Book Review: Competing Against Luck

The Book

Innovative products fail. Sometimes they fail quickly and spectacularly. Sometimes they fail slowly and quietly. Failure has become so commonplace it’s become an accepted cost of doing business. “Fail fast, fail often” is held up as a badge of honor among proponents of being ‘agile’.

Studying failure, and success, is what Clayton Christensen did before writing his latest book “Competing Against Luck”. Professor Christensen and his co-authors, want us to know that we can increase our chances of introducing a successful innovation, and thereby reducing the incidence and cost of failure, by adopting the “Jobs-to-Be-Done” approach.  

Professor Christensen believes we need to change the way we look for opportunities to innovate. The change might be radical. It might be painful. But, if Christensen is right, it will be worthwhile.

Jobs Theory

The basic premise behind Jobs-to-Be-Done is that consumers do not “buy” products or services, they “hire” those products or services to do a job for them. To get hired, companies must offer products or services that do the job better than all other alternatives.  Consequently, innovation starts with gaining an understanding what the job is. Or, as Christensen puts it, writing the job spec.

In this context, Christensen defines a “job” as the progress a person is trying to make in a particular circumstance. Within this definition, the term “progress” was a considered choice. Progress implies a journey. Our job spec must tell us where the consumer is coming from and where they are trying to go. Equally important is the consumer’s “circumstance”. The job spec must also tell us the barriers that are standing in the way of progress and what alternatives consumers are currently using to get where they want to go.

Once we accept the premise that we are competing to get hired, we must also consider the hiring process. Disruptive innovations don’t always involve major product changes. Sometimes they simply improve the purchase experience. Consequently, our job spec must also include the social and emotional dimensions of the job. What are the anxieties involved in the purchase process? What are the rewards the consumer is seeking?

My Five Take-aways

The book is full of interesting stories of companies that have implemented, or not implemented, a Jobs-to-be-Done approach. Each and every reader will find something different that is meaningful their particular experience. I’d like to highlight five concepts that I found particularly meaningful.

1) The importance of the experience

I already discussed how understanding the consumer’s experience is a critical component of the Jobs-to-be-Done process. Allow me to offer up an example from my own experience.

In 2014, a new mattress company appeared on the scene. When considered individually there was nothing particularly innovative about any part of the Casper product or the purchase experience. Foam mattresses were a dime a dozen. The technology to compress a mattress and put it into a box had been around for years. Consumers could easily order a mattress from any number of retailers with a comfort guarantee. What the leaders of Casper had that the leaders of the legacy brands and retailers didn’t have was an understanding of the barriers that were preventing consumers from buying a new mattress and a will to innovate against those barriers.  

Casper’s innovation was not the product but in pulling together all the existing technologies to create a better purchase experience. By innovating against the barriers to purchase and the anxieties consumers had, they brought non-consumers into the market and took existing consumers away from the less-than-perfect alternatives. 

2) Becoming a “purpose brand”

The concept of a “purpose brand” is almost a throw-away in the book. It takes up less than five pages about halfway through but it caught my attention. 

A purpose brand is one that has become synonymous with the job it is intended to do. As Christensen writes “a well-developed purpose brand will stop a consumer from even considering looking for another option.” A purpose brand is the consumer’s first choice and the choice they pay a premium for. As examples the authors cite Uber, Lunchables, FedEx, and Sawzall. Creating a purpose brand should be the stretch goal of any new product development effort.

As I read this section, I flashed back to the struggle my previous company had in developing our “brand purpose”. It would have been so much easier had we had the insight to know the job consumers were hiring us for. Instead we argued endlessly about what would make ourjobs meaningful. We had it backwards.

3) The errors we make

The book lists three “fallacies” that can drive companies off course. The first fallacy is the tendency of companies to design systems that measure their own processes and not how well they are doing at performing jobs. When a successful company is first established, the processes created are usually aligned to the job. Measurement systems are developed to drive efficiency in those processes. However, jobs are not stable. They change. If the processes do not change with the job, then the measurement system falls out of alignment with the job. When managers become overly focused on these measures, they lose focus on the job. Innovation opportunities are missed.

The second fallacy is the seemingly innocuous desire to increase sales to existing customers. The danger of this is that the focus of the innovation effort shifts to the customer and away from the job. I saw this personally as my previous company put great effort and expense into developing an array of products with the stated goal of attaining greater placement at existing retailers. Meanwhile, the disruptive innovator was making all these retailers obsolete by focusing on the job the consumer was hiring for.

The third fallacy is an over-reliance on data to make decisions. One of the unintended consequences of creating a culture of data-driven decision making is that data can and will be found to support any proposed action. As a marketing analyst I experienced this fallacy repeatedly. The example that drove me crazy was when I would provide a chart to a salesperson for a sales presentation only to see the same chart show up later in an internal strategic presentation. Under the guise of data-driven decision making, the strategic leader went looking for data to support a course of action most beneficial to their position and found it in a sales presentation. Had I known the chart was going to be used to drive company strategy, I might have created a very different chart that supported a very different course of action.  

4) Our misguided measurement systems

Companies, and investment analysts, put a lot of time and effort into measuring existing internal processes with little regard to their relation to the job consumers are hiring for. Honestly, the relationship between a product’s share of the shelf and it’s ability to do the consumer’s job is pretty tenuous and yet that metric is often front and center in many product launch scorecards. Putting resources around driving to a numerical goal not aligned to the job is not an efficient allocation of resources.

In Christensen’s view, any measurement system is dangerous as it risks taking the company’s focus off the Job-to-Be-Done. However, if a measurement system must be used it should be carefully constructed around the jobs consumers are hiring for. Consequently, measurement systems should not be standardized across products. They should be designed only after the company has aligned a product around a particular Job-to-Be-Done and designed their internal processes to successfully meet the job requirements.  

5) It’s not the number of reviews; it’s the content

While not core to the concept of Jobs-to-Be-Done, the book does contain an interesting insight into consumer product reviews. The authors contend that consumers are less influenced by the pure number of five-star ratings than by the individual reviews of people who were hiring the product for the job they have. For example, a single five-star review from a consumer who clearly stated they were hiring a home printer to print on card stock is more relevant than ten reviews from consumers that simply say they love the printer. Consequently, companies should encourage more detailed reviews and sort by content.

What’s missing?

The book does an absolutely wonderful job of selling us on the Jobs-to-Be-Done approach to innovation. In fact, it’s even a little annoying that they never stop selling it. At some point in time we don’t want the hard sell anymore. We want to know how to implement it. It’s not that the authors don’t recognize this. They try. They just don’t succeed very well. It is as if they sold us on a destination but only gave us a partial map on how to get there.

The chapter on “Integrating Around a Job” is particularly confusing. All of the examples in this chapter seemed to be there to emphasize the importance of designing processes around a Job-to-Be-Done. However, the examples provide precious little detail on how these leaders went about designing and institutionalizing these processes. The authors seem to think that once the hard work of spec’ing the job is done, everything else just falls into place. No reorganization is that easy.

Personal Summary

This book has forced me to seriously rethink my previous job as a Director of Consumer Insights. Generating quantitative data has become so easy and inexpensive that, in the competition for scare resources, we often skimped on qualitative research in favor of more quantitative data. Our innovation charters became filled with data; data about the target consumer, data about the incidence of a specific need, data about a willingness to pay a specific price. All this data wasn’t going to make us more innovative.  

Innovation requires inspiration and a will to change. In my experience, inspiration is more likely to come from qualitative research. Jobs-to-Be-Done provides a framework for conducting and presenting qualitative research in way likely to inspire innovation. Generating a will to change? Well, that comes from distributing copies of this book and getting your leadership to read it.

Better Research through Automation?

As I was preparing to launch Morsights, I spent quite a bit of time updating myself on some of the latest trends in the Insights field. There is no doubt that the role of Insights Director has changed from one of managing suppliers to one of managing data. One of the many manifestations of this change is in the explosion of firms promising to automate research processes.

What is Automation?

In some respects, the trend toward automation is just an evolution of the DIY research trend. Survey Monkey and similar online research tools made it easy for anyone to program an online survey, no coding experience required. From this beginning, it’s easy to see how question phrasing, sampling, report production and even statistical analysis would be the next candidates for automation. This is work that previous generations of Insights Directors would have outsourced to a full-service research firm.

Now, firms such as Zappi, GutCheck and Conjoint.ly have introduced automated, online tools for concept testing, price elasticity testing, advertising testing, prediction modeling, and more. Instead of outsourcing the work, the Insights Director simply needs to select from a menu of choices and watch their email box for a report. 

What is the appeal of Automation?

Developers of automated research tools typically make two claims regarding their tools. The first claim of is greater agility. “Test early, test often” is a phrase one often hears when talking with vendors of automated research tools. Having introduced automated idea screening and TV ad testing to Serta Simmons Bedding, I can personally verify that automated research is faster, simpler, and cheaper than outsourcing. When compared to the alternative of using a DIY survey, it’s faster, simpler, and, sometimes, even cheaper.

The second claim heard from developers of automated research tools is that by automating “mundane tasks” their tools will free up time for the Insights team to conduct deeper analyses or to build better stories. In other words, a promise of better research. For the client-side Insights Director this claim is problematic because it’s validity depends upon the alternative.

If the alternative is utilizing a DIY tool then the claim of time efficiencies may actually be valid. An Insights team member who previously had been writing surveys or building charts might now be in a position to reallocate that time toward looking for interrelationships in the data or perfecting the internal presentation. However, this assumes the person is qualified. 

If, however, the alternative is outsourcing then the claim falls apart. When outsourcing a project to a full-service vendor, the Insights Director is not just outsourcing mundane tasks they are also hiring a team of vetted, knowledgeable analysts and storytellers. In other words, a full array of expertise that may not exist internally.

What is the lesson here?

The lesson I’ve learned is that before choosing to automate a research process, the Insights Director must carefully consider the level of expertise of their team. If the team has sufficient analytical knowledge and ability to translate insights into memorable stories then automating a research project may be something worth exploring. If that level of expertise is not in-house then using an automated research tool may be faster, simpler and cheaper, but not necessarily better. In it’s current form, automation is a poor substitute for expertise.