Posts

Avoid Segmentation Missteps to Boost List ROI

List segmentation is key in targeted direct marketing, which is why the AccuList team offers clients help in defining best-performing customer segments via predictive analytics services and data management services. Over the years, we’ve learned that the secret to success is as much a matter of strategic mindset as technical expertise. A recent MarketingProfs article by Mitch Markel, a partner in Benenson Strategy Group, makes that point by identifying some of the common strategic errors that can trip up a segmentation effort.

Obvious Parameters and Old Strategies Dig a Rut

Marketers need to be aware that segmentation models can slip into an ROI rut. Use of obvious profiling parameters and assumptions is one reason. Certainly, demographics (or firmographics), stated needs, and past purchase behavior are essential in grouping for likely response and lifetime value, but people don’t make decisions solely based on these factors. Markel urges research that also looks at fears, values, motivations and other psychographics in order to segment customers or prospects not just as lookalikes but also as “thinkalikes,” which can be especially helpful in crafting personalized content and messaging. Markel cites the examples of car buyers grouped by whether they value safety over performance, and food purchasers sorted for whether they stress healthy lifestyle or convenience. Past success is another reason segmentation can get stuck in a rut. Because segmentation requires an upfront investment, marketers tend to want to stick with proven targeting once the segmentation study is completed. But today’s hyper-personalized, digital environment has accelerated the pace of change in markets, perhaps shifting customer expectations and preferences away from an existing segmentation model. Markel advises an annual “look under the hood” of the segmentation engine to see if segments are still valid or need appending/updating. An annual audit can avoid the expense of a broader overhaul down the road.

Big Data Blindness Ignores Potential Audiences

One outcome of segmentation based on existing customers is blindness to potential audiences. Segmentation research often uses the existing customer base and surveys of people that marketers assume should be targeted. This can create marketing campaigns that miss groups that Markel calls “ghost segments,” people who could be among a brand’s best prospective customers. Markel suggests a periodic look at non-customers for conversion potential as one way to capture these “ghosts.” And, of course, if a new product or service is in the works, research should ask whether it will attract new groups differing from the existing customer profile. Another reason ghost segments are common is that marketers, overwhelmed by the task of sifting “big data,” fall back on whatever data sets are handy. Markel suggests that it would be better to bring in big data at the tail end of segmentation. He advises analysts to start by creating segments using primary research, add existing customer “big data” to target those segments more efficiently, and then plug segments into a data management platform for insights on other products, services, interests, and media that may correlate.

Analytics Miss Without a Companywide Strategy

Finally, Markel stresses that a segmentation study that ends up residing only with a few marketing decision-makers will fail to live up to its ROI potential. Customer and prospect insights have relevance for multiple departments and teams, from sales to customer service to finance. In order to deliver a seamless, personalized customer experience, Markel suggests creating 360-degree customer personas and promoting them throughout the organization. Management can start with workshops to educate employees on the use and importance of those personas both for their departments and the organization, and then can schedule check-ins to show team members the resulting benefits of segmentation and targeting implementation. If segments are made relatable, it will ensure they are used and embraced across the organization.

Predictive Analytics Can Harness Data for Marketing ROI

Beyond list brokerage, AccuList can support direct marketing clients with “predictive analytics,” meaning scientific analysis that leverages customer and donor data to predict future prospect and customer actions. It will scientifically “cherry-pick” names from overwhelming “big data” lists and other files. For example, AccuList’s experienced statisticians build customized Good Customer Match Models and Mail Match Models to optimize direct mail results for prospect lists, as well as one-on-one models for list owners to help acquire more new customers or donors. Plus, predictive models aid other marketing goals, such as retention, relationship management, reactivation, cross-sell, upsell and content marketing. Below are some key ways predictive analytics will harness data for better marketing ROI.

More Swift, Efficient and Effective Lead Scoring

Lead scoring is too often a sales and marketing collaboration, in which salespeople provide marketers with their criteria for a “good” lead and marketers score incoming responses, either automatically or manually, for contact or further nurturing. Predictive analytics will remove anecdotal/gut evaluation in favor of more accurate scoring based on data such as demographics/firmographics, actual behavior and sales value. It also speeds the scoring process, especially when combined with automation, so that “hot” leads get more immediate contact. And it allows for segmentation of scored leads so that they can be put on custom nurturing tracks more likely to promote conversion and sales.

Better List Segmentation for Prospecting, Retention and Messaging

With predictive analytics, list records can be segmented to achieve multiple goals. The most likely to respond can be prioritized in a direct mail campaign to increase cost-efficiency. Even more helpful for campaign ROI, predictive analytics can look at the lifetime value of current customers or donors and develop prospect matching so mailings capture higher-value new customers. Predictive analytics also can tailor content marketing and creative by analyzing which messages and images resonate with which customer segments, identified by demographics and behavior, in order to send the right creative to the right audience. Finally, analytics can develop house file segmentation for retention and reduced churn, looking at lapsed customers or donors to identify the data profiles, timing inflection points and warning signs that trigger outreach and nurturing campaigns.

Optimizing for Channel and Product/Services Offer

Data analysis and modeling can also be used to improve future marketing ROI in terms of channel preferences and even product/services development. By studying customer or donor response and behavior after acquisition, analytics can identify the most appropriate promotion and response channels, communication types, and preferred contact timing by target audience. Plus, a customer model can match demographics, psychographics and behavior with product and offer choices to tailor prospecting, as well as upsell or cross-sell opportunities, to boost future results.

Committing to a Good, Clean Customer Database

Reliable predictions require a database of clean, updated existing customer or donor records, with enough necessary demographics/firmographcs and transactional behavior for modeling. So, to prevent garbage-in-garbage-out results, AccuList also supports clients with list hygiene and management, including hygiene matching for DO NOT MAIL, NCOA and more, data appending of variables from outside lists, merge-purge eliminating duplicates and faulty records, response tracking with match-back, and more advanced list screening options.

Why Participate in Modeled Cooperative Databases?

Today’s modeled cooperative databases offer big advantages for B2C and B2B direct marketers, which is why AccuList now represents 18 private modeled cooperative databases that clients can use to optimize direct mail results. These databases include millions of merged, deduped, and “modeled and scored” hotline names from thousands of commercial and nonprofit participants.  At no charge, each can match the client’s database, model client postal addresses, and deliver optimized “look-alike” names.  The database will prioritize those modeled names by decile or quintile to help clients further identify targets most likely to respond to an offer or fundraising appeal.

Fear of Sharing Misses Optimizing Opportunities

Marketers sometimes hesitate to participate because of unfounded fears of sharing exclusive/unique customers, catalog buyers, subscribers or donors with membership-based database participants. Note that these databases generally match a marketer’s names against the cooperative database files and share transactional data. If there are matches, only transactional information is added to the cooperative database records; and if there are no matches, the unique names are not added to the pool.  Why do cooperative databases opt to incorporate only multi-occurring or duplicate records? Because that is data that tends to be far more predictive, with proven response. Plus, the reality is that very few names are unique to a firm, publication or fundraiser. About 80% to 90% of consumer prospects are multi-buyers and so are in the database already, and 90% of nonprofit donors give to two or more organizations and so also are already included in cooperative data. On the other hand, by participating to access a huge pool of names rich with demographic and transactional information, marketers can tap many more optimized prospects, improve list segmentation and testing, bump up response and conversion, hone creative and offer targeting, and increase mailing efficiency.

Modeled Data Offers Cost-Effective Prospect and House Mailing

Acquisition campaigns clearly can benefit from netting look-alike prospects from the large cooperative database pool, a real boon for regional or niche mailers who struggle to find acquisition volume. The large universe also allows for more segmentation to target not only higher response groups but more valuable response segments. In the case of nonprofits, that could be high-dollar donors, for example. Profiling and modeling can create better results from house names, too. Instead of mailing the whole house file, current customers, subscribers or donors can be flagged for likelihood of response and upsell, for channel and messaging preference, for risk of lapse/attrition, and more. Plus, modeled databases offer cost efficiency via an attractive list CPM; recent, clean, deduped records that lower mailing costs; and optimization selects (or deselects) that also boost mailing efficiency and ROI. Check out these arguments for nonprofit participation in modeled cooperative databases, as well as these useful best-practices tips for commercial mailers from Chief Marketer and Target Marketing magazine posts.

Choosing One (or More) Modeled Cooperative Databases

As an industry-recognized list brokerage, AccuList now represents a long list of private modeled cooperative databases, some specializing in B2C, some in B2B, and many offering modeled names for both B2B and B2C campaigns. In addition, as a value-added option, some modeled cooperative databases feature omnichannel targeting services that allow matching of optimized direct mail names with digital media, including Facebook. We can help you choose the right solution to fit your marketing goals with the following leading cooperative databases:

  • Abacus Alliance
  • Alliant
  • American List Exchange (ALEXA)
  • Apogee
  • Dataline
  • DonorBase® (Founding Member)
  • Enertex
  • I-Behavior
  • MeritBase B2B Cooperative Database
  • OmniChannelBASE®
  • PATH2RESPONSE
  • Pinnacle Business Buyer Database
  • Pinnacle Prospect Plus
  • Prefer Network
  • Prospector Consumer Fundraising Database
  • Target Analytics
  • TRG Arts
  • Wiland