Value-Based Pricing: Requires Brains or Brawn?

Brains or Brawn | Software Pricing

If you're like most software companies, you probably have challenges getting paid for the value you deliver. Historically, pricing may have evolved from complicated cost models, or more simply, from how the founders arbitrarily decided to charge for their software.

Add in a few years of development, some new opinions on the matter, and voilà, out comes an ugly looking monetization strategy that your sales team has trouble explaining and prospects have trouble understanding.

Regardless of where you're currently stuck, the question is: how do you set prices based on the value you deliver? In other words, how do you migrate into a value-based pricing approach?

Value-Based Pricing is All About Perceptions

Value-based pricing is impossible unless you first understand, at a very intimate level, the value you deliver in the eyes of your customers.

At its core, studying someone else's perception of value is an exploration in human behavior, because it deals with people's perceptions. And people's perceptions come from their world views and opinions. This is why tools that attempt to automate your decisions on how to price your software fall short as a complete solution (and why dating sites still find it very difficult to find most people a decent significant other).

As it turns out, understanding people's behavior has a lot to do with how you capture information, as well as who you get information from. But once you leave the comfort of the office walls, virtual or physical, how do you wade through the multitude of customers' opinions of the value you provide?

The findings may surprise you.

Interview with Dr. Peter Huber

When it comes to customer research, Dr. Peter Huber has been around. During nearly 30 years as principal of PB Huber Analysis and Research, he has studied everything from pickled sausage to complex software in a variety of business sectors: shipyards, aerospace, oil and gas, radiology, mobile app tracking, content management and more.

He has plenty of experience fitting the approach to the problem. So we interviewed him on how software companies should approach the problem of understanding the value they deliver to their customers.

Chris Mele (CM): Can you tell me how you determined value for one of your more complex clients?

Peter Huber (PH): Sure, I'll use Husqvarna as an example. Today Husqvarna is a multi-billion dollar industry giant, but they didn't start out that way. Over the years, they have worked with us on a series of projects to better understand the value their products provided to their many different types of customers.

What made our research so successful was that we adjusted our approach for each situation. For example, we were asked to learn how many pulls people think it should take to start a chain saw.

One misguided approach would have been to just survey a ton of people, ask them for a number, and calculate the average. But a better way was to test several products with roughly twenty people, observe how they actually started the chain saw, and discuss their views.

CM: Did you learn something unique using that approach?

PH: We ultimately learned that the question was not meaningful – the number of pulls varies with conditions – basically, it takes more pulls when it is cold. But what customers really cared about was that Husqvarna chainsaws were easier to pull.

You can get so deep in the weeds of product engineering sometimes that you overlook key product benefits that are hugely important to your customers. Had we not talked with those customers face to face, we would have completely missed the boat.

“A lot of people find a comfort in large numbers – the more data points, the better. But this comfort is misplaced.”Dr. Peter Huber

CM: Many SaaS companies use large sample surveys in their marketing research efforts. Is this really the best market research technique to use to better understand your customers?

PH: There is a wide-spread illusion that large sample surveys are more reliable and more efficient. I think a lot of people find a comfort in large numbers – the more data points, the better. But this comfort is misplaced.

CM: Can you explain that?

PH: Sample size is one of the elements that affect statistical reliability, and the easiest to measure, but it is not really the most important. The most important element is generating a sample that is a random representation of the group you want to study. But to get this right, you have to be picky about where you find participants and how you screen them. This can be expensive, and the expense can lead people to substitute quantity for quality.

With smaller sample targets you can afford to be pickier and have tighter control in terms of both quality of participants and randomness of selection.

CM: At my former software company, we often performed market research, internally and with the help of vendors, to better understand our customers and prospects. It seemed like surveys were the most effective, but over the years I realized our team was missing important nuances. Why is it that surveys seemed to be so ineffective in helping us really understand our customers?

PH: Surveys tend to be so highly automated that they leave little space for exploration and serendipity. Most survey questions are closed questions, with limited choices given for answers, so they leave little room for explanation of context. But, explanations are the richest and most valuable source of information about your customers' world and values.

It is easy to create a survey that gathers and tabulates a lot of correct answers, and yet misses the main point.

“It is easy to create a survey that gathers and tabulates a lot of correct answers, and yet misses the main point.”Dr. Peter Huber

CM: What else affects reliability?

PH: Variance. The more variety in a population, the larger the sample you need for reliable representation. Highly specialized markets, like the ones where many software companies operate, show far less variability than broad consumer markets and can be reliably represented with much smaller samples with a lot of commonality.

CM: So, blasting out a huge number of surveys can actually work against you?

PH: For certain applications, surveys can be useful, especially in the consumer space. But if you're trying to understand the complexities of how your customers obtain value from your intellectual property, you're much better served with more qualitative approaches and smaller sample sizes. You can only process open-ended explanations from so many people. If you talk with one hundred people, you need to capture most of the information through closed-ended questions in order to be able to process it effectively. With twenty people who share a common worldview, you can synthesize a lot of qualitative information.

CM: What kinds of things make qualitative research so powerful?

PH: The interaction is just better. Research participants have a lot to say if you let them. They have knowledge, enthusiasm, and keen interest relative to your products. This is a valuable resource, and the best way to tap into it is through guided, conversational interviews that encourage participants to describe and explain their viewpoints in their own words.

CM: What kind of structure do you use in qualitative research?

PH: We work from a detailed background brief and discussion plan. Interviewers know what we are trying to learn, and why, and how we expect to get there. Participants often respond to one simple question with an explanation that also answers several others. We let them talk and capture what they tell us. If they bring up topics we weren’t planning to cover but they think are relevant, we find out why. If they have already answered a question, we don’t ask it twice.

CM: Do you read a script and ask participants the same questions?

PH: We don’t use scripts because they can be boring and frustrating for respondents, and can short-circuit the discovery process. Some people feel that having interviewers read a script assures that each question will be asked exactly the same way for each participant, and every answer will be within fixed parameters. This provides a false sense of control, at the expense of responsiveness.

Instead, we use an intelligent guide, good briefing and highly experienced interviewers to provide control, but without sacrificing responsiveness. A conversation with the right balance of leeway and control permits the identification of common themes and areas of difference, and a sense of what customers need. These findings inform product development, packaging, pricing and promotion. Intelligent data extraction and coding can actually produce highly reliable metrics on proportionality of key attitudes, market penetration and share, and market and opportunity size – even from relatively small samples.

CM: What should software companies look for in vendors helping them with customer research?

PH: You need a partner that can adapt their capabilities to your situation, and develop a customized plan with a clear rationale.

Most research situations have distinctive elements specific to your business and industry. You need a partner who can grasp these by learning about experiences with clients in your industry, and with clients in other industries in similar situations. They should be tailoring their approach to your situation, and they should be able to explain to you how and why they have done that.

If they can’t, walk away.

Interview with David Zerfoss

I met with David Zerfoss, former CEO of Husqvarna Professional Products, on a Saturday over a coffee to get his take on some of the keys to high growth. After all, he grew Husqvarna from $24M to well over $500M before his retirement.

David is an engaging, thoughtful and ever-positive guy, the kind of guy you desire to be around just to chat. He exudes an energy level that just makes you feel great and smile. You can tell him your challenges, and he has a way of coaxing out the answer you have buried inside of you, rather than telling you what you should do next.

I asked him about his extensive work with Dr. Huber.

CM: How important was the research component to Husqvarna's growth?

David Zerfoss (DZ): It was fundamental to our growth. To grow that quickly, you really need two things: first, you need a new decision-making framework for your management team to think more effectively and quickly - and second, you need an enormous depth of understanding about your market.

CM: How did you get that kind of depth?

DZ: There is no survey on the planet that can give you that. It's not buried in PowerPoint charts or graphs. It takes an enormous amount of critical thinking and exploration to truly understand the value you deliver in your customers' eyes. Dr. Huber helped us uncover the many nuances of how we were delivering value over the years.  We used these insights to make many crucial and powerful decisions that led to our growth.

Peter also helped us determine the potential in our marketplace. This guided our strategy and resources. Husqvarna wouldn't be the multi-billion dollar company it is today if we hadn't effectively engaged with our marketplace. This had impact around the world.

CM: In summary then, when it comes to understanding value, it's kind of like brains over brawn?

DZ: [laughs] Something like that, yes.

About Dr. Peter Huber

Peter Huber and his research team are part of Software Pricing Partners' think tank on Software Monetization, a research and development effort specifically focused on helping software companies quantify and routinize the process of understanding the value their intellectual property brings to the marketplace.

Living in Durham as an Anthropology graduate student, he designed and wrote the code for a factor analysis of longitudinal study measures at the Duke Center for Aging and Human Development – an early venture into big data. While living in Papua New Guinea, he studied the complexities of human behavior, mapping out complex social landscapes and the intricacies of symbolic ceremonial life.

He created Shady Brook Farms Turkey Brand – did the research, developed the value proposition, and managed the communications campaign - and his research and insights while working with Husqvarna helped them grow to a multi-billion dollar industry giant.

About David Zerfoss

David Zerfoss joined Husqvarna as president of its professional products division in 1991. During his 18-year tenure as president, sales grew by 1,700% (hundreds of millions of dollars). He led the integration of distributors in the U.S. and Canada and created coast-to-coast direct distribution to more than 10,000 dealers throughout North America.

After retiring from Husqvarna in 2009, he formed The Zerfoss Group, which offers executive coaching and strategic visioning sessions for CEOs and their executive teams.

He is a sought after keynote speaker who touches, moves and inspires audiences. He is also the author of two thought-provoking, inspirational books, Stress is a Choice and Stress Less and Enjoy Each Day.

Share this Post

About the Author

Chris Mele

Twitter Google+

Chris is a software monetization and pricing expert and former CEO of an award-winning SaaS company. When not working with fast growing software companies as co-owner of Software Pricing Partners, he's working on his first book to help founders create more powerful businesses & personal lives. An avid scuba diver, Chris should probably move closer to the beach.

Download the Top SaaS Mistakes eBook

Learn how to avoid the most common pitfalls of a SaaS company

Download Now

8 Comments on “Value-Based Pricing: Requires Brains or Brawn?”

  1. A thought provoking piece. I found a lot that resonated. In the discovery phase of a sales compensation plan design project I try to do both interviews (definitely) and a survey (if practical). The interviews provide a lot more detail and insight and are the source of many good ideas that end up in the design and communication of the plans. But the surveys are also useful because they often give a more accurate picture of which parts of the sales compensation program are well received and not so well received. Some sales people complain loud and long about their sales comp plans (it’s the nature of many good sales people) and as a result management thinks they know what their sales people don’t like and how much. But they are only hearing from the squeaky wheels and the general population may not agree. Although the surveys go out to the entire sales force and generally get a 50% to 70% response rate, because the respondents are employees and not customers or prospects it is usually a manageable population. That allows us to include a couple of verbatim questions at the end so everyone can emphasize what they feel is most important, which may not be possible in a quantitative survey.

  2. Great post Chris (and Jim)!

    “Data Driven” vs. Customer Value Driven . . . the former can only help latter, indeed not replace it or outweigh it.

    Looking forward to the Papua New Guinea post as well 🙂

  3. Thanks for this post! An excellent topic! I’m particularly intrigued by the “no script” approach to interviewing. I get the why behind it but still worry about consistency. An experienced interviewer can help a lot. Do your clients ever want to have their PMs listen in on the interviews to 1) hear the interviews first hand and 2) to learn how to do them so they can do future research on their own?
    Thanks again and please do keep them coming!

    1. Hi Bob,

      Unless you’re actually intimately involved in the interviewing process, it gets old fast. We have had clients listen in, but they rarely want to put the time into it. Scheduling also gets more difficult when you’re trying to coordinate additional calendars.

      Instead of listening to the interviews, we recommend you review the findings. That’s what you’re really after anyhow. Beyond that, though, consistency of inputs is not really a goal. This is why the survey approach isn’t optimal in these types of projects. Structuring inputs is required to print out fancy graphs and charts but it misses the goal of the value-based pricing exercise.

      Learning different things from different people is what you’re really after provided you learn enough to put the differences in perspective and understand the various points of view. At the end of the interviewing process, you’ll see consistency of outputs in the form of patterns and common themes. But this requires a highly skilled human brain to put it’s thinking cap on and codify the results.

      As for doing the interviews on your own, I wouldn’t recommend it. I leave that to the anthropologists and highly skilled interviewers. It’s too easy to let internal biases creep through or to “lead the witness” if you’re not careful.

      On a side note, many companies desire to have help from research companies who have domain expertise (i.e. expertise in their industry). That’s not always a great thing either. What you’re really after is a team well experienced in the harder research skills sets and, more specifically, in uncovering value and how value scales within the nuances of the software business model.

      Hope this helps!

    1. Hi Daryl, I’m glad you found this post helpful. Some of the best thinking you will do on your business is how to monetize what you have to offer. I went ahead and sent you an eBook that should help you on your way. Best of luck and call us if you get stuck!

  4. This is a fantastic post! This completely resonates with the kind of attitudes i also encounter with clients. People would rather have statistical significance on things that don’t matter than have a small sample size to answer questions that do.

    1. Hi Jana, I’m glad you liked it! Is it statistical significance or emotional comfort? It’s almost as if we feel better with larger numbers because it makes us feel like we did our homework (i.e. left no stone unturned). As Dr. Huber puts it: comfort in larger data points is misplaced.

      More and more we’re seeing market research firms that don’t really have strong skills in the harder core sciences. I guess in the end it’s just easier (and more profitable) to sell the idea that surveying a gazillion people in 5 nanoseconds is a “data-driven” approach. Unfortunately, using too many data points can be misleading.

      Peter also told me a great story about how the Papua New Guinea government hired him to figure out how to get the natives to grow tobacco — and there was no monetary currency of any sort in their tribe. We’ll try to get that out in a future blog post since it has some great lessons learned too.

Leave a Reply

Your email address will not be published. Required fields are marked *