Facebook Group: 8065544817 FeedBurner: tagthink/WeWn TAGtv: http://tagtvonline.com Linked In Group: 41590 Twitter: TAGthink
Wednesday, May 23, 2012

Focused conversations within the Georgia technology community.

Text Size
Send
Print
Written by VayuMedia
Relational_Segmentation_V2_0709

Created 23/09/11
Author Name viewpoint
Author Company viewpoint
Body of Topic
viewpoint White Paper
logo How Relational Segmentation Techniques Help Achieve Higher Sales at Lower Cost
logo

Sophisticated relational segmentation techniques balance the principle of statistics with realities of today’s marketing budgets, and can predict the likely success of B2B marketing programs.

This informative white paper can save significant dollars, time and resources by providing you valuable information that helps you:

  • Avoid the missed potential that traditional database clean-ups miss.
  • Understand why a clean list doesn’t equate to a high-performing list.
  • Identify your most valuable segments, and apply that knowledge predictably to generate higher return.
  • Achieve a higher number of more profitable sales in a timely manner—at a lower cost.
logo
How Relational Segmentation Techniques Help Achieve Higher Sales at Lower Cost
logo TREE ANALYSIS

A tree analysis can be used to identify firms likely to be members of a layer, or class, as in pinpointing the highest response segment defined by a specific revenue range and SIC code.

Tree programs are valid for small samples, easy to execute, and easy to understand.

Trees do not support analysis or interrelationships of multiple variables at one time.

For comparison purposes, reference Regression Analysis (page 3) and Cube Analysis (page 4).

Achieve higher sales at lower cost through relational segmentation

Are you wasting dollars, resources and time trying to clean up outdated, in-house databases to generate profitable leads at a reasonable cost? Often, these “black holes” have become final resting places for hundreds, and even thousands, of records characterized by sparse, inaccurate information.

Is there an efficient, cost-effective way to run clean-up and marketing initiatives that will generate higher returns? There is. Sophisticated technology from PointClear can help you wisely test your customer and prospect databases and establish predictors of success—before you deploy broad-based clean-up and lead-generation initiatives.

The PointClear Relational Segmentation (PCRS) approach provides companies with the market intelligence they need to fully fund and roll out programs targeted to high-return segments. This model has increased individual campaign results by up to 50 percent and simultaneously decreased costs by as much as 35 percent.

Traditional clean-ups miss potential

The mandate to clean up and use in-house databases is grounded in a rock-solid objective: Leverage existing customer and prospect data to drive cross-selling, up-selling, and new sales. Easier said than done, however, as companies often struggle with many disparate customer, prospect, and partner databases.

These databases are often characterized by outdated, inaccurate, or sparse information about customers in four critical areas:

  • Current pains and visions for addressing them
  • Current technology environment
  • Correct decision-making team and buying process
  • Plans for short or mid-term purchases

Still, your gut tells you there is opportunity hidden in these databases. The question is, “How can we best clean them and mine them for value?” Conventional wisdom calls for running a Phase I database clean-up initiative followed by a Phase II lead-generation program, setting in motion a one-two punch with high expectations. Yes, data is a little better, and some leads will shake out. A few deals may even close.

logo

But, all too often, return falls far short of potential. Campaigns that should do well don’t, and time, resources, and dollars are wasted. Why? Traditional clean-up programs only focus on replacing dirty or absent data with fresh, correct data. They do not add segmentation or prioritization value needed to predict success. As a result, marketing campaigns can only target all cleaned names, because best names have not been identified.

Head-to-head comparison

ABC Software’s three in-house databases help contrast traditional database clean-up/marketing with PointClear’s approach. Table 1 describes a license customer database, a maintenance customer database, and a prospect list purchased from a technology list vendor.

Table 1

ABC Software Databases
List code List descriptio
C1

Customer list 1

  • Software licenses only
C2

Customer list 2

  • Maintenance contracts and licenses
P1

Prospect list 1

  • Purchased list for telemarketing
  • Example: Computer Intelligence
  • Names match targeted vertical, revenue, and geography criteria

A traditional database marketing program targets all cleaned names in the database, when only a fraction of the names warrants investment.

ABC Software knows there are many opportunities for new sales, up-sells, crosssells, and point sales in its databases. It has the following objectives for Phase I database clean-up:

  • Call and update “firmographic data” for companies in the databases
  • Verify decision-makers and contact information

… along with these objectives for Phase II database marketing:

logo REGRESSION ANALYSIS

Regression analysis looks for relationships among variables, weights them, and then scores each company considering the impact one variable has on response rates compared to weights of other variables.

Its higher costs are usually justified only for larger sample sizes that can range to hundreds of thousands and millions of names. While it helps identify characteristics that define cubes, it does not identify values of characteristics (i.e. exact revenue range) that work best.

For comparison purposes, reference Tree Analysis (page 2) and Cube Analysis (page 4).

logo CUBE ANALYSIS

PointClear Relational Segmentation (PCRS) uses neither pure “tree” nor pure “regression,” but a hybrid directional analysis, often referred to as cube analysis, that balances principles of statistics with realities of today’s marketing budgets.

Designed to work well with lists of 1,000 to 20,000 names, it is sophisticated, objective and cost effective in predicting the likely success of B2B marketing programs.

For comparison purposes, reference Tree Analysis (page 2) and Regression Analysis (page 3).

  • Identify current addressable pain and projects
  • Segment the results by opportunity, timeframe, and budget
  • Distribute the hottest opportunities to field sales
Strategic targeting approach needed

Using a traditional database clean-up and database marketing approach, ABC Software cleans up one list at a time and runs a lead-generation program into it. Predictably, less than ideal outcomes occur.

While the cleaned database now has updated contact information, the lack of a strategic targeting approach means the clean-up has added no segmentation or prioritization value to records. The large investment of dollars and resources has failed to provide high-return direction or projected potential for Phase II database marketing. This assumes, of course, the clean-up initiative hasn’t broken the budget and left nothing for a Phase II marketing initiative.

Without the intelligence needed to hierarchically rank the valuable records that deserve to be contacted, a traditional database marketing program targets all cleaned names in the database, when only a fraction of the names warrants investment.

There may be limited success with lead generation, as shown in Table 2.

Table 2

Response Rates from Traditional Phase II Database Marketing
Source database or list Lead rate by source
C1 - License-only base 5%
C2 - Maintenance base 7%
P1 - Prospect list 3%

But return will never match potential because there is no deployment of dollars and resources against segments known or predicted to generate best return. No one knows why small successes were achieved or what portions of the list generated above or below average returns. Without benchmarks, there are no performance metrics and no wisdom to measure success and apply to future programs.

Bottom line: a lot of dollars, resources, and time are wasted as results fall

… along with these objectives for Phase II database marketing:

logo

dramatically short of potential. But there is a more efficient, cost effective way to run clean up and marketing initiatives that will generate higher returns.

The PointClear Relational Segmentation (PCRS) approach predicts success

While traditional clean-up and marketing initiatives take a freestanding “flat file” approach, PCRS marketing links multiple customer and prospect databases in a relational manner.

The underlying assumption is that various “cubes,” (groups of like companies) can be tested with differentiating characteristics to determine the most valuable segments, and the knowledge can be predictively applied to generate higher return on future programs.

Steps include the following:

  • Identify discriminating characteristics among the databases and lists
  • Segment the lists into small homogeneous cubes, or layers, of like companies
  • Conduct tests to profile and uncover opportunity in the cubes
  • Analyze cubes to find high-return segments and rank them as separate mini markets
  • Use this intelligence to fully fund the right model for future programs

Table 3 takes ABC Software’s three databases and relationally segments them into five cubes for testing and analysis.

Table 3

Setting up Cubes for Testing
Test protocol Predictive Variable: Maintenance Predictive Variable: No Maintenance
Match2 of 2 Test 1:C2 Maintenance customers
and
P1 Prospects
Test 3:C1 License-only customers
and
P1 Prospects
No Match Test 2: C2 Maintenance customers Test 4: C1 License-only customers
Test 5: P1 Prospects
logo
nBase Marketing Equally Sized Samples
a. Cube test b. Sample size c. Cube lead rate d. # of leads e. ROI option 1 f. ROI option 2
Test 1
C2 & P1
200 9% 18 32 leads or
64% of results
with 40%
of spend
42 leads or
84% of results
with 60%
of spend
Test 3
C1 & P1
200 7% 14
Test 2
C2
200 5% 10
Test 4
C1
200 3% 6
Test 5
P1
200 1% 2
Totals 1.000 5% 50

While databases are being tested in this example, any of the following can be tested with cube analysis:

  • SIC codes
  • Revenue or number of employees
  • Annual growth percent
  • Decision maker level
  • Geography
  • Offer (price, bundling, terms, or delivery mechanism)
  • Media (telemarketing, direct mail, e-mail, or print ad)

Table 4 depicts PCRS marketing response rates for five equally sized test cubes. This presents a dramatically different picture from traditional single-database marketing response rates.

While a traditional program can only generate response rates for a single database, PointClear’s cube approach has revealed three segments, or test cubes—Test 1 (9 percent ), Test 3 (7 percent), and Test 2 (5 percent)—that generate returns equal to or above the aggregate average lead rate of 5 percent. This provides the market intelligence needed to fully fund and roll out programs specifically targeting high-return segments.

Also note ROI data captured in columns e. and f. in the table. Column e., for example, shows that tests 1 and 3 generated 32 leads, or 64 percent of the program’s 50 leads, from 400, or 40 percent of the 1,000 prospects contacted. Similarly, column f. shows the top three cubes generated 84 percent (42 leads) of the total (50) with only 60 percent of spend. This information helps marketers balance return against investment and determine optimal program deployment and funding for future campaigns.

While Table 4 shows cube results from equally sized samples, this model can be even more useful by predictively weighting test segments. For example, any company name in the Test 1 (Table 3) cube appears in two databases (C2 and P1). Intuitively we know that a company of this nature is likely to be a more highly qualified prospect, and we increase the sample size of this cube and the Test 3 cube in Table 5.

Conversely, we expect names appearing on single-occurrence lists (Tests 2, 4, and 5) to generate a lower return than multiple-occurrence lists and only test 100 names for each cell. As lead rates carry over from our equally sized sample testing, the value of predictive weighting becomes apparent. Given the same total sample size (1,000) and lead rates from the previous test, predictive weighting has now generated 66 leads or 32 percent more leads than the 50 generated in the previous test.

Table 5

nBase Marketing Equally Sized Samples
a. Cube test b. Companies in cube c. Sample Size (% of segment) d. Cube lead rate e. # of leads
Test 1
C2 & P1
1,000 400
(40%)
9% 36
Test 3
C1 & P1
1,000 300
(30%)
7% 21
Test 2
C2
1,000 100
(10%)
5% 5
Test 4
C1
1,000 100
(10%)
3% 3
Test 5
P1
1,000 100
(10%)
1% 1
Totals 5.000 1,000 6,6% 66
logo

about PointClear

PointClear is a prospect development company. Founded in 1997, the Atlanta-based company helps B2B companies fill their sales forecasts with qualified opportunities. PointClear closes the gap between marketing and sales— nurturing leads, engaging contacts and developing prospects until they’re ready to close.

The PointClear Relational Segmentation (PCRS) approach provides market intelligence needed for companies to fund and roll out programs that generate a high ROI.

At the same time, we’ve learned Tests 1 and 3 generated the highest lead rates with Tests 4 and 5 performing poorly. Marketers have again gained the intelligence they need to fund and roll out future programs that will generate the highest ROI.

Summary

Companies often struggle with disparate customer, prospect and partner databases as they try to leverage existing lists to drive cross-selling, up-selling and new sales. Often, traditional clean-up programs only replace outdated data with fresh data, but do not target “best names” to ensure profitable campaigns.

Now, there’s a better, more effective alternative to either contacting all names in the databases or throwing out every name and starting over. PointClear’s sophisticated relational segmentation techniques balance the principles of statistics with the realities of today’s marketing budgets. They can predict the likely success of B2B marketing programs, helping eliminate wasted dollars, time and resources.

Ultimately, PointClear clients can achieve a higher number of more profitable sales in a timely manner at a lower cost.

Developing Prospects. Driving Revenue.

PointClear, LL C 3550 Engineering Drive, Suite 300 Norcross, GA 30092
Toll Free: 877-582-9909 Fax: 678-533-2703 Call: 678-533-2700
inquiry@pointclear.com

www.pointclear.com


logo
TAG’s article library is optimized by search engine optimization company Vayu Media, provider of SEO Services to technology companies nationwide. Vayu Media develops online strategies to grow businesses through increased sales, leads and internet brand awareness.

 

Member Status

Facebook Fans