5 Reasons Why Marketing Segmentation Projects Fail

Posted by: Mark Price | On July 18, 2014

It’s an uber-personalized world out there.  Which means customer segmentation is the name Customer Segmentationof today’s marketing game. But if segmentation is so essential, why does it fail over 50% of the time?

First a quick review.  Segmenting your customers means dividing them into smaller groups that are more alike than the rest of your customers.  This process yields critical insights about your best customers and helps you identify other segments that can become best customers over time. With this information, you can customize messaging and provide offers that appeal to smaller, more valuable groups, increasing engagement, retention and ultimately customer value. While customer segmentation has been accepted as a critical marketing tool for over 20 years, more than 50% of these efforts end up in failure – wasting hundreds of thousands of dollars and much of your sponsor’s goodwill.  Make sure your customer segmentation initiative is a success by avoiding these five common roadblocks and landmines:

  1. No strategic goal
    Good journeys begin with a clear focus on the destinations.  If you do not know the overall business driver behind segmenting your customers, you will not succeed. Many companies start a segmentation project because someone else did it. But what is the strategic need in your company — is it more efficient lead acquisition or reduced attrition of best customers?  Do you know who your best customers are (not your highest revenue, but your best)?
  2. Lack of buy-in from the teams responsible for sales
    Ultimately, the “peddle hits the metal” when someone has to sell something to someone else.  If the teams that sell to your customers — whether web site development, Sales or telemarketing — will not support a segmentation pilot, then you will not succeed.  Segmentation must reach customers to matter, and marketing cannot win this battle alone.
  3. Lack of an executive sponsor
    Eventually, all segmentation efforts face pushback from other departments whose goals and metrics are not the same as those of marketing.  The only way to ensure buy-in is to have an executive sponsor who can intercede with Finance, with the Creative team, with e-commerce, to make sure the pilot program can see the light of day.  An executive sponsor, in our experience, is critical to your success.
  4. No ROI/success parameters
    Another key factor contributing to segmentation failure is a lack of quantification of potential value.  What are your success metrics?  Without a forecast of potential upside, supporting groups (and potentially Marketing as well) can get sidetracked meeting their own metrics or senior management special requests.  Identifying the potential upside, based on company performance and best practices from similar companies, gets you through the hard times, when everyone else would like to abandon your initiative.
  5. Inability to link data to individual customers
    The most common mistake of all is a focus on an “attitudinal” analysis rather than a behavioral one.  This approach is useful to provide insight into customer beliefs and needs, but you can’t use it to target specific customers for database marketing efforts. It simply doesn’t work.  Build your segmentation on parameters that you want to change — behaviors, and you will be quickly on the road to success.

Yes, segmentation can be transformative to an entire organization.  Viewing your customers as groups, rather than a single “she” or “he”, allows the company to better tailor products, services and communications to your critical customer classifications. Since any customer segmentation effort can easily be derailed, keep these five factors in mind, and success will be on the way.


10 Reasons to Personalize Email Marketing and 1 Way To Get Started

Posted by: liftpoint | On June 20, 2014

It still surprises me today when I run into retail marketers who are sending the same email to all their customers.  They are the ones who are surprised at the low levels of their open and click rates.Top 10 Reasons to Personalize Email Marketing

While there are probably a 100 reasons to begin to personalize email marketing today, here are the top 10 reasons to start now. Remember, in digital marketing, time is not your friend!

  1. Your customers are receiving personalized emails from your competition.  As personalized emails become more common, consumers will begin to shift their spending to brands that seem to know them the best.  That better be you!
  2. Your customers don’t know many of the products that you carry. Often, even best customers do not know your entire product assortment.  By highlighting products that analysis suggests they might be interested in, you both intrigue and reward your better customers with more information.
  3. Your retail experience is largely unpersonalized. No matter how much training you give your store associates, regular customers are seldom recognized, and their service experience is often generic.  Email campaigns give you the opportunity to add a personal touch not available in-store.
  4. Your customers may be running out of your products right now.  You do not know what the consumption/usage patterns of your products are and cannot easily anticipate when customers may need a refill.  Every time they run out of a product, they have to make the choice to repeat or seek out other alternatives.  Help them make the choice to come back by letting them know that you know who they are and care about them.
  5. Email personalization does not require programming knowledge any more. “Back in the day,” you often needed a combination of HTML and SQL skills to execute personalized email marketing.  Today, the new generation of email marketing tools allow you to “point and click” your way to basic email customization (images, offers, text).  Don’t let software installation stop you, either.  Most of these new tools operate in the cloud.   You can get up and running quickly.  You can load your email list, host your images and copy and build your campaigns as you need.  In addition, you can download open/click files for further analysis as you wish.
  6. Your customers think you don’t really know who they are.  Why should they?  Do you?  If you knew that the patterns of a mere 10% of your customers would determine not only your bonus but whether you stay in business, how different would you act?  You can identify those customers with some basic analysis and then focus on them for your targeted marketing.  Begin to convince those customers that you know them by referring to their prior purchases and other things you know about them, and follow-up with other programs, both online and in-person.
  7. Your customers are angry that they don’t receive the discounts “the other guy gets”.  Why is it that companies give the largest discounts to prospects, who they don’t even know, and the lowest to their prize customers, the ones who “keep the lights on”?  By the way, if you think your best customers do not know what other customers and prospects receive in offers, you are wrong.  The Internet takes care of that.  Preempt those concerns with special offers for Best Customers, based on their unique needs.  That approach will help with retention.
  8. Bored customers try the competition.  “Hey, the other guy has some shiny new bells and whistles!  I think I will go try them out!” And there goes your Best Customers, trying the competition’s new offers or products while you send the same mass boring emails to them over and over again.  Use your customer knowledge to customize the offers in ways a competitor can’t.
  9. Personalized emails make you more money.  Over and over again, we see in our clients that personalized email subject lines and offers drive up to a 50% higher lift in open rates, click through rates and conversions.  Yes it takes more work.  Yes it will make you more money.
  10. Personalized emails contribute to customer retention.  Customers, simply put, want to return to places that know them, speak to them as individuals and would miss them if they were gone.  That is the goal of all marketing.  Personalized emails are one small step in that direction, but they are a clear step, one you can’t afford to miss.

You probably realize by now how critical personalized email marketing is to the growth of your business.  Now, how to start?

In our experience, the one place to begin with personalized offers is to offer customers exactly the same product category they purchased before.  That approach may be counter-intuitive, but in client after client, the last product category purchased shows up as one of the products categories the customer is most likely to purchase again, even for categories with longer purchase cycles.  So take the shortcut, and offer your Best Customers a product from the same product category as before, with personalized communications and subject lines.  Sometimes the easiest approach is the best.

Personalized email marketing is a critical link to retaining and growing relationships and business with your Best Customers.  Don’t wait to start.  Your competitors already have.

Data-driven Lead Scoring: What Makes a Best Customer?

Posted by: liftpoint | On May 28, 2014

Lead ScoringB2B Sales and Marketing  always seem to be disagreeing about lead quality.   Marketing claims that the leads are “highly targeted” and Sales claims that the leads aren’t accounts that can help them meet their revenue and growth goals.  Either they are too low in value, or not well qualified.  Lead Scoring can help resolve those issues, but carries some unforeseen risks.

Given that leads are the lifeblood for growth of B2B companies, what is the solution to this dilemma?  How do you get targeted leads that actually close?

One approach is to use lead scoring software.  These software companies provide a solution that allows you to “score” leads based on factors such as their stage in the buying cycle, SIC code and the prospect’s level in their company.  Those scores will help you find prospects that resemble your highest revenue customers.

But what if your highest revenue customers aren’t really your best today, or the best customers for the future of your business?

To identify the best prospects for the company’s future, you need to start with your past.  

  1. First, identify current customers who are “right” for the company as it grows.
  2. Then you use data about those companies to determine what they have most in common.
  3. Once you have found those common elements, it is not difficult to identify and rank prospects on those same factors.   

The hardest step is the first one — actually determining what customers you would consider to be “your best.”

To identify best customers, I recommend incorporating the following data:

  1. Depth of Product Purchases
    How many of your product lines does a given customer buy? Your BEST customer buys across your product offerings.  Customers, who buy a lot of the same product, don’t share that distinction.  By ranking customers according to range of their purchases, you begin to build a customer profile with an insight that a simple “who bought how much” analysis doesn’t give you.
  2. Retention
    How long has a given customer been your customer?  Chances are, the longer the relationship the better the customer.  By measuring the time between the first and last purchase, you gain valuable data in building your best customer profile.
  3. Loyalty
    How many channels does a given customer use to access your company?  Has this customer written positive reviews on social media?  By taking a cross-channel look at your customer database, best customers will rise to the top – providing another insight in fine-tuning your best customer profile.
  4. Strategic Fit
    How does a given customer fit with your company’s strategic direction?  Does a given customer’s demographic profile match your corporate “voice of the customer”?  What psychographic data can you append to your customer database?  If you’re selling tractor equipment, your best customers probably aren’t lawyers.

Combine all of this data and you’ll have a pretty clear profile of your best customer.  Analyze those customers to determine common elements — that is the key to the most effective type of prospect targeting.

Once you have determined the commonalities between those best customers, you can apply a lead scoring algorithm to rank your prospects based on those factors.

You see, the hard part of this process is not the lead scoring itself; rather, it is the thinking and analysis that helps you lock down who are your real best customers.  That insight can be company changing, both in how you treat your current customer base and how you target.

Companies who have used this approach have seen double digit improvements in revenue from new leads.  The insights from understanding what really makes a best customer also helps the company better target new products and services to that base.

So before you score your prospects, make sure you are scoring them on the right factors — your “true” best customers — those that meet the needs of your company now and for the future.  Otherwise, as they say, “be careful what you pray for, you just might get it!”

Read some of my previous posts about Sales and Marketing alignment:

Sales Alignment – An Essential Link to Build Marketing Credibility

Grow Your Marketing Credibility Before It’s Too Late 

Is Privacy the End of Database Marketing?

Posted by: liftpoint | On May 22, 2014

Why can’t customers and companies just get along?

These days, when there are so many stories about data security breaches and privacy, effective database marketing is at risk as well.  To succeed, marketers need to make the case, more convincingly than ever, why it is beneficial for consumers to share information about  preferences and interests.Privacy and database marketing

And the case is not hard to make.  We can make it from our personal experience.  How many times have you been swamped by web ads that have nothing to do with you?  We are graduating our last child from high school this year — please do not send me diaper ads!  I love to ski and play tennis — why are you sending me golf offers??

We are awash in irrelevant offers wherever we turn, yet  consumers balk at providing the very information that could improve their experiences and perhaps give them some valuable offers on things they are interested in.  When companies genuinely ask, consumers are often proud of refusing.

When companies try to make the case, their communication can be so full of legalese that it serves to frighten consumers even farther away!  It is more critical than ever that companies today, when asking customers to take a perceived risk (whether true or not), to be authentic, to talk as real people, one to another, to acknowledge the fears that consumers have and to make genuine promises.

Once you have made the promise, what do you have to do to succeed?  Keep on promising.  Like they say with shampoo, “repeat as necessary.”  And you should assume a lot of repeating will be necessary, with examples and support, to build up credibility that has been damaged for the marketing industry.

Companies interested in using database marketing to personalize communications need to keep making a clear, effective promise and then ask customers about their needs and interests.  Once a consumer has told you what they want, the best way to keep the relationship is to use that information, and not fall prey to the tendency to send out blitz offers in the futile hope that they may change their minds.

Companies are also at fault for not using their data about what a customer opens and clicks on to improve the communications accuracy for an individual customer.  Campaign management software makes such personalization possible today, yet even when a company has that software, they often return to a modified version of “spray and pray” — barraging customers to see if they will click on anything.

What marketers fail to do is to take the extra effort and test — actually do the hard work of email optimization and analyze clicks/opens and ask for preferences, and then customize their communications based on that information.  Sure it is a lot of work, but consider the possibility that the results may be strong enough to reprioritize your entire approach to marketing.  That is what Amazon, Overstock and other best practice ecommerce companies have discovered, and you could too!

If anyone wants to know, I am interested in Crossfit, tennis, skiing, cooking, theater, science fiction, travel, European soccer (Go Liverpool!), holistic health approaches and Jewish education.  PLEASE use that information to send me the right offers  – I can sure use some vacation discounts right now!   :).

To read some more about database marketing, please check out these posts:

Fight Corporate ADD with Rapid-fire Analytics

Are Marketers Responsible For the Last Mile of Customer Experience?


Fight Corporate ADD with Rapid-fire Customer Analytics

Posted by: liftpoint | On December 2, 2013

MotivationDo you feel like you can’t get anything completed before the next “corporate priority” derails the effort?  If so, you are a victim of “corporate ADD” — the tendency of senior management to change priorities based on personal experience, something they read or in response to a business change that may or may not be a trend.

The symptoms of those suffering from corporate ADD include rapidly switching priorities, forgetting recent conversations or commitments and a high level of excitement around the latest “silver bullet.” Those who are responsible for dealing with such individuals also suffer from a host maladies: anxiety, insomnia and PTSD among them.

Intervention with Corporate ADD must be done swiftly and with consistency, and customer analytics can be a big piece of the solution.  If you can show progress towards your executive’s key metrics, your project will stay on the “top burner” and not be shelved.

The problem that most marketers experience when attempting to intervene with customer analytics is the time lag. From the start of an initiative, marketers often have to wait months before receiving the performance metrics that will justify the work.  In that time, marketers are at risk of another episode of Corporate ADD, taking them off of what they know to be important and moving their focus and resources to another of the CEO’s pet projects.

What’s a marketer to do, to keep focus on the most important initiatives in the face of the ever-changing corporate direction?

That is where “rapid-fire” customer analytics comes in.

Rapid-fire customer analytics is an approach that recognizes that there is no time to waste in reporting and analyzing the progress of marketing initiatives.  Whether number of customers participating, total revenue driven, increased AOV, etc. — each of these metrics, while not telling the whole story, suggest positive results and keep executive attention until you can get a more complete picture of results.

The best companies using this approach employ the following strategies:

  • Development of a list of 4-6 key metrics that will determine the success of the initiative
  • Identify where that data must come from (source system)
  • “Yank” sample data off the source system
  • Calculate the necessary metrics using Excel or a similarly easy tool to use
  • Create a “dashboard” worksheet to highlight the key metrics
  • Determine necessary frequency, and pull the report 1-2 times before the program launches to make sure you get it right

This approach is the easiest if you have fewer than 1million rows of customer data.  If you do not, then use a sampling approach to get you started, all in Excel.

Note:  Do all of this before the initiative launches, in order to have things in place when time gets tight.

Repeat as necessary.

This approach is seldom used in marketing departments, where measurement is an afterthought, once the big push to get the initiative launched is completed.  But following the “begin with the end in mind” approach, rapid-fire customer analytics will lay the groundwork for a successful program that keeps going, rather than one that falls by the wayside in the next reprioritization.

Not only will rapid-fire customer analytics feed management with performance metrics before the next onset of corporate ADD takes place, but this approach will also give you an early window into program performance, which can allow early course corrections to improve performance.


Want to Raise Prices? Customer Segmentation Provides an Answer

Posted by: liftpoint | On September 4, 2013

Customer_SegmentationIn an article in the Wall St. Journal entitled, “How Companies Can Get Smart About Raising Prices“, Professors Paul Farris and Kusum Ailawadi suggest that many companies will be forced to raise their prices in the coming years, since cost of goods and transportation costs have been rising steadily while prices have remained flat to declining.  Cost savings have been benefiting these companies, but now the cost savings are running out.

Since so many customers have become price-sensitive in the past recession, what’s a company to do?

Some companies try to “hide” their price increases through a number of strategies: cutting promotions, reducing quality, reducing package sizes, etc.  But all of those strategies have been shown to backfire, since consumers have a sense of what is a “good value”, and are sensitive to “price gaming”, when a few consumers figure out “the trick”, they let EVERYONE know.  Then the company has to deal with a negative reputation to go with the pricing issue.

So how can you raise prices right?

Professors Farris and Ailawadi recommend that companies use customer segmentation to target their promotions to the right customers.  As they write, “After raising prices, companies should rely on discounting to keep their coupon-clipping customers—the ones most likely to jump ship if they think they’re getting a bad deal. That means taking a close look at who their customers are and who should get what promotions.”

You don’t have to tailor promotions on a 1:1 basis, “It’s enough to group customers into segments based on things like their purchase history and how sensitive they are to price.”

The Power of Customer Segmentation

In fact, promotional sensitivity is one of a number of factors that can be customized through customer segmentation.  Product preferences, communication channels, offers, timing, seasonality, messaging/positioning and others also become clear when you group customers by their transaction behavior (see my post “Customer Segmentation: Using Behaviors to Drive Data-driven Marketing” for a deeper perspective on behavioral segmentation).

How do you gain a deeper understanding of your ability to increase price or conversely, reduce promotional activity, at the segment level?  Many retailers use some combination of the following three approaches:

  1. Analyze historical promotional response at the segment level to find out which segments show incrementality and profitable results to price promotions. By taking overall promotional analysis and drilling down to the segment level, it is possible to see which segments actually respond to a promotion at all as well as which segment’s purchases are actually incremental, rather than simply replacing existing purchases with a greater discount.  If you do not have control groups in your data, there are two ways to roughly estimate how a specific segment responds to different kinds of promotions. The first one is to examine customers using the promotion versus customers that are making purchases without the promotional redemption. Match them by their prior purchases to make them roughly equal and examine what happened during the promotional purchase and beyond. In addition, you can analyze customers who receive the promotion versus new customers who could not have received the promotion to see what the relative lift would be.
  2. Test and control your way into knowledge. If you want to make sure that your estimate of segment response to promotions is accurate, you can always take a subset of your customer database by segment, break that into a test and control group and run an email marketing campaign. You can even test different levels of promotion versus each other and versus the control group all by building a test matrix, executing the program and reading the results.
  3. Analyze past pricing actions at the segment level.   Identify markets where pricing has changed and evaluate customer purchase patterns before and after the announcement of that change. You will have customers ranging from those that abandon the product category after price increase to those who did not change their behavior at all. By understanding segment response to prior price actions, you can get a sense for what the net impact of a pricing change at this point would be as well.

Professors Farris and Ailawadi lay out a convincing case for retailers to increase prices after such a long period where the retailer (or the manufacturer) have been forced to manage the consequences of a highly price-sensitive market. Now that costs have risen over a number of years, it is appropriate for retailers and manufacturers to cautiously, carefully and deliberately increase prices to help to absorb those margin hits since 2008.

By varying promotions and discounts, the most valuable, most price-sensitive customer segments can be somewhat protected from the impact of these price actions. The strategy of segment–specific marketing programs also benefits the retailer or manufacturer because it communicates to consumers that they are known and cared for by the organization.

Price actions are very scary in today’s environment. Follow the guidelines in the Wall Street Journal article, and analyze and execute your marketing promotion and communication plan at the segment level – and you will be just fine.

In fact, customers in the highly protected segments may appreciate the personalized care and attention to their needs. In this way, counterintuitively, you may be able to grow the business while taking a price action at the same time.

Full disclosure, I studied under Prof. Farris at the Darden School of  Business at the University of Virginia, and we have kept in contact over the years.

Customer Segmentation: Using Behaviors to drive Data-driven Marketing

Posted by: liftpoint | On July 24, 2013

segmentation action vs wordsThe saying “actions speak louder than words” has never been more true than when it comes to segmentation.  In trying to understand customers, both words and actions have a role; when it comes to driving revenue, it is all about the actions.

Are you in danger of confusing attitudes for actions? 

Many companies face this problem when it comes to customer segmentation.  They conduct research on consumer attitudes and beliefs, with the goal of “better understanding their customers.” That attitudinal research is then grouped into segments with commonalities in attitudes about the category or just in general.  Finally, those groups are described as personas, and the company fills the walls with those pictures in an effort to make the company more “customer centric.”

This type of research is valuable when developing mass media creative, identifying opportunities for new products or new markets for the company as a whole. The problem occurs when marketers try to target one specific group for an offer, product or communication.  The problem is:  Who are the consumers who have those attitudes?

The research company may have provided the company with a “typing tool” — a questionnaire that, when answered, can place a consumer into a specific group.  The problem with that approach is that you cannot get enough consumers to take the survey to be able to create a meaningfully sized group.  Net of all, companies find that attitudinal segmentation is rarely actionable for database marketing.

The end result is that the attitudinal segmentation gets relegated to the back shelf and seldom referenced again.  By the way, did I mention that such segmentation projects can range from $250,000 to over $1 million?

What’s the alternative?  To avoid such situations, the best database marketers follow these three specific rules:

  1. Measure what you want to change.  If the goal is to change behaviors (particularly transactional behavior), then group customers with similar behaviors rather than attitudes together. This approach will let you first identify different behavior patterns and then, importantly, identify customers with those patterns — each and every one of them. Then you can apply the rules of database marketing to drive additional retention, cross-sell and referenceability from them.
  2. People are as they act, not as they say.  Research has shown that consumers who claim to have the same beliefs (and even same claimed purchase behavior) frequently differ in their purchases when it is measured.  Behavior is the one factor that cannot be debated — as Joe Friday said in Dragnet, it is all about “just the facts, ma’am, just the facts.”
  3. The more you know, the more you know.  Don’t stop with customer purchase behavior with your company; expand your knowledge base by identifying website behavior, email open/click behavior, store credit card behavior and so on.  The more you know about a customer from their behavior, the most you will be able to complete the picture of that customer.

It does not matter if the customer is Hispanic or White, young or old, rural or urban. Customers who purchase and behave the same are effectively the same from a marketing standpoint.

Behavioral segmentation lays the groundwork for a deeper understanding of your customers: segment migration over time, seasonality, promotional response, product preference, and so on.  Each of those pieces helps you build a fundamental understanding of the levers that can be pushed with your customers to help build stronger,long-lasting, profitable relationships — the end goal of data-driven marketing.

Please check out my post “Customer Segmentation: So Much Analysis, So Few Results” if you want to learn more about making customer segmentation effective.

Customer segmentation is a critical component of marketing credibility.

To view this as a Slide Share presentation click here. 

Customer Segmentation Projects: So much analysis, so few results

Posted by: liftpoint | On July 1, 2013

Customer_Segmentation_3I was speaking with a friend yesterday at breakfast about the customer segmentation initiative in his organization. The analysis was proceeding smoothly.  There was significant momentum for that initiative both in the “rank and file” and in senior management as well.  Everyone was looking for this initiative to be the “secret sauce” to take the company to a new level.

But my friend is likely to fail.

As I thought about customer segmentation projects that we have worked on over the past 10 years, I realized that while the customer segmentation analysis may have been successful in virtually every instance, the actual customer segmentation initiative has succeeded in fewer than 50% of those companies.

You see, with few exceptions, most companies do not want insightful analytics; rather, what they want is to grow their business. The analysis is only as valuable as how it is used to change customer behavior and grow revenue/profit.

So a customer segmentation analysis may be deemed successful if it identifies clear and distinct segments who behave and think differently from each other. However, for the customer segmentation initiative to be successful, that segmentation must be used by the organization with specific actions that will add value – clear, identifiable customer value.

That is where segmentation projects frequently fail. I know another company where over the past four years they have built segmentation analyses more than one time. The work done there defined those segments in a way that was both plausible and actionable. But as of today, I would deem both projects failures. Because no one did anything with the results.

There is a wide range of reasons why organizations are resistant to implementing customer segmentation initiatives. These three reasons represent some of the greatest barriers to driving customer and company value out of customer segmentation:

1. Mismatch between Metrics and Desired Behaviors – A conflict between metrics and segmentation often leads to an early flameout for the initiative. The problem is you see, “people will do first as they are compensated, second as they are measured, and third exactly what they used to.”

If the company metrics are quarterly-based, or even worse, monthly-based, employees will be unlikely to spend time on initiatives that can yield higher returns in the long term but may impact the cost line in the short period.  Since segmentation increases marketing costs (creating different offers, creative, etc. for each segment), and the returns may not come for several months down the road —  you can see the problem.

2. Conflicting Priorities – Conflicting priorities are something that every marketing organization struggles with, and segmentation is often the victim. When a segmentation initiative is introduced it may be the top priority.  However, over time, other projects become more important in the short term.  If you have enough short term projects falling one right after another, segmentation runs the risk of falling off the list entirely.  Since segmentation at the beginning is an unknown, and workloads are already high, the project is often delayed in order to “make the numbers, now.”

3. No repeatable process - most companies can execute any email program on a one-off basis.  After all, they have been trained by management to deliver email programs almost at a moment’s notice, often to a fault.  But the problem starts in the follow-up to the campaign — measurement — and extends to be a lack of process to repeat the same type of program as the normal course of business.

If the marketing department does not have a consistent process for evaluating marketing programs on a regular basis and they do not measure the segmented effort, then the effort as a whole is doomed to fail.  After all, if you do not determine that a program is successful, why should you repeat it.

In addition, segmented marketing programs are much more work the first time around.  The only way they get easier is if you plan for multiple versions and then slot the work into its time slot. (Even then, segmentation is more work than blast programs, but less than if you have no process). When a marketing department is either overwhelmed or just basically reactionary,  any approach but the simplest goes by the wayside in the name of experiency.

And there goes segmentation…down the drain.

Customer segmentation clearly has its challenges, but it also has its own rewards. As the company begins to understand the differing needs of different segments, and tailors products, services and experience to match those needs, customers will respond with increased loyalty, share of wallet and customer value. The reward to the marketing department is not only the sense of a job well done, but the rewards that come with direct, measurable, trackable revenue driven by their initiatives.

So if you are starting a customer segmentation initiative, or find yourself in the midst of one right now, you should take a few minutes and examine these three factors to see if they apply in your organization. More often than not they will.

What will you do about it?

Focus is the key to avoid drowning in Big Data

Posted by: liftpoint | On May 21, 2013

focus cameraLast week, I met with a company that had the most complete data that I have ever seen.  From web traffic to email/direct mail to transactions and so on, this company had invested in their data infrastructure as a key asset.  But when I met with them, they were still dissatisfied…

They weren’t making money from the data.

The complaints that the digital marketing team had were the same ones that similar teams face in companies with less complete data:  “Management keeps asking questions that are interesting but not actionable.” “We haven’t had time to create the measurement approach,” “Control groups are not part of the company culture here,” and so on.  Despite having Big Data in better shape than most companies, they were still bogged down in the same issues as their less successful brethren.

The main complaint boiled down to one thing: “We don’t know the right questions to ask.

This issue is not uncommon.  And very frustrating.  And as Yogi Berra once said, “If you don’t know where you are going, you will wind up somewhere else.”

When you look at companies who are the most successful at leveraging Big Data, they have the following characteristics:

1. Fewer questions to start, rather than more.  Rather than investigate everything under the sun, these companies have taken a “step-by-step” approach focusing on identifying opportunities for quick wins from relatively quick analysis.

  • The goal of this effort is to increase company confidence in the accuracy of the data and analysis as well as to demonstrate how many profitable actions can be identified and taken based on data-driven insights.
  • One example of this type of analysis include product gap analysis, to identify customers who have purchased Product A without the accompanying Product B.  Cross-sell promotions of this type are usually immediately profitable.

2. Begin with the end in mind.  Rather than research “why people buy more red t-shirts in Tulsa on Tuesday,” each analysis should begin with a hypothesis of what the answer might be, and an explanation of the action that could be taken based on that finding.  That way, the ROI of data-mining Big Data can be clear and meaningful.

  • This approach does not mean that the analytics team should be the arbiter of whether or not a proposed action is valid; rather, the analytics team should work WITH the marketing teams to help make sure that the proposed action is as valuable as possible.

3. Measure and publicize the results.  Analytics does not exist in a vacuum; the value of Big Data is based strictly on the actions that can be taken by analyzing it.  For that value to be recognized, the actions must be driven cleanly by the insights (so others can see how the analysis led to conclusions which lead to marketing programs) and must result in a change in customer behavior that drives incremental revenue and profit.

  • To ensure that the results are valid, they must be measured.  Whether pre/post, vs. prior year or vs. control group depends on the approaches in your company (BTW — control group is the most valid, but many of the best companies use all three).

Big Data presents us all with challenges.  Organizing the data to be able to answer questions is one of the challenges.  But the greater challenge is one of focus.  We will, more than ever, be able to ask almost any question about our businesses and the data will be there to answer.

What are we going to ask and what are we going to do with the answers?

The answer to that question may be the most important one of all.

Customer Data Analysis: Overcoming the Imperfect Data Trap

Posted by: liftpoint | On May 15, 2013

Have you ever felt like you had all the facts to make a decision?  If you are like me — never.  Whether for a new smartphone, a new car, someone to prune your trees — there is always more information available on the web, another friend to speak with, another competitor to consider.

Eventually, you just have to “suck it up” and make the best decision you can make with the information you have.  You will NEVER have all the information — you just have to feel like you have enough to make an informed judgment.

We all make those sort of decisions in our lives outside of work – that is why I find it so surprising that marketers frequently get “stuck” and cannot make a decision without every possible piece of information.

That’s why I recommend applying the “good enough” strategy to data collection and customer data analysis.

In the early stages of customer data analysis, the data will not be very “clean” and the tools you will use will be very basic (think Excel). On the data side, often there are duplicate customers, bad or NULL data in certain fields and incomplete physical and email addresses, among other issues.  On the technology and skilll side, your team may only have basic Excel skills so far (note the “so far”).

Yet Marketers can’t wait for perfect data.  The pressure is on, NOW, and we need to find insights and take actions on them in the short-term for there to be a long-term at all!

I’m assuming you already have a marketing strategy with campaigns in place. You need to know which campaigns are the most effective to which customers and why.  In order to get answers to those questions, you have to find a way to “duct tape” together a way to get that data.

What are the challenges and how do you overcome them?  The most common data-usage barriers we see with our clients fall into three general categories: skill, technology and data quality.

Let’s look at each and explore some options for overcoming those obstacles.

1. Lack of Skill Set. If you don’t have the right person on your team to answer your questions, make your questions simpler. Try to stay more macro. Ask questions that can be answered with one analysis at a time.  For example, ask “what are total sales from customers redeeming this coupon,” first.  Then take the data on just those customers and look at average order value.  Then look at how many dollars were spent on the promoted items, and so on.  By setting up your questions as a series of basic analyses, you can make them easy to do for a team member with basic Excel skills.

2. Lack of Technology. Everyone has Excel, and most marketers can do the basic analyses and a couple of the “If-Then” and Lookup sorts of activities.  The key is to make sure that you structure your IT request to ensure that the data that comes can fit into Excel (less than 1mm rows and not too many columns).  If necessary, ask for a random 10% sample of customer data to make sure it fits in Excel.

3. Lack of Data Quality. By doing some basic Excel work, you can clean up your customer data to permit the basic analyses we discussed above.  The approach is not to solve “the problems of the universe,” but to simply identify and exclude customers whose data is obviously incorrect. Eliminate addresses without common extensions: org., com., net. or edu. Eliminate customers with spending less than $1 or more than $10,000, customers with negative transaction counts or 100s of transactions.  Doon’t try to fix them; just exclude them for now and worry about them later.  As you do more and more of this sort of analysis, you will build a list of exclusions that you can repeat time and again, faster than before.

Just get rid of as much bad data as you can, and move on.

A 70% correct solution is 100% better than what you had before. Just about every marketing publication and blog talks about the critical importance of data to enhance effectiveness of marketing campaigns.

You know you need data. Get it any way you can.