By now, everyone knows that cross-selling (including upselling/cross-selling a new product) is an unbelievable source of profitable revenue growth. Yet, there is a challenge. To be successful, a cross-sell sales play or marketing campaign must provide highly targeted, very tailored offers to each prospect. “Carpet bombing” (oversaturating) existing account buyers and prospects with company-specific messaging won’t work; you’ve got to hit the right buyers at the right time with the right offer and right content. Simply put: your buyers know you, and they expect you to know them.
In Episode 6 of our Killer Slide Series short video, we walked through the below customer intelligence platform we help companies develop to scale cross-selling efforts. In short form, marketers and sales reps need to answer four questions, and without automation, that requires a rep or marketer to spend at least 20-30 minutes per prospect researching as many as 10-15 different data sources for answers. That’s not scalable…period.
The methodology we use (we call it PlayCaller), answers a basic set of questions needed to drive cross-sell:
- Who should they target this week?
- What product(s) should each prospect be offered? What is the best content needed to advance and close a deal?
- How and how often should they be contacted (i.e. channels)?
- Where are we achieving measurable success?
The general approach to answering these questions is outlined in the above graphic, but it’s how this gets scaled and acted on that’s critical. In this blog, we focus on three key components to the cross-sell intelligence platform: 1) Data Sources, 2) Decision Analytics, and 3) Activating in Workflow.
1) Data Sources:
Most companies likely already have the critical data sources they need to double or triple their cross-sell effectiveness. The challenge is integrating these data together in either a structured data warehouse or a semi-structured data lake. Four of the most powerful sources of cross-sell data (or “signals” as we call them) include:
- Purchase History: Purchase history includes timing and cadence, what was specifically bought, and prices paid. These are obviously valuable data, but sometimes the analytics are surprising. Did you know that oftentimes increased purchase frequency is a strong predictor of customer flight to a competitor?
- IoT/Product Usage History: As IoT and other usage tracking technologies explode, these big data have already become a powerful signal identifying which customers are ready to be sold to.
- Customer Service Inquiries: By tracking both service needs and complaints, product upgrades and cross-sell can turn around client dissatisfaction.
- Web Visits and Downloads: Your existing clients visit your website more than anyone. Tracking who is reading what—and following up with highly relevant “add-on content”—can turn content consumption signals into new deals.
2) Decision Analytics:
Too often, companies deploy basic machine linear models, such as optimizing which customers should be mailed to in order to maximize ROI, and stop there. But when it comes to data science, simple linear thinking just doesn’t cut it today. Marketers and sales channels need to know exactly which customers in their territory are ready to buy what products, and through which channels they want to be sold to. MarketBridge typically starts with four models when building cross-sell platforms for our clients:
- Buyer Segmentation: There are usually 5-7 buyer types present in a given industry. These clusters are typically driven by functional need (what), psychographics and buying style (who), and channel usage (where). Understanding and scoring these segments helps attain better cross-sell performance by getting the right content and offers to the right buyers. For more detail, check out this blog on content segmentation and this approach to buyer segmentation.
- Buying Proclivity: The proclivity to buy helps determine when to send a specific “action” outreach vs. a nurturing, “teaching”-type outreach. The basic elements of RFM (Recency, frequency, monetary value) analysis are typically critical inputs into this score.
- Next Logical Products: Understanding the specific SKUs and bundles of SKUs that a buyer is likely to need next helps drive relevancy.
- Customer Attrition Risk: The risk that a contact, buying center, or entire account is about to stop buying is critical. When a “warning level” is reached, be prepared to spend more on high-value touches to intervene. The best attrition-avoiders validate, research, and respond to attrition warnings, in-person if necessary.
3) Sales & Marketing Activation:
The best data and machine learning are useless unless they can be easily understood and acted upon by your sales & marketing channels. Delivering predictive analytics into your existing customer contact workflow platforms is critical. We have found that loosely coupled architectures are dramatically better over the long run than tight integrations with SaaS martech platforms. Think about it—if you’re using the exact same sauce as all of your competitors, you won’t be able to maintain your advantage. This is what the SaaS vendors want, of course—for everyone to be on their system, using identical proprietary algorithms.
Additionally, most companies are using multiple platforms to manage the end-to-end customer journey. The need to insert insight into each one of the platforms—and to gather feedback from the customer interaction directed by those platforms—is critical to managing the process holistically, and understanding where the customer is in the process at any given point in time.
The better alternative is to deliver analytics via a set of standardized endpoints that any CRM or martech platform can use, and then writing quick integration layers for specific systems; this is the MarketBridge approach. This way, when that next great piece of technology is rolled out—or when Salesforce raises its prices by 20%— it’s no problem. We just need to spend a couple of days writing an adaptor layer, vs. tearing out a bunch of proprietary APEX code from Salesforce and trying to remember what the developer was thinking.
Fortunately, all CRM and marketing automation systems—including Salesforce—share the same basic architecture. The objects Account, Contact, Lead, Opportunity, Product, etc. don’t really vary, and haven’t for 25 years.
Cross-sell recommendations also share the same DNA. Typically, the API calls for cross-sell include several microservices that, when taken together, form the basis of the contact / content strategy woven into the CRM / Martech platform. We’ve listed a few of the most common and useful here:
|Account||Top cross-sell contacts?||Ranked list of contacts|
|Contact||Segment?||Segment for this contact|
|Contact||Top products?||Ranked list of products|
|Contact||Likelihood to attrit?||Probability|
|Product-Segment||Right offer?||Offer structure|
|Product-Segment||Right content?||List of top content pieces|
In summary, this framework should help sales and marketing executives think about cross-sell in a systematic way, and set about driving continuous improvements across all three.
- What data are needed to derive the signals that make good cross-sell decisions, and how do we go about getting them all in one place, relatively cleanly, and keyed to the level where data scientists can use them for modeling?
- What questions should executives be answering with analytics and data science, beyond the basic “scored list” approaches that dominate the space today?
- How can I get data out of and into my operational systems, without tightly coupling?
- How should I measure the end-to-end process across platforms and use that information to optimize my performance?