Mastering Market Research and Pricing Strategy

Automatically transcribed

Vanessa: Quick disclaimer before we start, this episode was originally recorded as a webinar and edited for podcast format. You can find the original webinar recording on our website at sawtooth. com or on our YouTube channel.

Now back to the show.

Dean Tindall: Thank you all for joining us for this month's webinar. Today we are going to be having the pleasure of the company of the indomitable professor, Tim Smith, who's the founder and CEO of Wcl pricing. He is also the adjunct professor of economics at DePaul University. He's the author of Pricing Done Right and Pricing Strategy. He's got degrees in physics, chemistry, and mathematics, as well as an MBA from Chicago Booth,

And he is also a really nice guy who really knows its stuff. I met him last year at the Professional Pricing Society Conference, so I know you're all in for a really good time. So without further ado, Tim, take it away.

Tim Smith: Come on. There we go. We're alive. So how's your check? Can you speak? Check yet, Dean, if you're going to Prague. Anyway. Yeah, let's go ahead, huh? No comment. Tim. No comment. It's one of the two languages I kind of know besides English, so I'm gonna go ahead and share my screen and get started. Share, replace current share?

Yes, sure. Share two. I think we're good. We're good to go now. So I'm gonna be talking about pricing and market research. Yeah, I've gotten in lots of trouble with this, but everything I'm telling you today is based upon academic research facts, not just something that I can sell, but facts about how these things work.

All right. Yeah. I got this photo taken in Prague. If you want to know where it came from, great guy. Roman. Roman Black. He'll take yours too if you want. Dean. What I'm giving you today is actually going to be a preview of my workshop. Not the one in Vegas, though, actually the one that's gonna be in Barcelona this fall.

I'll be talking more deeply about how this can be done. Price setting, price setting. Consider the value-based pricing framework. Price setting is just one part of it. All right, so the value-based pricing framework was released in 2016. It's been adapted by many companies across the globe and many different industries.

Think about the pricing questions. A good price management team addresses. We're gonna start over here at price execution. Do I have the right price on the, you know, shelves and the cash registers in the, invoices. Then I got price variance policy. Some people like to call this commercial policy.

I'm not gonna argue with them, but it's more than commercial policy. I'm looking at how my promotions are doing, my discounts are doing, my channel strategy is doing. It's looking at a lot. And then there's market pricing, the go to market price that you have. Now, I've worked in many different companies and I know that sometimes your list price may be 80 bucks, and that's true maybe on your website, but it has nothing to do with what you sell to a retailer that may be 40.

It has nothing to do with what you might sell to somebody who's gonna install it. That may be four. So, I mean, we have different parts. Market pricing that need to be considered and how that flows back into the rest. Then moving up another square, we got pricing strategy. Most people think about that. Skim, neutral, and penetrate.

That's part one. Part two is the structure of pricing. Am I doing unit two part tariff dying, tying arrangements good, better, best versioning, bundling index based pricing, yield management, subscriptions, consumption, or pay what you want or just demand supply match. All of these different ways in which we can price these products out there.

Oh, and the third part of it is what's my competitive strategy? We're living in a weird time where some interesting economic things are occurring, and so when your competitors raise your price, do you raise your price? If your competitors lower your price, do you, you have to think through the competitive response matrix as well.

All of this feeds up to the overall business strategy and how pricing can help the executives meet their goals. I've been doing pricing speedometers for three years, multiple times. I found 5 billion US dollars of value could be created by better pricing. So we're gonna try to create a little bit of that extra value for shareholders in today's meeting.

And yeah, pricing is a shareholder issue. So in market pricing, that's where we're doing our market research generally in that area. There's other parts, but this is where we're gonna use most of our tools for market pricing, and in some ways it's gonna affect the pricing strategy and the two are connected.

If you've been listening to the Hula about AI pricing, how do you price AI bots? You realize, hey, wait. They need to rethink their pricing strategy, their structure, and how those structures work. Before they go into the market research, but the two are related. So let's move forward. Anytime I ask you a pricing question, I'm asking one of three questions.

Actually, I'm asking you all three, what's my alternative? Are you better or worse? And do I care? See if you're better than the alternative. You can hire charge a higher price. If you're worse, you charge a lower price. But that's only if the customer cares. That's only if the customer cares. So wig left to tech meso American dude over here.

That's who he is. You may recognize him from another band also. He's asking these basic questions in our market research. We need to ask those questions too, from the customer's perspective. What do they perceive the alternative to be? What is the price of that alternative? What do they perceive the price of that alternative to be?

How do I relate to that product? Am I better or worse? And do they actually care about these points of differentiation or not?

Before I begin any pricing market research that goes to the market and tries to ask questions of people. I wanna have a hypothesis about what the answer is. The point of market research isn't just to ask the customer and learn, it's to actually have a hypothesis and test that hypothesis before you go to market.

So we need a hypothesis. The best hypothesis way of creating called economic value to customer. It's kind of qualitative. There is ways you can use research to improve it. It, but it is the basic tool of creating the hypothesis about what price you should charge from customers. If these AI people had tried to use this approach, they might not have made as many stupid mistakes as they made.

Anyway, start with the question of what is the economic value to the customer from their perspective with different segmentation at different times. You ask, well what is this economic value to the customer? We have, you know, a good out there an offering. It has some total utility. Price is gonna be much lower than the total utility.

Think about the price of water. It's really cheap utility, it's called life. Consumer surplus is usually so big it doesn't even matter. So you can stop thinking about that. I know we teach it in economics. Stop worrying about it. It's not important when we're price setting. What is important is the price of the alternative.

Hence, the first point of your parking pricing research should be to understand the actual price of your alt competitors and maybe the best competitors yourself. But that's fine. I need to know those prices, or at least what they are. The accuracy of my research is gonna be depend upon the accuracy of my knowledge of how my competitors are pricing.

So let's start with that because that forms the reference value the mind that the price that customers have in the back of their head and their dorsal stratum when they start to ask 'em questions. This is what they're thinking of. This is what they're thinking. The price of the alternative. If you're gonna do conjoint, you need to have this as an input for your modeling of your choice.

Share when you're actually going to do that research. Yes, there's marginal cost. The only rule is, is price above it. What you're looking for is that differential value and that differential value, plus the price of the alternative. 'cause you're displacing the price of the alternative that forms the economic value to the customer.

Some portion of that you can capture in price, then yeah, you get a contribution margin outta that. The contribution margin is the result, not the input of good pricing. You start to realize that if you strip it away, there's just really two questions. You wanna know what's the price, the alternative, and what's my differential value.

From there, I can kind of estimate where that price will be. That's it. Now, how do I understand my differential value? You conduct mental experiments, you conduct thought experiments, you create a model of the value, and then you try to go out there and quantify it for different market segments. Here's an example.

We have, uh, a GE came up with a new evolution series. Locomotive, they were displacing their own locomotives that were short priced around for the case, for the purpose of the case study, I'm gonna say 2 million US dollars Now, these things had a huge fuel savings. You can calculate the amount of fuel savings of a, of a GE locomotive over the course of a year, and you can discount it because they're gonna use a locomotive for more than one year.

So you can look at it for a present value of future years, and that fuel savings amounted to about 3 million US dollars over the lifetime of this vehicle. There were extra maintenance costs because it's a more complicated machine. Maybe that de detracted some value that I estimated in this model at 800,000 US dollars.

That told me the total economic value to the customer, the value put on the table for the customer to enjoy was $4.2 million. That's my hypothesis about the value of this to a customer. Now different customers perceive value differently. You may have a segment out there who loves benefit one, two, and three, didn't like benefit four, you were missing it.

You just didn't offer it. That forms one calculation of your exchange value or your economic value to the customer. Remember though, there's commodity buyers out there who never, ever care about your points of differentiation. You can put all the money in it you want to, they don't care. They're just gonna buy it at whatever they consider the cheapest price in the market.

That's a commodity buyer. You have a utility buyer out there as well who actually did care about benefit one and two, didn't care about benefit three and four, and they're willing to pay a pretty darn good price. They're creating a lot of value out there for that customer. Now think about what I'm doing with this study of the economic value to the customer.

I'm creating a hypothesis about what benefits matter to the customer. I'm creating a hypothesis about the value of those benefits. I'm even creating a hypothesis about how to segment the market. Now, all of this is necessary if I'm gonna design a good market research study. 'cause I wanna know who I wanna ask, what range I wanna ask them about.

And what benefits do I wanna talk about and actually test to see if they are real benefits for that segment or not. This is how you form the foundation of doing a good market research study. Now, the economic value to the customer tells me the value on the table. The question is, is how much of it can I take?

I gotta least selling the value for the customer if I'm going to encourage them to purchase. Shorts and longs rules. There's about a 15, 30, 15 heuristic on this. It's been tested in different situations. I've tried it with different areas. Many people use it with high success. For every dollar you put on the table of differential value, you can set your list price at 50%.

You set your expected closing price at 30% of that extra dollar. And you say, I am not selling. If I can't get at least 15% of that extra dollar of value, or that's a walkaway price, this approach works pretty well. Phillips Healthcare uses a 25% rule for the price setting from an economic value to customer viewpoint.

GE has done something very similar to this. I've seen it. I've used it. I worked in software back in nineties selling software to utilities, and I asked, how much would you pay for extra dollar of value We create, and yeah, they had the power. They pushed me down to 15%. It worked. The methodology works. How do you do this?

You talk to your product managers, you talk to your salespeople. You try to understand how the market really thinks about the value proposition. You can use your own alternatives as a default. You can use your competitors if you understand how they're priced. You know that that pricing of competitors offering is often and especially in B2B, questionable value accuracy.

But it gets you a ballpark, and that's pretty good. You're, look, you're looking at the problems that they sought to solve with your software. Hard good product. Scanning machine, I don't care. You're trying to figure out what the problems we're trying to solve. You get them to describe it in words, and your job as the pricing researcher is to create a math problem from their word problem, which you then solve and tell them the answer to their word problem that they didn't even know that they had.

So it's a bit of creativity you get to do here. After you have that word problem in the model, you can ask customers, well, how much do you pay for fuel? Like back to the GE locomotive, how much do you pay for a fuel? How much do you drive a train in a day? How much do you drive it for long haul, for commuter, for, I don't know, real yard operations?

How does this work between different trains? How do you, how long do you use the train? And you can start. Clarifying those parameters you have in the economic value to customer through that direct questioning, through an informant and a qualitative research approach. And that'll help create a better hypothesis that then you can go out there and test with quantitative research.

And we got three different methodologies, which are commonly accepted in the market for pricing products. Gabbert Granger. Van Western door price sensitivity meters and conjoint analysis. I've seen all of them. I have done all of them, and they have D trade-offs. We'll get there. All right, so let talk about some of the research we have.

Suppose I just simply said, how much would you pay for my glasses? How much would you pay for a bowl of soup? Direct question, just how much would you pay for this? What happens is you just get that reference price coming outta their head. It's gonna come from the dorsal, lateral stratum, you know, it's your quick memory or your lizard brain with, and it doesn't have a lot of thought behind it because what am I not asking them?

And I say, how much would you pay for a bowl of chili? I'm not asking 'em what their alternative is. I'm not asking them what they're comparing me to. I'm not asking them to even make a comparison. I'm not asking about the benefit differential. I'm just saying tell me a number that you think of when I give you a bowl of soup.

Ah, it's horrible. The next problem is, is they start to they quickly engage in bargaining behavior. Well, I know that I pay 10 bucks for a bowl of soup, but I don't, come on, I wanna pay six. Ah, that's the words. I want to pay. What you're measuring here is how much they want to pay, not how much they would actually pay.

It's a different measurement. So what I'm getting is their price expectations, their price want to pay. I'm not getting their willingness to pay. It doesn't simulate a real purchase decision.

the price I will identify from direct questioning will always be biased low. How much low? I can't tell you. I've seen pretty big differences out there, but it'll be below their willingness to pay. The Van Webo methodology relies upon direct questioning techniques.

That should make you hesitate. See what the van we price sensitivity meter says is. I got a chili, TJ's Chili. I like to make chili. Okay. TJ's Chili truck food Truck sells 12 ounces of servings of Chili's, both vegetarian and beef. I live in Chicago. We got a lot of vegetarians. Although if it's a vegetarian chili, I think the proper name for, it's called beans.

But anyway, I'll, I'm not in Texas anymore, so we got both vegetarian and beef chili out there with all the trimmings corn, chips, geso, chihuahua, cilantro sibo chopped onions. I'll put it on top and you kind of mix it up and maybe if you want some, some sour cream and hot sauce out there.

Sound pretty good. Great. Then we ask four questions. What price would you consider this offering to represent a good value to be getting expensive. So expensive? You would not consider it so cheap. You would question as quality. I ask these four questions. I do cumulative distribution functions and one minus the cumulative distribution functions.

These are one minus the downward sloping. These are cumulative, the upward sloping for the, considered expensive, considered too expensive, considered cheap, considered too cheap. Down there you find the points of intersection, and then you give 'em fancy names like the prance of marginal cheapness, the price of marginal expensiveness, the optimal price point, and the indifference price point.

And voila, you've made a measurement, but of what? You've made a measurement, you've done the accuracy. You may have done this perfectly, even segmented your market very, very well. But what does this really tell you? The optimal price point isn't the optimal price point. I didn't consider my cost. I have asked for no trade offs.

I have the indifference. Price point isn't the point where people are indifferent between buying and I mean it is. What did you learn? All you can learn is that people expect to pay for this chili somewhere between six bucks and nine 50. I do teach at DePaul. I have taught market research. We. Have done Ben Washington or price sensitivity meters with my classes on bowls of chili, on soups, on pizza.

Guess what? I get the same answer every time and I could have learned the exact same information, but simply looking across the street and looking at the prices of McDonald's for a hamburger, I didn't learn much. But, okay, I will do this. I will be glad to do this. If you wanna pay me. Just realize that the answer is not gonna tell you where to price.

I've heard people say, just use this as the demand curve. Well, it's not the demand curve. I didn't even ask. I haven't tested that. It's flawed. So why do people do it if you know it's flawed? 'cause it's easy. Because I can get responses pretty quickly without tiring up my audience, my, my informants. I can usually get decently converged answers on those four price points with just 20 people responding.

So small samples and it's quick to execute. Most software companies will allow you to, most survey companies will just, boom. Out it goes. It does deliver a market informed range of price expectations between the price of marginal cheapness and partial marginal expen expensiveness. But you know what?

It doesn't tell you what the willingness to pay is. It doesn't deliver a reliable demand curve for doing price optimization, and that's whether I'm doing price optimization for market share price, I'm sorry, profit or revenue. It just doesn't do it. I can't interpret the curves reliably as demand curves.

These names of the intersections, as I said, they're meaningless and it has all those flaws of the direct questioning methodology on price, bargaining behavior, lack of trade-offs, usage of the reflex memory, p. I know of a very large Chicago based hamburger franchise who deployed van. We price sensitivity meters to try to price hamburgers across the United States and then said the results were bad.

Let's fire the, the research company. That wasn't me, by the way. And the answer is the results weren't bad. What you asked for was the wrong thing. Sometimes you have to tell clients. What a better looks like to help them move forward. So let's draw a little bit better. Let's try a bio response survey, where I show you the item.

It's a pair of reading glasses. I don't know if you can see it. It's a simple pair of reading glasses. There they're on and I give you a price you use reading. Glasses are 30 pounds. Okay, fine. All I ask then is. Well, you purchased this pair of reading glasses at 30 pounds. No. Assuming you're in the market by buying reading glasses, you may or may not.

You may find that a little expensive. Well, the research by academics, tell me again. Respondents engage in bargaining behavior and they provide lower prices and they say no more often than they actually would. What we'll express is, again, our price expectations, not our willingness to pay. And again, I did not ask you to make a trade off between one pair of glasses and another.

You know, there's plenty of styles out there. So what I'll identify is below the willingness to pay something. When I use GABA Granger, it's basically asking a direct questioning. Technique. It's just going for him. What does GABA Grange look like? Back to my chili truck. TJ's Chili truck sells 12 ounce servings of chili, both vegetarian and beef, with all the trimmings of corn chips, queso, chihuahua, salad throw, and chopped onions on top.

Tasty. How likely are you to purchase at a price of, and you start at like eight. And then you say very likely, somewhat likely, neither. I mean, this is a liquor scale right here. Pretty standard. As long as they pick the top two box, you go to the next one. And how likely are you to PRI purchase? At a price of nine 50 they said yes.

So you say, okay, great. We'll try the next one. How much, how likely are you to purchase at a price of 11? They say no. Okay, fine. I'll score nine 50 for that respondent, and then I'll have other people who says eight bucks. Oh, no way. Six 50. Yeah, sure. I'll pay six 50. Five bucks. Can I even make chili for that cheap?

Yes, this is the answer. But what am I gonna get out of this if I do this enough, I can create a demand curve from their GABA Granger of the price they want to pay, not their willingness to pay. So here's my demand curve for TJ's Chili at different prices, and that tells me my profit and my profit is being maximized around 12 bucks.

So let's price it at 12 bucks. And I note that, you know, my choice share be kinda low. Maybe that's a good way to go for it. Okay, we can do that. We can do that. Why do we do this? Because it's easy. It provides a price expectation, informed demand curve that I can optimize against. Unlike Van Western or price sensitivity meter, it's easy to execute, it's easy to get responses.

I can again, converge with relatively small samples of 20 people or so. It's quick and execution, and it does deliver market informed price expectations, so, okay. Yeah, but it is still flawed. It's flawed by the bargaining behavior and the lack of trade-offs. See, I didn't ask you about that bowl of chili versus a hamburger at McDonald's.

I need to actually ask about real trade-offs if I want to understand how people are thinking from the three questions. What's my alternative? Better or worse than do I care when I'm doing conjoint analysis? I'll actually ask respondents to make trade-offs between two or more offers as to which they would purchase at what price?

Well describe the products as, collection of attributes. Price is one of those attributes. One's tabulated the simulations, calculate the fraction, which I love. Saw. Saw tooth here. 'cause you have nice simulation tools. Okay, I've used them Once. Tabulated simulations calculate the fraction of choice, share of the target.

Offering a different prices against the true competing offers, attributes, and prices. The choice share can then be interpreted as the demand curve. And by requiring requiring trade offs from respondents, I get them out of the lizard brain into their actual thinking brain. I get them to simulate an actual purchasing decision.

Therefore, I get a more accurate, a better understanding of the market's true willingness to pay from my offering. So here's an example of a conjoint that I just made up. So we have, you know, different artists. We have Rothko, Picasso, Marri, Kunz, Montreal. You get the attribute utility levels, you get different colors, like the blue mode, the Picasso, et cetera.

You get the gray scale. Is it dark? Ooh. Like the scream? Or is it bright like a nice, uh. Or Rothco. Rothko made both dark and light ones out there. What shape is the I just made it up, right? I'm trying to mask an actual study and then I got a price range a has a higher utility and a higher price has a lower utility, and you get the attribute importance.

This comes out of the sawtooth type work and you understand. What matters. What matters is the artist's name. I don't know why I chose shape as an important part, but the other thing that mattered was the price. Got it fine. From all of that, then you can run a market simulation knowing the price of the competitors and the shapes and the stuff out there.

You get this little choice share given your, offering, given the real competitive situation, then you can actually try to go ahead and optimize profits and look here. Now it's at five 20 for this particular offering, even though it's at five 20, I'd probably advise the customer to try six 80. Why?

Because it's easier to drop prices. It's hard to low it, and this difference is relatively small. Why not take the extra margin? You can always drop it later on. If I'm wrong,

I like conjoin. It does simulate a real purchase decision. It creates a choice share, which I can reliably use for price optimization. It is the single best method to create a simulated market demand curve considering the trade-offs for an offering. I like it, but it does have trade-offs. The sample sizes need to be up to the a hundred, sometimes bigger.

And if I don't design it poorly properly, my responses will deteriorate. My respondents will suffer from survey fatigue and stop really thinking carefully.

So start with this work of economic value to the customer. Create a hypothesis about what customer's willing to pay. Or what they should pay. That's really what they're asking. What should you pay? What should you pay? By segment, by market, geography, what should you pay? If you wanna do Van Wew Price Institute meter, I got it.

It does reveal the price expectations. It is market feedback, but realize that is not necessarily what you wanna pay at. I mean, what happens, and I've seen this occur when your Van Wew price I meter says the product should be priced at 10, but your economic value of customer says that my, my walkaway price should be 30.

At that point, you realize that you have two different measurements. One, what they wanna pay and one what they should pay. Then you have to make an intelligence assessment about how you wish to proceed into the market. Do you wanna start low and try to raise it later? Do you wanna start high? Knowing that you may have to be negotiated down?

It helps clarify how you go into the market. G Granger at least gives you a demand curve, which, a market informed a price expectation, demand curve, which I do prefer over the van wash store. It's relatively easy to execute, but again, it's not actually their willingness to pay. If I wanna get their willingness to pay using my hypothesis from the economic value of the customer study, I'll go to conjoin.

It reveals a real realistic demand curve. Willingness to pay in light of the trade-offs.

So thank you for listening. Pick up my book Pricing Strategy. It's an e-textbook if you're in Europe or in the States or in Latin America. Or you can buy the actual textbook if you happen to be in India or China or get my pricing done right? Figure out how your organization can. Get pricing done right and stop wasting value.

How was that Dean? Is that what you wanted?

Dean Tindall: That was exactly what I expected it to be. Tim, thank you very much, delivered with all of the energy that I know you bring to a presentation. Seth, thank you very much for that.