|
|
 |
Marketing optimization
|
As more accountability pressure is put on marketing functions,
the demand for technologies that allow the accurate measurement
and optimization of marketing effectiveness is increasing. The Business
Optimization group focuses its research and development in the following
key areas in marketing.
|
|
|
| |
|
|
| |
|
The purpose of direct marketing
optimization is to develop advanced customer lifetime value and loyalty
metrics, which allow companies to
| · |
segment customers based on their lifetime value
in addition to variables such as demographics, behaviors, preferences,
etc., |
| · |
value individual customers over variable time
horizons using state-of-the-art customer lifetime value models, |
| · |
analyze historical data and derive optimal targeted
marketing plans (sequences of marketing actions) for each customer
(segment), |
| · |
allocate a given marketing budget over an optimal
subset of customers and marketing actions such that marketing
ROI can be maximized. |
We have developed the Customer Equity & Lifetime Management (CELM)
solution, which is a methodology and software that allows marketing
managers
| · |
to assess the efficiency of existing marketing
plans and their impact on customer transition to higher loyalty
and value states, and |
| · |
to derive the optimal marketing plan and budget
per customer segment for variable planning horizons (e.g. weeks,
months, quarters, etc.). |
CELM has been used by IBM Research and Global Business Services
in several consulting engagements with customers in various industries.
In 2005, CELM was awarded the Best Practice Prize of the INFORMS
Society for Marketing Science for the Finnair case study.
|
|
|
| |
|
|
| |
|
|
The purpose of mass-marketing optimization is to design an optimal
mix of marketing actions and media to maximize awareness, product/service
trial, sales and market-share levels. When attempting to measure
and optimize mass marketing effectiveness, the key challenges most
marketing practitioners face can be expressed by the following two
questions:
| · |
How can we effectively estimate a quantitative
(statistical) relationship between each marketing tactic (actions)
and impact metrics (sales, market share, etc.)? |
| · |
Assuming such statistical models have been estimated,
what would be the optimal media diversification policy to maximize
ROI for a given amount of marketing resources? |
The Business Optimization group has developed a mass marketing
optimization solution that includes advanced technologies from machine
learning and portfolio optimization theories. This solution is a
way to address the above two questions effectively when marketing
data is available. This solution is being used by IBM customers
to support marketing decision-makers in their efforts to design
optimal marketing plans and budget allocations.
|
|
|
| |
|
|
| |
|
|
The purpose of causal discovery is to assess and explain the effect
of marketing actions (e.g. campaigns, rebates). Current tools and
methods are based on predictive models and other data mining techniques
that do not model how sales are impacted by marketing activities.
They generally do not take into account the context (e.g. past campaigns,
season, marketing policy, etc.) in which past decisions have been
made.
For that reason, the Business Optimization group has developed
an approach based on causal discovery techniques that makes it possible
to assess the impact of marketing actions under various conditions,
for example
| · |
when future marketing policies are not the same
as in the past, |
| · |
when external factors such as product availability,
weather, or season also influence sales, and |
| · |
when the business environment changes. |
Such changes can only be managed if the effect of marketing actions
is explained as opposed to predicted. Causality is applied across
both marketing areas (direct and mass marketing) and is currently
focused on sales time series.
|
|
|
| |
|
|
| |
|
| [1] |
Labbi, A., Tirenni, G., Berrospi, C. and Elisseeff, A.
"Customer Equity and Lifetime Management: The Finnair Case
Study"
to appear in the Marketing Science Journal, 2007. |
| [2] |
Labbi, A.and Berrospi, C.
"Customer Lifetime Value based Marketing Planning and Budgeting"
to appear in the IBM's System Journal, Special Issue on Business
Optimization, 2007. |
| [3] |
Labbi, A.
"Customer Experience Management — Best Practice Research
for Customer Retention"
3GSM World Congress on Telecommunications, Barcelona, Feb. 13-16,
2006. |
| [4] |
Tirenni, G., Labbi, A., Berrospi, C., and Elisseeff, A.
"Efficient Allocation of Marketing Resources using Dynamic
Programming"
The SIAM Int. Conf. on Data Mining, Apr. 21-23, Newport Beach,
CA, USA, 2005. |
| [5] |
Elisseeff, A., Gomez, M. and Labbi, A.
"Optimal Campaign Planning: A Machine Learning Approach"
The INFORMS Marketing Science Conference, Jun. 16-18, Atlanta,
GA, USA, 2005. |
| [6] |
Labbi, A., Lindfors, K. and Iverson, B.
"Taking
CRM to a Higher Level"
Hospitality Upgrade Magazine, Fall 2004. |
| [7] |
Labbi, A. and Rump, V.
"Building Equity across the Customer Lifecycle: Dynamic
Marketing Optimization for Telecommunications Service Providers"
IBM Global Services White Paper G299-0693-00C, Aug. 2004. |
|
|
|
| |
|
 |
| |
|
|
|
