of savings logic looks like this. Plug it in whatever the numbers are. If you can’t answer this set of questions before it’s time to price your product you haven’t done your homework. Go see Randi about interviewing a bunch more folks. So let’s do one just with the numbers so we can see what that looks like. Here’s a similar one. I have some magical piece of software and it doesn’t matter what it is or how it works. That if I gave it to your support organization that takes phone calls from users with problems somehow it’s going to make those calls get finished – how much – 30% faster. Right. Maybe it looks up answers better it’s machine learning nobody cares right. There’s some activity here where our magical software is going to make your tech support phone team 30% more efficient. Right. We need numbers so I made up a bunch of numbers here’s the numbers.
This is what we need to know we need to know how many calls you take every year and maybe how many hours that takes. I’ve imagined that you’ve got a support team of four people and you pay them £40,000 a year – I don’t know why – so we can compute for you that you’re spending £160,000 a year answering the phone for tech support right. What’s the next line? Saving! So I’ve invented my 30%savings somewhere and here we’re going to tell you that I’m going to claim 30% savings and therefore. Your savings as my corporate customer be to be enterprise is going to be about £48,000 a year so. Tell me what this product costs. Twelve? Right. So it’s going to. I’m going to take 20% just for the average right. So 20% of £48,000 is what £9,600 call it £10,000. Notice that I didn’t ask what it took to build. I didn’t ask how many engineers I have and how many smart people in which countries I have. If this is worth £48,000 to my customer then I should probably be charging about £10,000 or £12,000 for it. More than that they won’t buy less than that I’m leaving money on the table. It’s all about value. Now we’ve also noticed that if I’m a bigger customer I should probably pay more. And so what units we’re going to use here for instance we might do this per seat but if we save them enough money then they’re going to have fewer seats. We might do it per call or per look up. The units are important here because we’d like to match our units to the way they measure savings because that’s the language they speak. Good so far?
All right. Again purely generic. If you’re selling business to business and you can’t compute this then you’re either depending on having really smart customers and I don’t recommend that as a marketing strategy. Or your sales team is going to invent this for themselves and often they’re not the strongest maths folks in the room or you’re going to depend on somebody to do it. As the product person in the room, I feel an obligation to supply this to my sales team so that we get it right. And again as a product person who may not have done this work upfront if we’re coming up on launch time and I haven’t done my homework. Shame on me. Okay let’s keep going. All right
Units are important
Units are important. I don’t know which side anybody’s from you can be on either side you like it’s useful to know what units your clients or customers measure their world in. If you’re in the hotel business the unit of economics of a hotel is what they call bed nights. It’s not just the beds we have to sell those beds every night and so what they look for is utilization of bed nights. How many are being slept in tonight and anything I can do to make their bed night score higher, they value directly because that’s how they keep score. If I’m in a university the unit of work is probably students per semester or term or trimester or whatever unit of time is. That’s how they measure things. So we might want to price our product or in the automobile assembly game. It’s usually numbers of cars per minute that roll off the assembly line. So some if I can unitize my product and saying well I can help you get one more car per hour off the assembly line. They immediately know what it’s worth instead of per seat because honestly they have no idea. Right. If you’re in a solar energy farm you’re probably measuring some kind of kilowatt or megawatt hours. And somehow our software better get you more kilowatts or megawatts because that’s how you measure life. High end retail shops and I know Debenhams had some bad news this week. So sorry for those folks who shop there
Mark Littlewood: No one does!
Rich Mironov: Apparently no one does. That’s right. The unit of work or the unit of economics for high end retailing where you’ve got some you know Knightsbridge store anybody no what we measure we measure it’s usually currency per floor area so it’s dollars per square foot in the US and it might be pounds per square meter here whatever it is. But the retail space is all about being in the most wonderful location and trying to sell enough really expensive handbags to cover it. Good. So knowing units is important the two default units the one the ‘All You Can Eat’ subscription. So this is the we didn’t think of anything else or we’re in the consumer business and so for $9 a month you can have as many movies as you can watch or for £12.50 you can have as much data on your phone as you want whatever the unit is right. The trick with all you can eat is you’ve got to drive lots of subscribers and you need to keep your costs low because the margin is everything you keep after your costs and you must put some limits and for folks who abuse volume. So if you watch too many movies or you use too much bandwidth or you do too much trading there’s got to be some limits and cut those folks off or charge them more. The alternative is a transaction based system. So if you’re buying shares or you’re doing e-commerce or you’re booking travel you don’t pay except when you do a transaction. And the challenge with transaction based is, as a vendor, I have to keep reminding you that the service exists because maybe next time you’ll book your travel somewhere else and then only make money when you do something. So, the continuous drumbeat of marketing is really important. Obvious. Okay, let’s keep going.
So this is a chart I’ve used every place. Does anybody know this chart? This is all SaaS products in the world. That are well packaged. So there’s a basic and expanded and advanced or there’s a bronze and a silver in the gold or a small or medium and large. Notice this is not about volume. This is not how many seats or how many transactions. The goal here is to say some folks are going to get the basic product and some folks are going to pay extra to get more storage better support features that aren’t available extra downloads conversions whatever they are. And then there’s usually an advance version that this is maybe the most important one right. So this is where the CEO invites all of the big best customers to golf together whatever it is. Right. Marketing makes this chart look beautiful by having all the columns be even there’s only one important thing here which is as the product person I need to know which features which couple of features are the ones that some people will pay extra for and other people don’t need . This is called segmenting. So I have to have a really clear idea of who in my audience is going to pay extra for what so I can put them into nice bundles. Everybody loves this because it’s easier to buy, it’s easier to sell, it’s easier to support. We know what we’ve got. The alternative is sixty five options and you have to choose every single one yes or no and figure out what they cost. Not so good.
So if you don’t have a model thatt looks like this you’re probably making life really hard on your customers and your prospects and your salespeople. So good segmentation because customers come in different shapes. Maybe they don’t. Is to figure out which group of your customers can get away with the basic and how we’re going to know who they are and which of your customer segments are going to pay extra for which features and why. Again if you don’t know that question we have to go back to the learning and discovery and insights and go talk to another 50 or 100 of your customers because you don’t know how to do pricing yet because you’re missing the data.
A great segmentation and pricing scheme lets a bunch of incoming prospects look at your website and see if we go back to this. So this is on your Web site somewhere and at the bottom is some numbers and really good segmentation means that people can come to your website. Look at this chart, figure out immediately which column they belong in, and then when they call your sales team or maybe even you don’t have a sales team you have a sales form they tell you what they need. And by the way the price is obvious. We’re not haggling over additional features. They’re calling up saying I need your intermediate package because I need features five and six. Great. Let me write that down for you and send you an invoice. This is what makes frictionless really true. And so we really want to have segmented our audience carefully so that they fall into the buckets automatically and they recognize themselves in our packaging. They sort themselves: it’s got to be visible, it’s got to be simple. In the Internet age and I know I sent the telex out to all of you in 1996 that the Internet was going to be important because we knew this in Silicon Valley but the majority of your users and buyers whether they business or consumer are looking at the prices on your website and deciding whether to call you if you don’t have prices on your website they assume your overpriced, they pass and they go to your competitors and you never hear from them and you never meet them. So the idea that you can hide your prices behind a sales force that wants to chat on the phone with people. Is a fallacy it’s a fiction it’s magical thinking don’t do it. Other than that I have no opinion. Okay good.
All right. So briefly because I have two minutes and 30 seconds left. Let me tell you a short war story a real war story. I have a client who’s in the mobile applications for churches business by all of my clients of the world leader in something. This is my client who is the world leader in mobile applications for churches. Why is this a thing? I didn’t know was a thing. It turns out that in a lot of churches they pass a wicker basket on Sunday and people take their wallets out and they put money in the wicker basket which makes perfect sense if you’re over 30, and own a wallet, and know what money is. But the churches have a problem which they’re now recognizing which my client got way ahead on because they’re six or seven years into this. And so they have an app that they build out for churches that does two things one is it lets you donate money to your church and you know there’s a click on the back in the transaction there for them and the others it lists all the events for the church. When is the youth group and when is the trip and when is the Bible reading. If you’re under 30 it’s not a real event unless you can see it on your phone. So they built up six or seven thousand churches now that they’re using their app which is very cool. What they did though over the first four years was every time they brought out a new feature they listed it separately with a separate price on the price list so that they could amortize off the cost of building it know how much it was worth. So at the end of four years they have 70 or 75 features each of which is separately priced and you can choose them or not and none of their prospects can figure out what to buy. And almost none of their sales team can figure out to buy and lots of folks are getting combinations of things that don’t make any sense or don’t work. And it’s creating a lot of friction and unhappiness. So I work with them to create a shockingly three package model here. But we were struggling with what are the three groups because if the groups don’t recognize themselves then it’s not so useful. So we had to go back in again interviews and learning and insights and looking at all the data and figuring out. And what we figured out, and I’m going to run over a minute here, is there’s one group of churches that honestly you’re not very tech oriented and there’s just one pastor and there’s nobody to feed all the data stuff and they just want this simple thing because they’re done with nobody use it. And the second one you’ve got a church with lots of folks who volunteer to do social networking and they want all the features. And a third group you’ve got the big American mega churches that have many locations and God wants us to be rich. And they need a whole lot of enterprise looking features around delegation and a round allocation of resources and all this stuff. And so it’s completely obvious that we can put the three columns up to say well you’re either a really small church with a single pastor or you’ve got a bunch of people work on this or you have multiple locations nothing do with size and every one of those churches could come to that web page and immediately know which column they were in and what we’re going to sell them. That’s how you know it’s good segmentation.
Okay let’s wrap, so takeaways because there’s our takeaway container for those who missed it. Right.
- Customers define value. Nobody cares how hard it was to build. Nobody cares what you spent. It’s either worth it to them or they’re not figure it out.
- Keep it simple. Right. Pick a natural unit don’t have 15 kinds of pricing schemes.
- Ideally you want to have three packages – independent of size so you can upscale the folks from the basic to the medium to the large and then the last one is if you guys know this right.
- Good pricing is not interesting. It ceases to be a discussion and you get to talk about what matters which is the cool stuff you do and how you add value.
Good. So that’s me. Come find me some other day. This is really really important this how all bad software companies die. And I’ve seen them all good. This is me. Anybody who needs a copy of the book and doesn’t have it come drop me a note. I’m in San Francisco which is a suburb of Cambridge. Good. And I’m just going to throw us. Do we have time for one question or are we done?
Mark Littlewood: I think we are going to give you a big round of applause!