Building the “Frictionless” Organization—an Interview with Bill Price

Bill Price is a longtime friend and industry colleague in the customer service space. Bill was Amazon’s first global vice president of customer service and is the founder and president of the consulting firm Driva Solutions. He and David Jaffe authored The Best Service Is No Service, Your Customer Rules! and, most recently, The Frictionless Organization.

Bill and David believe that customer contacts provide valuable insight into where organizations can reduce friction through better processes, services, and products—a view I’ve long shared. In The Frictionless Organization, they contend that “very few organizations effectively measure the reasons customers are forced to contact them for help. Fewer of them have data on how well these contacts are resolved, and fewer still have analyzed the causes of those contacts and who in the organization is responsible for fixing the problems.”

Bill Price
Bill Price

Here’s the dialog Bill and I had on this opportunity. (Note: The conversation appears as is, and if some of the terms or abbreviations are unfamiliar, you’ll find contact center and customer experience glossaries here.)

In The Frictionless Organization, you write that becoming frictionless is “a strategic imperative.” How do you define frictionless, and why is it so important?

We define friction as unnecessary contacts that customers have to make and, more broadly, anything that interrupts customers’ lives by introducing confusion or the stuff that doesn’t work. My co-author and I interviewed more than a dozen companies recognized for delivering excellent customer experiences around the world and then conducted secondary research into many more organizations that are striving to become frictionless. Among these organizations are Airbnb, Amazon, Apple, Australian Tax Office, Blizzard Entertainment, CableOne, E.ON, N26, Nike, RedHat, T-Mobile, Tesla, Trek Bikes, Uber, United Airlines, USAA, Vodafone, Xero, and Zoom. What we discovered was their customers are increasingly expressing their frustration when they encounter friction, including taking their business elsewhere, and that removing friction creates these great customer experiences. 

The other strategic imperative is that organizations with less friction save money. They incur lower operational costs and can reinvest this in products, services, or customers. It’s hard to compete with businesses who have costs 20% or 30% lower than yours. In many industries, we have seen newer businesses enter an industry and disrupt with these lower-cost and lower-customer effort models.

David Jaffe
David Jaffe

When companies are frictionless, it is less obvious because, by definition, everything works like it’s supposed to work, and customers don’t need to contact the company that often anymore. In every industry, innovators are promising, and in many cases delivering, a far simpler customer experience as their differentiator vs. the incumbents or entrenched players: Like the T-Mobile “Un-carrier” strategy removing contracts that sent AT&T and Verizon crazy trying to keep their customers under contract. This means that companies that have been around for a while, and even recently-formed companies, need to be on the lookout to make sure that their proposition and go-to-market program is not eclipsed by these innovators.

What’s your approach to measuring the level of friction in an organization?

You can measure the amount of friction by looking at the rate of assisted contacts over a time period overall and by issue or by reason. We recommend between 25-50 reasons, in the customer’s exact language, such as “My service keeps dropping out” or “Where is my claim?”  We define this rate of contacts as CPX. C = the number of customer-initiated contacts across all channels that require live support (inbound phone, chat, e-mail, or social), and X = the number of customers or orders or other business driver. For example, if an insurance company has one million accounts and two million contacts a year then the CPX is 2. Lowering that number will demonstrate that they have reduced friction and made themselves easier to deal with.

Frictionless organizations want to see CPX decline over time. For example, Uber defined “defect contact rate” as their CPX and told us that it had declined by 90% over a five-year period, enabling them to grow dramatically on the top line without needing to add more customer service staff. Clearly, there needs to be another CX metric, such as Net Promoter Score (NPS) or Customer Effort Score (CES) to keep yourself honest so that you’re not removing some of the positive reasons for contact. These metrics are also important to keep tabs on how your automated contact channels are working. 

You and David point out, and I concur, that the metrics many organizations use—e.g., average handling time, first contact resolution, even net promoter score—can hide important details. Please tell us what you mean.

When we met with and studied these organizations, they confirmed to us a longstanding belief that averages mask the varied nature of reality. Over the years, we have created way too many metrics, almost all of them abbreviated with three letters, such as ASA, AHT, NPS, FCR, and emerging metrics such as CES or PRR (Positive Response Rate). In many cases, organizations simply look at the overall number, the average, and if the average goes in the right direction, they move on to the next topic. However, lurking behind these averages are the long tails; for example, if NPS is 36 there are still a number of customers who will gave 0, 1, or 2 scores, which today are called deep detractors. The frictionless organizations are ones that focus relentlessly on identifying and reaching out to these deep detractors to figure out why they’ve had such bad experiences so that they can turn them around or, at the minimum, figure out how to make sure other customers are not affected in the same way.

Similarly, as I mentioned earlier, FCR might be 70, which means 30% of customers, on average, do not get resolved the first time (where companies can measure FCR, which many can’t). However, in order to produce that average, there are customers who might have to contact the organization six times or more in order to get their issues resolved, and clearly they are very dissatisfied! In addition, it’s important to look at FCR across channels that a customer uses to contact you. Many organizations measure FCR in each channel in isolation. For example, if the customer goes to a chatbot first to get an answer and it doesn’t work, they needed to place two contacts. If they navigated through an IVR system that has potential answers to problems and it didn’t work, that also did not produce resolution. Then, if they contacted the support center and still did not get what they wanted, they failed the third time. Unfortunately, many organizations simply collect and report the FCR in the final assisted channel and not in the upstream channels that the customer already navigated.

You write that frictionless organizations ensure that being frictionless is a “whole of business” requirement. When product and marketing teams are tasked with more revenues, that can create conflict, right?


Courtesy of Bill Price and David Jaffe

Exactly! It should create some tension. The risk of product and marketing driving revenue relentlessly is that it comes at a high cost that the customer support areas have to handle. The cartoon illustrates the point that marketing and product teams take shortcuts or mismanage expectations and then support has to pick up the pieces. That’s why it’s even more important that organizations that want to become frictionless bring together all of the different functions to solve the customers’ problems.

In most organizations, customer service doesn’t cause the problems that it has to fix. The causes are in marketing, product design, delivery, web design, and so on. The organizations we interviewed and studied showed us that they gather their cross-functional leaders on a frequent basis to review CPX and other customer experience metrics as well as to assign responsibility—what we call “owners”—to each of the customer reasons that customers are expressing to the organization. For example, “why is my bill so high?” might be the responsibility of the head of billing or product marketing, while “why didn’t the salesperson tell me about X?” might be the responsibility of someone in retail. In those situations, it’s important for the heads of all departments to be working closely with the customer service team to determine what the rate of contacts is for the reasons assigned to them and then to be able to come up with the appropriate action plans to deal with them.

You say that “digital first” should be replaced by “eliminate first.” Tell us what you mean.

Ah, yes, another interesting debate! Over the last five to seven years, we’ve heard that “digital first” should be the primary strategy for organizations since customers are increasingly going first to websites, apps, portals, or other digital locations. However, a significant percentage of customer reasons fall into the eliminate action category. These are ones that are irritating for customers, and companies do not want to have to handle them with live support. These issues start with statements that are near complaints, like “where is my X?” and “why didn’t you do Y?” We found that these reasons have the highest costs and the lowest CX results, including the highest levels of churn. By contrast, the digital or digitized reasons are ones that are valuable for customers but, equally so, are ones that organizations do not want to have to handle with assisted support, and their costs are lower and CX higher in the first place. These digital interactions are often transactional, like placing orders, making payments, or checking balances.

Therefore, it’s important to distinguish among the different customer service reasons to determine which ones should be eliminated versus which ones should be digitized.  The reason “eliminate” reasons aren’t tackled is that the causes are harder to tackle and owned in multiple areas. For example, customers asking “where is my X” may need help from logistics, inventory control, distribution and the website owners in the case of a retailer. The improvements may be hard to manage or need change across an entire workforce or network. It’s much easier to add some digital transactions than change the behavior of 5000 sales people! In the research that we did for the book, and our own consulting experiences, 30% to 50% of the volumes and even more of the cost of contacts fall into the eliminate action vs. 25% to 30% that fall into the digitize action. Therefore, it makes more sense to determine how to conduct root cause analysis to remove or eliminate reasons—but it is hard work.

Similarly, using the FCR concept, it is very important to figure out whether the containment rate (those who complete the action they wanted) in the digital channels is high and is therefore satisfying customers. Many organizations cannot even measure or report the digital containment rate, but the frictionless organizations do so, sharing with us that in some cases they had a 96% containment rate in the digital portals (such as the business software provider Xero based in New Zealand and Australia). This is an astoundingly high number but one that they measure so closely in order to get it asymptotic to 100%, which they see as the best customer experience.

Despite predictions that the percentage of calls vs. email, chat, or social would decline, we still see significant inbound call volumes. What’s your take on why that is happening?

There are two key reasons and then a range of subsidiary reasons. The first key reason is that phone calls are still the best way to solve problems in near real-time. If customers need help or something isn’t working, they still call (and hence the eliminate reasons are often dominant in contact centers). The second problem is that the digital channels often drive new and different contacts, like customers unable to log on or digital interactions not working as they expect. As digital security gets more complex, then support for things like multifactor authentication are growing. One bank recently rolled out a new authenticator app for its digital platform, and the complexities involved doubled the call volumes!

There are several related reasons adding to the inbound call volumes. We’ve covered some of these already but let me summarize them:

  1. FCR for e-mail and chat are typically much lower than phone calls so customers who don’t get what they need through those channels often wind up placing a phone call as well;
  2. Some organizations don’t bother to interact with their customers via social media, so that induces additional contacts typically into the inbound phone channel;
  3. FCR in the call channel is often below 80% which means that you will see repeat contacts, or what we like to call “snowballs,” come back into the center;
  4. Organizations that do not adequately eliminate or digitize customer contact reasons are bound to get even more phone contacts in the future as they see their products and services become more complicated and their sales growth increase.

Each of these four reasons leads to specific actions that organizations should take in order to reduce the number of inbound call volumes on a CPX basis as well as on an absolute number.

Why hasn’t “omnichannel” support taken off in many organizations?

We’ve all been hearing about the promise of omnichannel for many years now. Unfortunately, in many organizations, these different channels are set up and managed separately and don’t even have the same knowledge bases or platforms to support them. In addition, most organizations do not store the information collected in earlier channels. For example, as United Airlines told us, many customer service agents do not even know that customers had previously been on the website or been in a chatbot to attempt to answer their problem. United and other frictionless organizations we interviewed are requiring that the same knowledge base be used across each of the channels and that any of the earlier channels’ results have to be fed downstream, especially to the assisted channels like phones.

The other problems are ones of technology and people. Many organizations haven’t joined up all the different channel technologies so that they can deliver true omnichannel experiences. A phone-based agent may not be able to send or see messages from the customer and may be restricted to work in their channel. The people challenge is that it is hard to equip staff to deliver omnichannel experiences. They need to understand what is possible in all the digital channels in order to support them. We often find staff trained in just a single channel and how it works, and that can be demanding enough. Unless you have great technology, it’s really hard to take a call one minute, send an email the next, then reply to a text, and lastly reply to a message. Some need verbal skills, some written skills as well as needing to know how each channel, portal, or app works. It’s a complex problem despite platform providers claiming they can offer omnichannel support. With high staff turnover, it’s often easier to train staff in just one channel. 

We understand that this is not easy to do, or else more companies will have done it. However, customers are increasingly frustrated when they have to repeat the same information from channel to channel or from agent to agent in the center. Just listen to your customers tell you “I just said that to you!” or “I was just on your website and I couldn’t get what I wanted,” and take heed—every time this happens, make sure that you can figure out how to connect these channels and make them truly omnichannel. Sometimes that can be as simple as asking the customer what they have done in other channels rather than ignoring it.

Looking ahead, where do you see customer expectations and customer support in three to five years?

I’ve been wrong in the past about crystal ball gazing including predicting the replacement of inbound phone channels with e-mail, but what the heck, let me try again! In three to five years, I see two distinct types of customer support in response to the same level of customer expectations. Customers will increasingly expect fast and easy solutions in the channel of their choice, connected channels, so they don’t have to repeat themselves and, overall, less need to contact organizations for support at all. Many customers are already saying, “Why can’t you be as easy to deal with as Amazon?” and I expect these “last contact benchmarks” will increase. 

However, there will be two different types of customer support from organizations:

  1. Frictionless organizations that will see a steady decrease in the CPX and increases in their customer experience metrics, as well as enjoying reduced costs and less need to have to invest in additional support operations; or
  2. Organizations that will be going out of business.

I know that sounds dramatic, but with the rapid acceleration of new companies and new ideas and the proliferation of online delivery options, organizations that are not operating in a frictionless way will rapidly be eclipsed by those who are. It’s as simple as that.

How can readers learn more about the book and these topics?

My coauthor and I set up a website, called, where readers can find more information about the book and new cases and stories that have come to our attention since we wrote the book earlier in 2022. In addition, I have written a number of posts in CustomerThink, and they can be read on after clicking on the Authors tab and advisors link to find my name, Bill Price. I’ve also been posting articles and podcasts on LinkedIn. Lastly, I’d be happy to receive any questions or comments directly from your readers to my e-mail address