Taylor Rodgers
  • Home
  • Books
  • Free Tools
    • Requirement Doc Templates
    • Quality Checking Guide
  • Blogs
    • Data Science
    • R Programming
  • Contact
  • Home
  • Books
  • Free Tools
    • Requirement Doc Templates
    • Quality Checking Guide
  • Blogs
    • Data Science
    • R Programming
  • Contact

Book Review: "The Good Jobs Strategy" Would Be Good for Data Science Too

5/28/2020

0 Comments

 
​I recently read a book called the Good Jobs Strategy. It was written by an MIT business school professor named Dr. Zeynep Ton. In the book, she explains why retailers who invest more in their employees outperform their competitors.

These companies, she says, are following the “good jobs” strategy.
Picture
I highly recommend reading the book, if your organization is investing more in data. Much of her advice easily translates to the major challenges that face organizations trying to build enterprise wide data solutions.
Dr. Ton’s core argument is that retail success depends on good operations. Good operations require dedicated employees. And dedicated employees require good pay and good training.

To make her case, she uses four companies that she calls “model” retailers. Those companies include:
  1. Trader Joe’s
  2. QuikTrip
  3. Costco
  4. Mercadona (a company in Spain)

The employees at these companies are well-compensated and well-trained. That leads to greater loyalty and commitment to the company and the company’s strategy. They then deliver good service on a consistent basis and manage in-store operations in a way that minimizes loss.

I actually worked for QuikTrip, one of these model retailers, in my early twenties. To this day, it’s the most well run organization I’ve ever seen. It also paid better than my first job after college. And most surprisingly, I received more on-the-job training there than any other job I've had.

Zeynep Ton explains that many retailers ignore employee investment in favor of other strategies. They focus on delivering low prices, offering a large variety of products, and emphasizing employee friendliness to customers.

She doesn’t recommend that be the primary focus of any retailer.

Customer service, she says, isn’t about how friendly the employees are to customers. It’s about making sure the shopping experience is consistently pleasant. That means customers can easily find the product they want. It means prices are reasonable. It means they don’t have to wait too long to check out. And it means that getting in and out of the store is painless. Employee politeness is important, but it won’t make up for failure everywhere else.

If a store fails to deliver this experience enough times, customers simply stop shopping there.

Companies can improve this experience on a consistent basis by focusing on operations and paying their employees well enough to deliver on those operations.

QuikTrip, for example, would have a rule of “no more than three customers in a line.” If a fourth customer enters a line, another employee would have to come to the register to help ring-up customers. That’s one simple and consistent process the company has to ensure customers consistently get in and out of the store quickly.

Companies like Wal-Mart, Home Depot, and the now defunct Borders are the opposite of what she considers a good retailer. Unlike the model retailers, their performance in areas that matter to customers are inconsistent. 

Wal-Mart, says Ton, is very good at logistics. It did a very good job at organizing and delivering products to their stores, which gave them a competitive edge as they took over American retail. But they now consistently drop the ball once the products arrive at the store.

Wal-Mart often has empty shelves. Their inventory system will read that they have products in stock, but the product is misplaced, lost in the backroom, or was never delivered to the store to begin with.

Finding Wal-Mart employees to help you inside the store is a challenge. There aren’t enough people working to help customers. When you look at their registers, most are empty and the check out lines feel like they take forever.

I stopped shopping at Wal-Mart years ago because of that. I loath going inside Wal-Mart now.

None of that is the fault of the employees, says Ton. It’s because these companies didn’t invest enough in their employees and didn’t build an operation that allowed the employees to perform well. Wal-Mart employees have high turnover and the employees often report being overworked. Much of that is because of pay. The rest is due to poor operations.

Her book has many better examples and gets more into the practical implementation of her strategy. It would be too much to summarize in one article, but the key points include:
  • Good operations matter
  • Good pay incentivizes employees to stay with the company and perform well
  • Less is more in terms of products, which makes inventory control easier and helps employees and customers know exactly what they sell and where to find it
  • Standardize as much as possible, which you can see this in the store designs of Costco and QuikTrip
  • Empower employees to use their judgement in non-standard situations
  • Cross train employees in-favor of generalization as opposed to extreme specialization (employees who can stock products and run the register vs employees who only stock products or run the register)
  • Operate with slack with regards to employee schedules

Why This Book Is Also a Good Guide to Building a Successful Analytics Practice

Zeynep Ton’s philosophy has a similar premise to my upcoming book, How to Manage a Successful Data Team. 

In the same way retail stores have trouble ensuring every product is stocked where it’s supposed to go and scanned as it’s checked out, data solutions require many different people in different parts of the organization completing key steps on a consistent basis.

In other words, good data solutions stem from good operations. And good operations require low turnover and well trained employees to deliver quality. It’s hard to accomplish one without the other.

If you replace Ton’s focus on delivering consistent service with delivering consistent, high quality data solutions, her methods often apply.

What would qualify a “good job” strategy within the analytics profession?

The most obvious thing that comes to mind is pay. You should at least pay the market rate to reduce turnover. I’ve discussed in a previous article with a technical recruiter who says below market pay will prompt employees to leave, even if they like their boss. You should also be more transparent about pay as well, which has been shown to make employees feel their compensation is more fair. 

Offering more training and cross-training will improve employee performance and also incentivize them to stay with your company longer. Many employees are more likely to stay with a company if they see personal growth there. (That’s one of the few things you can offer that actually outweighs pay)

I personally go back and forth on employee specialization. Does having one database developer focused on ETL, one report developer on reports, and one analyst on analysis really improve efficiency? 

I don’t think it does. Vacation time and sick days cause road blocks further down the pipeline if your individual team members are overly specialized. More cross training between these skills would allow employees to feel they’re growing and improve efficiency.

As Dr. Ton points out, focusing exclusively on employee pay and training really isn’t enough though. Like retail, data solutions (especially at the enterprise level) really are dependent on operations and finding ways to make sure both one off projects and support tasks are consistently completed.

My favorite example from the book is when Zeynep Ton writes about her conversations with Nichan Bakkalian, the head of processes at Mercadona. He claimed they recently put out a new placement offering stew vegetables near the stew meat in every store.

Zeynep Ton was skeptical that the placement had been set up in every store. In Dr. Ton’s past research, she found retail stores typically achieved this for 50% of their locations. She found that Mercadona had achieved it a 100% of the time.

When I read that example, it sounded like the same issue with setting up campaign tracking at a marketing agency. While it’s considered a normal process and something that’s required for each new campaign, it’s rarely close to a 100% execution rate. (See my article about operation issues in marketing analytics)

Companies that focus on operations can improve the execution rate, but they have to actually verify that it was done. And management has to clearly indicate it matters to them. That involves building checks and balances that allow people to follow-up and making sure it’s known by the key people involved that it’s part of their job responsibilities.

I recently had a conversation with the co-founder of a big data consulting firm who told me that data quality is an issue at most companies he works with. He said that most executives will state “oh, we have good quality data.” Once they dig in though, the executives are surprised that it’s not as good as they say. Had they known earlier that quality is directly related to operations, they could have improved their data sooner.

The other way this consistency is important is in the more routine work.

QuikTrip, for example, wanted to focus on making customers feel welcome in every transaction. So their employees are given a customer service bonus based on whether they say “Thank you, see you later!” at the end of a customer interaction. They also have to greet the customer when they walk in the door. They use secret shoppers to measure this. It improves the overall customer service bonus for the store employees and the individual employee at the register gets $50 if they say everything they’re supposed to say.

Employees do this without even thinking about it after a while because the company trains them the first week to say it. It then consistently rewards them for doing it. Shortly before I left the company, they had started offering trips to Hawaii as a reward for perfect secret shopper scores. It was amazing how customer service scores improved after that new incentive.

This training followed by a reward helps transform processes into ingrained habits. In the same way QuikTrip focuses on saying “Thank you, see you later!” and mopping the bathroom floors every thirty minutes, data teams can focus on building habits around requirements gathering and quality checking. That will go a long way to delivering a better experience for the stakeholder.

Some people think this means turning the team members into robots. I won’t lie, that did sometimes feel like the case at QuikTrip. When everything is standardized and routine, it does get a bit boring and you begin to feel like a cog in the machine. (I think I was too analytical for a job like that. The very extroverted people I knew loved working there and never got bored.)

Zeynep Ton acknowledges the boredom that could result. She emphasizes the need to standardize and empower though. Even though you’re standardizing some aspects of the job, you’re giving employees more leeway to use their judgement elsewhere.

If you set up processes correctly for data teams, the team members are empowered to use their judgement and creativity on their projects. That means they get to make the recommendations to the clients, choose the visualizations they want to use, and program the way they want to program.

The processes are simply meant to reduce data quality issues and limit the time spent in development hell because there wasn’t enough structured discussion early on in the discovery phase. Doing that means your staff can focus on the fun stuff and not chase down every data quality issue found or updating a project because a key requirement was forgotten.

Is the Good Jobs Strategy the "Best" Strategy?

​Zeynep Ton’s book is advocating for the good jobs strategy. Because of that, she really doesn’t acknowledge other strategies that seem to work in the retail space. 

Wal-Mart clearly doesn’t follow the good jobs strategy. But it is still the largest retailer in the United States. Revenue-wise, it still outperforms Amazon. I don’t know what Wal-Mart’s strategy is, but it seems to work for them.

To her point though, people like me don’t shop at Wal-Mart because of the very poor customer service experiences she outlines in the book. Even though they’re the most successful retail company, they could probably be larger if they followed her strategy.

One thing service oriented companies can do that’s more important than the good jobs strategy is real estate, which Zeynep Ton doesn’t mention.

For example, let’s compare McDonald’s to In-And-Out Burger. McDonald’s does not follow the good jobs strategy, but they’re very good at real estate placement. They make a lot of money because of it. In-And-Out Burger does follow the good jobs strategy and they had one of the first drive-thru restaurants. It is far smaller than McDonalds, despite doing what Zeynep Ton recommends.

I remember when I worked at QuikTrip, they explained to us how store location accounted for 75% or 90% (I can’t remember the exact number) of a store's success. Even with their amazingly well operated business and their employee-centric strategy, the majority of their success is still based on store location.

7-11 does not follow a good jobs strategy, but they seem to do well for themselves. They were founded only a few years before QuikTrip. While 7-11 probably doesn’t make as much on a per store basis as QuikTrip, they do probably make more than QuikTrip in total revenue. I couldn’t say for sure since QuikTrip isn’t a publicly traded company.

I recall reading a book by QuikTrip’s founder Chester Cadduex. I believe he did admit that paying well meant they couldn’t grow as fast as they wanted. But he said it was more a question of morality to him. 

That’s probably the ultimate thing the good jobs strategy has going for it. It’s the moral thing for a company to do. It’s a recognition that there’s dignity in every job. Even those in the service industry.
0 Comments



Leave a Reply.

    ABOUT

    A blog about the non-technical side of data science.

      SUBSCRIBE

    Confirm

    ARCHIVES

    March 2022
    October 2021
    September 2021
    August 2021
    March 2021
    February 2021
    January 2021
    November 2020
    August 2020
    June 2020
    May 2020
    April 2020
    February 2020
    January 2020
    December 2019
    November 2019
    October 2019
    September 2019
    August 2019
    July 2019
    June 2019
    May 2019
    March 2019

    RSS Feed