Converted

1. Marketing must shift from single transactions to ongoing conversations.

Interactions between the best companies and their customers are changing from quick messages demanding an immediate response to deeper, more lasting conversations.

Beyond the single click. Traditional digital marketing often focuses on optimizing individual moments, like getting a click or an immediate sale, treating every interaction and customer the same. This short-term view, while measurable, misses the larger picture of building enduring customer relationships. Companies act like someone proposing marriage on a first meeting, hoping sheer volume yields results.

Relationships matter. Just as in real life, meaningful connections with customers build trust and loyalty over time. Customers think of brands and websites almost as people, forming opinions and attachments. Businesses that prioritize conversation stand out from competitors stuck in the "buy now" mentality.

A necessary shift. The market is evolving, and customers are becoming apathetic to relentless, generic tracking and messaging. Marketing leaders must learn to recognize and respond to the signals customers send, moving beyond simple statements to engage in dialogue that fosters deeper, more lasting relationships.

2. Start simple with data, focusing on people and identification.

Customer conversations are not about capturing every single interaction.

Avoid complexity paralysis. Many companies get bogged down in trying to collect and perfect every imaginable piece of customer data before taking action, leading to expensive, multi-year projects that yield no immediate results. This "digital transformation" often fails because it prioritizes data perfection over progress. The goal isn't to capture everything, but to focus on what truly matters.

Prioritize people and money. The most valuable data is straightforward: who spent money, how much, and when. This transaction data, tied to a unique customer ID, is the foundation. Start with simple tools like spreadsheets and add data with purpose, focusing on identifiable customers rather than channels or campaigns.

Identify your customers. Knowing customers by name (or a unique ID like email or loyalty number) is crucial for connecting interactions across systems. Offer incentives for registration and use single sign-on to ease the process. Focus on identifying as many people as possible, finding the right balance between data collection and customer friction.

3. Ask questions to deepen understanding, going beyond surveys.

Some of the most successful marketing practices share this grounding in curiosity and inquisitive conversation.

Data is a window. While data provides insights, it doesn't tell the whole story. Successful marketers actively participate in the conversation by asking questions to understand customer goals and advance the dialogue. This knowledge provides a significant advantage over competitors who only interpret existing data.

Beyond the annual survey. Don't limit questioning to infrequent, generic surveys with low response rates. Integrate questions into website interactions (e.g., checkout pages), rotate questions frequently, and use modern tools for fast, targeted surveys of specific audiences.

Strategic questioning. Ask questions that reveal customer intent, potential value, and preferences.

• "Are you buying this as a gift?" (Gift buyers spend more later)
• "How much do you spend on [category]?" (Reveals share of wallet opportunity)
• "Why do you keep coming back?" (Identifies journey stage)
• "What do you like most about us?" (Positive questions increase sales)
  Phrase questions carefully and avoid asking too many, focusing only on those whose answers will inform your next action.

4. Embrace irrational human nature in marketing strategies.

But we’ve actually found a lot of opportunity in looking beyond the data to recognize the reality of human behavior: it’s often irrational.

Marketing isn't purely logical. While data might suggest customers are perfectly rational beings driven solely by price, speed, and features, human behavior is often nuanced and irrational. Recognizing this provides opportunities to connect with customers on a deeper level. A faster website isn't always perceived as better if the effort involved isn't visible.

Behavioral science techniques: Incorporate insights from behavioral science to influence customer perception and action.

• Tease the finish line: Frame processes to suggest progress has already been made (e.g., "Your account is 90% set up!").
• Stress scarcity/loss aversion: Highlight limited availability or potential loss to drive urgency (e.g., "Only one room left," "Don't miss out").
• Gather the crowd: Use social proof like testimonials, reviews, or showing how many others are interested (e.g., "15 people looking right now!").
• Plant a seed (Priming): Expose customers to stimuli that influence future responses (e.g., asking "Are you a gamer?" before showing a gaming ad).

Guide, don't manipulate. Use these techniques ethically to guide customers, not trick them. Understanding that customers are emotional beings, not just data points, allows for more effective and human-centered marketing.

5. Learn to read subtle customer hints beyond explicit actions.

But you can still look at combinations of other hints—where small things come together in surprising ways—and carry the conversation forward.

Beyond the obvious. Customers don't always explicitly state their intentions. Relying only on direct questions or obvious actions (like adding to cart) can be misleading. The automaker who focused on the car customizer tool missed the real signal: looking for financing information.

Complex signals matter. Subtle behaviors, when combined, can reveal true intent.

• Product returns: Habitual returners, ordering multiple sizes, or lack of engagement with product details (like zooming in) can hint at future returns.
• B2B intent: Visiting specific pages (like pricing or features), time of visit, or adding team members to a trial can signal purchase likelihood.
• Gift buying: Apologetic messages can indicate price insensitivity.
• Credit cards: Waiting a few days after an offer might signal higher long-term value than immediate sign-ups.
• Cart curation: Modifying items in the cart can be a stronger purchase signal than just adding items.

Machine learning helps. Analyzing thousands of signals manually is impossible. Machine learning can identify combinations of behaviors that predict outcomes, helping marketers understand which hints matter and which are just noise. Start by defining the problem (e.g., predicting purchase likelihood) and use data you trust.

Insights aren't eternal. Customer behavior and market conditions change. Continuously look for new signals, rerun models, and challenge assumptions. Don't just eavesdrop; use hints to anticipate needs and guide the conversation effectively.

6. Not all customers are equal; measure their long-term value (CLV).

The metric we use to understand customer relationships is known as customer lifetime value, or CLV.

Beyond the transaction. Just like in personal life, not everyone you meet will be equally valuable. In business, a small percentage of customers often drive the majority of value (the Pareto principle). Treating all customers the same is inefficient and prevents you from nurturing your most profitable relationships.

CLV quantifies relationships. Customer Lifetime Value predicts the total worth of a customer relationship over its duration. It's a crucial metric for understanding if marketing efforts create sustainable value or just facilitate one-off transactions. It shifts focus from immediate sales to the long-term health of the customer base.

Calculating CLV simply. You don't need perfect data or complex systems to start. A basic CLV model requires just three data points: customer ID, transaction date, and transaction value. Use historical data to predict future behavior and validate the model's accuracy. This process allows you to segment customers by predicted value, revealing the distribution of worth within your customer base.

7. Focus acquisition efforts on finding more high-value customers.

Finding great relationships—like the customers we’ve learned to identify—is much easier than trying to change someone into a better person.

Acquisition is key. While customer development and retention are important, acquiring the right customers from the start is often the most effective strategy. It's significantly easier to find people who are likely to become high-value customers than to transform low-value ones into top spenders.

Use first-party data. Your own customer data is your most valuable asset for identifying characteristics of high-value customers. Analyze CLV alongside acquisition channels, initial purchase categories, use of coupons, or engagement with specific features.

• Customers acquired through paid search might have lower initial spend but higher CLV than those from affiliate marketing.
• Customers who engage with certain product categories first might be more valuable long-term.

Target strategically. Use insights from your CLV analysis to refine acquisition campaigns.

• Target audiences similar to your high-CLV customers using ad network tools.
• Weight campaigns to spend more aggressively on signals associated with higher CLV.
• Update conversion tracking to report CLV instead of just initial transaction value.
• Identify and potentially exclude characteristics associated with low-value customers.

Trust first impressions. Initial behaviors (first purchase, acquisition channel, use of promotions) are often strong predictors of future value. Prioritize behavioral attributes over demographics when building customer profiles.

8. Be realistic about changing customer behavior; prioritize development wisely.

Changing a customer’s behavior is almost the same thing as changing a person—trying to change them into your soul mate through effort and determination when it just isn’t meant to be.

Avoid the unicorn trap. Companies often make the mistake of acquiring large numbers of low-value customers, believing they can later convert them into high-value ones. This is incredibly difficult and costly, as seen in the cautionary tale of the unicorn company that failed trying to make low-value customers profitable.

Focus on realistic growth. While transforming a terrible customer into a top-tier one is unlikely, you can nudge existing relationships to become slightly more valuable.

• Give best advice: Use recommendation engines to increase transaction value or suggest complementary products. Amazon and Netflix excel at this.
• Find more to offer: Cross-sell complementary products or services. Allstate found cross-selling existing customers was more effective than acquiring new ones.

Target development efforts. Don't try to develop every customer relationship. Identify which customers have the potential for growth and focus efforts there.

• Use signals like time between purchases, return rates, or initial purchase category to target cross-selling.
• Ask about "share of wallet" to understand potential for increased spending in a category.
• Avoid investing heavily in customers who cost more to service than they return.

Optimize, don't transform. Be realistic about how much customers will change. Focus on making great customers slightly better and average customers a little more valuable. Don't let optimism override good judgment; invest where the potential return is highest.

9. Strategically retain valuable customers, knowing when to intervene or let go.

When it comes to forging lifelong relationships, customer retention is the glue.

Retention's importance. Retaining existing customers is significantly cheaper than acquiring new ones and is crucial for long-term profitability. Companies that master retention build loyal advocates who not only spend more but also promote the brand to others.

Identify risk signals. Don't wait until a customer explicitly cancels. Look for subtle signs that a relationship is in trouble:

• Specific website actions (turning off auto-renew, visiting cancellation pages).
• Decreased engagement (lower service usage, fewer visits, lower email open rates).
• Increasing time between purchases.
• CLV models can predict the probability of future transactions, highlighting customers at risk.

Intervene wisely. Not every at-risk customer is worth saving. Prioritize intervention based on the customer's predicted lifetime value.

• Focus efforts on high-value customers who show signs of leaving.
• Test different interventions (small gestures often work better than large discounts).
• Don't overspend to save low-value relationships; the cost may outweigh the potential return.

Know when to say goodbye. Relationships have end points. When a customer's predicted value is nearly realized, or the cost to retain them exceeds their potential future value, be comfortable letting go. Don't chase the last few dollars from a relationship that has run its course.

10. Listen to the voices of your most valuable customers, not just the average.

It’s not calculating lifetime value or targeting more customers with ads that elevates you into a customer-centric business.

Beyond the average. Optimizing marketing based on average customer behavior can be misleading. High-value customers often have different preferences, motivations, and responses than the average customer. What resonates with the majority might alienate your most profitable segment.

Segmented listening. Use CLV segments to understand what matters to different customer groups.

• High-value customers might prioritize quality, service, or brand legacy.
• Low-value customers might be driven primarily by price or promotions.
• Test messaging and creative specifically tailored to high-value segments. An ad campaign with a lower average open rate might be more successful if it engages a higher percentage of your top-tier customers.

Build new best practices. Your company's marketing rulebook should be based on insights from your most valuable customers. Analyze past campaigns and experiments through the lens of CLV impact, not just immediate conversion rates.

Share insights widely. Arm product development, sales, and service teams with knowledge about high-value customers. Zappos learned their most valuable customers had high return rates and adapted their policy to cater to them, recognizing the long-term value outweighed the short-term cost.

11. Foster a culture of rapid, iterative testing and learning.

Seek progress, not perfection.

Testing drives learning. To understand customers and adapt strategies, companies must test new ideas. However, organizations often face bottlenecks: fear of risk, bureaucracy, and a desire for perfect data before acting. This inertia prevents learning and allows more agile competitors to gain an advantage.

Think small, move fast. Don't wait for the perfect plan or perfect data. Embrace small, iterative changes and tests. The opportunity cost of inaction (missing potential sales) often outweighs the risk of a small test. An imperfect step forward is better than standing still.

Unleash ideas. Create a process to capture and prioritize test ideas from everyone in the organization, not just leadership. A simple form asking for hypothesis, supporting data, test plan, and action based on results can surface dozens of opportunities.

Incentivize ideas, not just results. Reward the best hypotheses, not just successful tests. This encourages bold thinking and surfaces potential breakthroughs. Separate experimentation budgets from marketing to remove pressure from quarterly targets.

Remove bottlenecks. Identify what slows down testing (approvals, budget allocation, technical implementation) and address it. Empower teams to run Type 2 decisions (reversible) quickly without excessive oversight, as Amazon does.

12. Understand the human and organizational factors influencing data-driven decisions.

Decisions are not made based on data alone, ever, and you’re not going to change the way people make decisions.

Data isn't enough. Even compelling data won't automatically lead to change. Decisions are influenced by egos, team incentives, fear of risk, and personal experience. A sales director's engagement metric failed because it didn't account for how salespeople would interpret and manipulate it based on their bonus structure.

Metrics can be manipulated. Any metric can be gamed, intentionally or unintentionally. Companies often focus scrutiny only on underperforming metrics, allowing inflated or misleading positive numbers to go unchallenged. Understanding the levers behind a metric and their consequences is crucial.

Get inside the numbers. Don't blindly trust dashboards. Use techniques like "red teaming" (borrowed from the CIA) where a separate group challenges the analysis and findings of substantive projects. This process reveals weaknesses, highlights subjective analytical choices, and encourages transparency from project owners.

Navigate organizational reality. Be aware of the political landscape. Understand stakeholders' incentives and perspectives. Don't make promises you can't keep, and agree on how test results will be acted upon before running the test to avoid paralysis and arguments later. You can't force people to be data-driven, but you can understand the rules of the game and navigate them effectively.

Last updated: May 1, 2025