Driving Profitable Growth: Marketing Best Practices

White Paper

When thinking about software deployment strategies, you have a number of options. The first is your project scope. Are you going to roll out a new technology across the whole organisation, or across one or more divisions to build experience? Our experts would encourage you to pick a single channel, business unit or customer segment. Learn first, then deploy at scale. No matter how great the technology, there will always be something that needs adapting - whether it’s the technology or the perceptions of others.

Get the download

Below is an excerpt of "Driving Profitable Growth: Marketing Best Practices". To get your free download, and unlimited access to the whole of bizibl.com, simply log in or join free.



When SAS asked marketing thought leaders and pracitioners to weigh in on the issues and capabilities required for driving profitable growth, four main issues emerged:

  1. Knowing your customer.
  2. Not all customers are profitable.
  3. Many marketers struggle with complexity.
  4. Profitable growth is everyone’s responsibility.

We explored these issues from both the thought leader and practitioner perspectives in the white papers Driving Profitable Growth: The Thought Leader Perspective and Driving Profitable Growth: The Practitioner Perspective. In the paper Driving Profitable Growth: Technologies Available to Marketers we delved into topics like following your intuition, using pen and paper, advanced analytics and solutions designed specifically for marketers.

In this paper, we discuss best practices and lessons learned, covering:

  • Deployment strategies.
  • Considerations that touch on skills, culture and organizational structure.
  • Approaches to avoid.
  • Approaches that exceeded expectations.

Those interviewed for this white paper series include:

David Reibstein – William Stewart Woodside Professor and Professor of Marketing at The Wharton School of the University of Pennsylvania and co-author of Marketing Metrics: 50+ Metrics Every Executive Should Master.

David Chong – Senior Vice President of Marketing at a leading Malaysian bank.

Helena Schwenk – Principal Analyst at MWD Advisors, a European IT advisory firm that consults with organizations to create tangible business improvements from IT investments.

Jim Foreman – Director of Circulation and Analytics at Staples Inc.

Charlene Li – Founder of Altimeter Group, a leading research-based advisory firm with a focus on disruptive technologies, and the author of The New York Times best-seller Open Leadership.

Warren Murray – Head of Business Decision Support at eBucks, South Africa’s leading multipartner rewards program.

Deployment Strategies

  • Focused scope.
  • Progression or maturity.
  • Test and learn.

Focused Scope

When thinking about software deployment strategies, you have a number of options. The first is your project scope. Are you going to roll out a new technology across the whole organization, or across one or more divisions to build experience?

Our experts would encourage you to pick a single channel, business unit or customer segment. Learn first, then deploy at scale. No matter how great the technology, there will always be something that needs adapting – whether it’s the technology or the perceptions of others.

When Staples implemented SAS Marketing Automation, Jim Foreman, Director of Circulation and Analytics, said the launch was divided into two phases. The first focused on an external prospect database hosted by a third party. The idea was to start small, and identify tips and tricks that would make things easier for phase two, which would involve the larger internal customer database. In short, it was treated as a dry run or experiment. It took six to eight weeks.

According to Foreman, the biggest benefit wasn’t tweaking the physical implementation, but the improved confidence related to intangibles such as confidence in the delivery team, business and subject-matter knowledge, and confidence that everything was anticipated and planned out.

Progression or Maturity

There’s another classic consideration here that’s akin to walking before you can run: If you haven’t covered the basics of data quality, it doesn’t matter how great your analytics is. Don’t forget the saying, “garbage in, garbage out.”

Helena Schwenk, Principal Analyst at MWD Advisors, looked at implementing analytics in terms of digging deeper, one layer at a time. Once you obtain clean data and have mastered basic analytics, you can move on to predictive analytics. If you have a good handle on structured data, expand into unstructured. If you are using a limited number of data sources, evaluate and plan to use more. The focus is on improving your understanding of behavior and getting to more accurate decisions in real time.

Do everything at once and you risk overload. Go from basic to super-advanced in a single step, and you risk confusion. A CMO of a UK bank put that last comment into perspective. He said that to truly make use of any technological insight, your staff must understand how and why that insight was created. And to do that effectively, they need to master various levels of sophistication over time. Each progression informs, educates and opens the mind to asking better questions – an essential quality required of any great marketing analyst.

Charlene Li, founder of the Altimeter Group, used Dell as an example of the amount of upfront work to be done. It made listening to customers a strategic issue back in 2005. Over the next six years, Dell explored what would work, what would fit within the culture of the organization, what kind of training would be required and how it would change the decision-making process and follow-up actions. All this exploration was completed before deploying the technology it has today. The process involved asking many questions and exploring many possibilities. In short, Dell mapped out the evolution and secured buy-in before deployment. Technology is the easy part.

Test and Learn

Whether your deployment strategy focuses on scope or progression, the key is to use the test-and-learn approach that David Reibstein, Professor of Marketing at The Wharton School, mentioned in the first paper of this Driving Profitable Growth series: Prove what moves the needle, one change or experiment at a time.

His approach resembles classic change management theory. John Kotter, an expert in change management at Harvard Business School, describes an eight-stage change process that can be applied to any change initiative. One of the most important stages is the need to demonstrate short-term wins. As Kotter puts it, there will always be someone on the fence, or someone who believes his or her project is more important than yours.

In the absence of evidence or facts that prove success, they will try to put a damper on your initiative so money and resources can be transferred to them. Beat them to the punch and prove success sooner rather than later.

You might be surprised at what you discover. Warren Murray, Head of Business Decision Support at South Africa’s eBucks, explained how eBucks used this kind of approach to determine which products to make available in its online store, and how it should engage specific customer segments.

Instead of promoting and stocking thousands of items, it now focuses on a relatively low number targeted at specific segments. Its “test and learn” loop proved that only 35 percent of spending came from the Web store – the majority of products were purchased in a partner’s bricks-and-mortar store.

Warren also talked about using the test-and-learn approach to recruit and retain partners. Before adding a partner, his team performs a sales analysis – effectively creating a benchmark. Once a partner is on board, the team reviews its performance to demonstrate lift and assess changes in strategy and tactics.

Additional Considerations

  • Culture.
  • Skill sets.
  • Organizational structure.
  • Communication.
  • Experiments and failure.
  • Measurement, value and ROI.

Want more like this?

Want more like this?

Insight delivered to your inbox

Keep up to date with our free email. Hand picked whitepapers and posts from our blog, as well as exclusive videos and webinar invitations keep our Users one step ahead.

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy

side image splash

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy


It all comes down to this: Without setting the right culture or tone from the top, you won’t get the best lift on your investments. Your initiative will take longer to implement and you will end up spending far more time dealing with internal politics.

Charlene Li believes culture is the biggest obstacle for most organizations – particularly when it comes to embracing social marketing. Think of the traditional definition of culture as a sort of “this is the way we do things around here” approach. That kind of thinking comes with certain norms about how you get things done, how you make decisions and how you solve problems.

If the culture in your organization is driven by personality and gut feelings, introducing analytics to help generate data-driven decisions will be a challenge. Don’t expect people to suddenly change purely because a new technology is implemented.

And when it comes to “tone from the top,” remember that actions speak louder than words. If your executive team members don’t change their behavior, it won’t matter how often they tell others to change. No one will believe them.

Gary Loveman, CEO of Caesars Entertainment, is famous for setting the right tone.

He says there are only three ways to be fired from his company: theft, sexual harassment and running an experiment without a control group. (This type of thinking is music to David Reibstein’s ears.) When your CEO demands that employees operate the business by analyzing data rather than leaning on hunches – and publicly talks about the consequences of ignoring the demand – you know analytics will be taken seriously and become part of your culture.

Skill Sets

Over the course of our interviews for these papers, the participants uncovered three basic qualities needed for analytics implementation. The first referenced a person’s ability or willingness to change and adapt. The second, as Staples’ Jim Foreman defined it, is a person’s analytic curiosity. The third was a person’s understanding of the core business. The rest of the needed skills can be taught or brought in from the outside.

Think of the first skill as table stakes. Based on the amount of disruption marketers are witnessing, your team will need to be comfortable with change. As you grow your team, don’t hire people who like doing things the old or traditional way. Look for diversity, creativity and an open mind.

Charlene Li would also remind you that change takes time. You can’t read a book and be an expert right away. Think about one of the challenges associated with social media, such as assessing sentiment.

The statement “that is just great” could be a positive or a negative statement. Tone of voice and context are needed to interpret correctly, and that requires a person’s judgment, which represents the culmination of years of experience. Technology gets it right about 75 percent of the time, but you will always need people to challenge and adapt the rules – a good segue to Foreman’s notion of analytic curiosity.

Everyone interviewed for this paper series referenced improvements in technology. But you don’t need an army of statisticians. The technology does much of the heavy lifting for you, and as Foreman put it, opens the first few doors to explore before you get into the deeper analytics.

What is more important than technology is a person’s analytic curiosity, or his ability to ask better questions and demonstrate an understanding of logic – the kind of person who doesn’t simply deliver a report, but questions results, looks at the bigger picture, and seeks to understand and explain what created the results and why.

Combine that curiosity with an understanding of the core business, and the rest can (quite literally) be taught. In fact, David Chong, SVP of Marketing at a leading Malaysian bank, said his bank would typically take someone from the business side of the company and teach them how to do the analytics rather than the other way around.

Organizational Structure

Three considerations worth reiterating:

  1. Move from a product- to a customer-centric structure.
  2. Remove channel and/or business unit silos.
  3. Consider creating a shared “analytic center of excellence.”

The first two in our list focus on putting the customer first, creating a better experience and removing internal competition.

The third recognizes that your analytical talent could be put to use across the organization, not just in marketing. In return, your analytic staff will get a deeper and broader understanding of how your business operates, and marketing (assuming this is where the team reports) will be seen as a stronger, more valuable partner: a win-win situation.


At some point, everyone learns the same lesson: They should have communicated more effectively (me included). If you have attended a marketing webinar or read a marketing book, you might be familiar with the concept of communicating seven times in seven ways before a message is truly understood and remembered. We start out with the best intentions, but how many of us effectively communicate? Let me rephrase that question. How many apply the same rules we use for external communication to our internal audience?

We all need to do a better job communicating, but getting our message across is not just about repetition – it’s knowing your audience, and ensuring relevance. The exact same considerations apply for marketers when it comes to interacting with customers. But how often do we apply that lens when communicating internally?

One stakeholder who came up time after time during the interviews for this white paper series was the CFO. This individual ultimately approves all corporate financial investments. Approval from CFOs is critical to almost everything you do, so keep them involved and updated, and use their analytical talent.

Another commonly referenced group can be described collectively as “whoever you impact.” That may be people in your call center, retail outlet staff, brokers or sales people, the folks in logistics – the list goes on. Involve all of them sooner rather than later. Don’t wait until everything is perfect or complete. Also, don’t assume that once your message has been communicated, it will be understood. Repeat it. Test it. Prove that your message was received and understood.

Experiments and Failure

An underlying theme thus far is that the market is changing and marketers need to adapt. That means forming a hypothesis of what you think needs to be done, then doing it.

If you define it as “doing something new,” your hypothesis could contain a number of educated guesses. That’s OK; you have to start somewhere. But over time, through a series of disciplined experiments, you need to get better at predicting outcomes. Technology can help, but the reality is that there is inherent risk, particularly in the early days when you are attempting to find the right data to prove what works.

An article in Harvard Business Review3 reported that the most successful entrepreneurs had redirected their strategy at least five times before they hit a solid growth trajectory. All devised low-cost experiments to disprove a concept before it was too late. Our experts could all relate to this point since their test-and-learn loops are all a series of experiments.

Chris Trimble, co-author of The Other Side of Innovation, goes further. He says organizations need to put learning first and profits second. He argues that organizations focus on getting a good result, which is a strategy that can either be a success or (as the HBR article points out) a quick and inexpensive failure. Unfortunately, many treat failure as a negative, unlike David Reibstein who sees it as a cause for celebration

If you don’t create an atmosphere where it’s safe to fail, few will ever risk trying anything new. If they don’t try, you won’t learn and you risk competitors getting smarter and more profitable. The key then is to not judge based on the outcome alone, but by your ability to run a disciplined experiment and adapt to the insights gained.

That’s a good segue to the next topic: measurement.

Measurement, Value and ROI

When it comes to measurement, it gets very personal. We are not going to attempt to articulate the many and varied ways of doing it. Rather, for this section we pause to think about how organizations measure and justify investment.

Both David Reibstein and Charlene Li firmly believe measurement should be defined in terms of affecting customer lifetime value (CLV). As we covered in our first paper of the series, there are challenges arriving at those figures, but the general idea is that we should be asking a simple question: “Are our actions improving or reducing CLV?”

Charlene Li said too many of us are hung up on measuring the return on investment (ROI) of a campaign or installing a new technology. Those figures provide a good snapshot for an activity or moment in time, but fail to reflect improvements related to your competitive capabilities or mindset – something far more valuable.

For example, one of our customers used SAS® High-Performance Analytics to score its customers in 60 seconds versus 4.5 hours with our regular analytics.

How do you calculate ROI? The results were the same. They just received them faster. Do you calculate ROI based on processing costs, your ability to process greater volumes, scoring your customers more frequently, reduced staffing costs, or something else?

How you spend the additional 269 minutes saved with high-performance analytics will ultimately determine your true ROI. But how many attempt to incorporate the value of “knock on” activities? How many can quantify or even correlate the relationship? You’ll never get a perfect answer, but if you don’t give it some thought you could be seriously underestimating value.

Li recommends taking the higher ground, and in the above example, simply asking, “If we can do it faster, how much more market share can we potentially attract?” If your initiative or actions improve customer satisfaction, ask “How will that affect CLV in the medium to long term?”

The lesson is to not be stuck in the present, and to think bigger. Marketing is about driving profitable growth today, tomorrow and well into the future. Measure the trend or progress on that dimension.

Approaches to Avoid

Perhaps we should call this section “approaches to think twice about.” Not all of them can or should be avoided, but they’re all worth exploring.

  • Logic versus loyalty.
  • Logic versus profitability.
  • Changing the rules during a recession.
  • Underestimating the importance of training.
  • Expecting one tool to do everything.
  • Preprogrammed pitches.

Logic Versus Loyalty

David Reibstein shared three examples that sound logical, but may not be the best incentive for loyalty.

The first was about an approach used by Amazon. New prospects or infrequent customers are presented with offers at the “best” price. Loyal customers, those who spend frequently, pay full price. As an incentive to generate new business, or improve frequency, the strategy sounds logical – until your best customers realize what you are doing. At that moment, your best customers may get a little upset. They may be thinking, “I spend so much money with you and you treat me like this?”

The second example comes from Coca-Cola. It used vending machines that were temperature-sensitive. When temperatures went up, so did the price. Again, sounds logical when you consider machines are likely to run out of inventory fast when it’s hot, creating a need to increase the frequency and cost associated with replenishing stock. But from a customer perspective, what message or feeling does it create?

The third example Reibstein used was about pricing in New York supermarkets. During the day, when everyone is at work, prices are cheaper. During the evening, prices go up. The logic is that those shopping during the day probably have more time to do price comparisons and pick the cheapest. Those rushing home from work just want to get home and will pay whatever is asked, within reason.

All three of these examples sound logical from a corporate perspective because all of them are designed to improve profitability based on demand. Do they seem logical when you think about the experience and emotion created within your customer’s mind?

As Reibstein pointed out, we don’t get terribly upset when we see price changes with the airlines because every seat has a different price. Hotel room pricing works similarly. But there is a risk that this pricing strategy could hurt your reputation – especially because people are increasingly sharing their thoughts via social media. It’s worth questioning which is better: making an extra dollar today, or keeping customers loyal to protect revenue tomorrow.

Logic Versus Profitability

In this example, we challenge basic assumptions. A Canadian wireless carrier assumed the accounts of customers signed up for its $100-a-month, “all you can use” plan were profitable. Using standard accounting practices, the logic it adopted was that if you applied the standard costs to usage, there was always margin left.

But as we highlighted in our first paper in this series, 20 percent of customers could be destroying up to 400 percent of profit. This could be blamed on the occurrence of nonstandard activities, or perhaps “standard costs” are simply averages, meaning some cost more and some cost less.

When taking a closer look at this specific offer, our wireless carrier discovered that 25 percent of its customers cost $200 a month to service. If marketing targeted other customers with similar characteristics, it doesn’t take a genius to realize it would soon go out of business.

The lesson is to never assume that standard accounting logic is perfect. It may stack up in aggregate, but when you look at specific customers or segments, things could change. As marketers get closer to a customer “segment of one,” it pays to sweat the details and challenge basic assumptions.

Changing the Rules During a Recession

During the first wave of the 2008 recession, Staples decided to pull back on acquisition efforts and focus on retention. As Jim Foreman put it, Staples wanted to identify its best customers and determine how to keep them happy and engaged. In other words, Staples went into defensive mode versus maintaining its proven ratio for growing the business.

Sounds logical because everyone knows it costs more to acquire a customer than retain a customer. Given the recession, why not focus on retention? There were two basic lessons learned here.

The first was a realization that no matter what you do, some customers will churn. This churn could be caused by a hundred different events, but it’s difficult to predict. And without an acquisition plan, customer churn leads to a shrinking customer base.

The second lesson learned was that incentivizing loyalty among your best customers via discounts or coupons effectively undermines your margin on products they would have bought anyway. They’ll take the discount, but won’t necessarily grow their basket – particularly during a recession.

Add the two lessons together, and unless corrected, your situation will get worse quickly. Fortunately for Staples, it implemented a change using the test-and-learn approach, which surfaced mistakes quickly.

Underestimating the importance of training

According to Charlene Li, one of the trickiest aspects of deploying technology is making sure you have the right training and structure aligned to achieve business goals, particularly when moving to a conversational marketing approach.

For example, Li mentioned an organization that had deployed a listening tool very quickly. It told its people, “This is powerful. This is a great way for us to engage with conversations. Please use it.” The organization didn’t tell them how to use it. It didn’t tell people about the constraints and issues they should consider. It was the equivalent of being given a Ferrari as their first car. They had no knowledge of how to use the power, and something very dangerous was now in the wrong hands.

Over time, the organization realized that people were using small bits of information. They were drawing their own, incorrect, conclusions. They made decisions without fully understanding the implications.

The lesson is to think through the basics. How are people going to use the new information? How are they going to share it? Who is going to have access to it? How will it fit within existing processes? And how will you ensure consistency and alignment? With those questions answered, put together a training program to ensure everyone is on the same page and can fully realize the potential of your investment.

Expecting one tool to do everything

The challenge is that few people prioritize needs. They expect to squeeze every last drop of value from their technology investment, whether it makes sense or not.

For example, Li said that you might purchase a good “scraping tool” to listen to what is being said within social media. It could be good at capturing everything under the sun, but hopeless at delivering strong analytical insights. Conversely, a great analytical tool may not capture everything under the sun. Expecting a single tool to deliver the whole is a flawed position to take.

What you need to do in this specific scenario is prioritize what you are going to listen to and monitor. Understand what part of a relationship you are trying to improve. Once that is clear, choose a solution, suite or combination of tools that deliver on both counts.

Preprogrammed pitches

Most marketers talk about the importance of segmentation and understanding customers, yet how many times has someone tried to sell you something that isn’t relevant or timely? Or, perhaps you were open to a sales pitch but found no one to deliver it. Happens all the time – right?

Helena Schwenk shared two personal stories that highlight these mistakes.

The first referenced her experience contacting her bank. She was having trouble logging into the bank’s website. After a help desk associate fixed everything, he tried to sell her a savings product. While Schwenk was a good candidate for the product, the timing and context wasn’t right. It was taking too long to log into the bank’s website, so the act of selling something totally unrelated to her task was an annoyance.

The second example she shared, which was triggered by her bank encounter, involved contacting another bank to discuss savings products for her children. She completed the transaction and set up an account – mission accomplished! Later, Schwenk reflected that the bank associate didn’t ask if he could help with long-term financial planning needs (a logical cross-sell opportunity). She was already thinking about savings and had reached out to the bank. The timing was right, and that sales pitch would have been relevant.

In both situations, the employees were probably working from a static script full of preprogrammed pitches. Get the timing or context wrong, and you risk losing a customer. Fail to identify and adapt next-best actions in real time, and you miss an opportunity that may be difficult to introduce later. Technology can help. They key is to embed real-time analytics (sas.com/solutions/crm/real-time-decision) into your processes to improve decision management.

Approaches That Exceeded Expectations

  • Understanding behavior.
  • Knowing your customer and consumer.
  • Empowering employees to reach out via social media.

Most of the following examples explore the use of analytics to identify and explain hidden patterns in your existing data. The message is clear – constantly challenge existing assumptions, search for the unknown and don’t fear experimentation.

Understanding behavior

When dealing with major change (think recession), it pays to challenge existing assumptions. Many organizations classify or segment their customers based on channels or information collected via reward programs, online account profiles or surveys. They use this information to select the most appropriate offers or methods for marketing. Yet some of those customers may change jobs, move or alter their traditional behavior. And, they may or may not tell you about it directly.

Using behavioral modeling, organizations can analyze existing assumptions to test whether a customer’s classification (e.g., business versus consumer) still holds true – and adapt what they offer via which channel(s).

Another approach is to challenge the notion of looking to your best customers for additional revenue. Use analytical models to identify customers who are less engaged or responsive. By focusing relevant campaigns and discounts on this segment, you could improve relationships, increase frequency of transactions and reduce churn.

David Reibstein shared an approach used by Harrah’s Casino (a longtime SAS customer). Harrah’s recognizes that everyone is different and applies extensive testand-learn experiments that analyze how people respond to different types of incentives. It scientifically looks to prove what drives additional loyalty and revenue – customer by customer – with great success.

Schwenk mentioned a telecommunications provider that analyzed why customers were late in paying their bills. Rather than being bad customers, or financially constrained, it found a large segment that had customer service issues that were not being addressed in a timely or acceptable fashion. Late payment represented more of a protest. By addressing those issues, it drove multiple improvements that reinforced each other, such as improvements in payment, satisfaction and retention.

Knowing Your Customer and Consumer

Li had worked with an online retailer focused on teenagers and women in their early 20s. After analyzing traffic across Facebook (people who had given appropriate permissions), this retailer discovered a group of women in their 50s who didn’t match any of its target customer profiles.

On further analysis, they turned out to be parents and grandparents who were buying for a younger generation. With that understanding, it created campaigns that would better appeal to them. By changing the tone, language and approach, the retailer was able to grow revenue from this group with great success.

Empowering Employees to Reach Out Via Social Media

This was another personal story – this time from Li. She wanted to buy her son a smartphone but wasn’t sure which one to purchase. Concerned about where he could go online, she posted a question to Best Buy’s @Twelpforce account on Twitter asking for recommendations.

More than 2,500 Best Buy employees have been empowered to answer questions via Twitter. They are not monitored by corporate. The answers aren’t scripted by marketing. They are genuine answers from employees who have volunteered to help.

An employee wrote back and recommended she look at an Android phone because it had some really nice programs and capabilities. Li had previously allowed Twitter to display her location, so the Best Buy employee asked, “Why don’t you come down to the local Best Buy store and bring your son with you so that we can look at these phones together. I’m available Sunday afternoon. What do you think?” Li said her “jaw dropped to the ground” because she never imagined a local employee would answer her question.

@Twelpforce had more than 40,000 followers as of December 2011. Best Buy’s employees had contributed more than 54,000 tweets. As Li puts it, Best Buy has made a successful transition to conversational marketing. It has more than 2,500 employees actively joining the conversation with relevant information at the right time and place – a marketer’s dream.

Closing Thoughts

The first best practice that stands out in my mind is to constantly challenge assumptions. The market never stands still, so why should you? The second requires humility, because on the way to greatness, you’ll make some mistakes. How you deal with failure is a key barometer of success. So our experts recommend you run disciplined experiments to prove what works quickly, then adapt, learn and move on.

Finally, technology has come of age. Whether your needs are small or large, simple or complex – there’s a solution just around the corner. There’s no silver bullet, but in SAS’ experience, the “Pareto principle” comes into play: 80 percent of results are created by 20 percent of our actions. SAS solutions have been engineered based on worldclass best practices that meet the needs of 80 percent of marketers. For the other 20 percent, the same solutions can be customized and adapted to reflect your own unique processes, or push the boundaries of your analytic curiosity.

Want more like this?

Want more like this?

Insight delivered to your inbox

Keep up to date with our free email. Hand picked whitepapers and posts from our blog, as well as exclusive videos and webinar invitations keep our Users one step ahead.

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy

side image splash

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy