|
- For
quick link
For organisations wishing to
undertake data related, analysis, targeting or reporting tasks in-house, rather
than outsourcing them, Tech4T can supply the specialist software, tailored
training and single contact 'one-stop' support. Software can generally be used
stand-alone, configured as a tailored combination of modules to meet specific
marketing, analysis and business objectives, or used to complement existing
I.T. systems.
Most software provided by Tech4T is
also used extensively by our team for the specialist services we offer - hence
the reason why we can offer 'single point of contact' support and training
across software that have different authors. NB. We take a consultative
approach to first understand your needs before recommending an appropriate
software and training solution.
File
conversion - If you
have to audit data or undertake file conversion - optionally computing
changes during the conversion - you need
DBMS/Copy.
This powerful utility also converts to/from over 80
different database and statistical analysis formats. For additional data
understanding and fast drill-down data exploration with tables, graphs and
reports, use
DBMS/Explorer included free with
DBMS/Copy!
High
speed sorting on huge databases and ETL data transformation
- CoSort
Need to read
files into SAS?
Then
DBMS/Engines
is what you need.
Database comparison
-
DBMS/Compare
- a superb piece of software that allows you to compare content - field
by field, record by record - across multiple databases or statistical
analysis files (inc. SPSS and SAS) in a single pass. This is perfect for
complex data audit work and to identify changed records prior to behaviour
change or fraud analysis.
For data
cleaning and removing unwanted characters
- optionally on
the fly for de-duplication of foreign data, see
F
Getting started
- or need the perfect reference book that helps you
apply statistical analysis to data-driven marketing, buy
The
New Direct Marketing
Lifetime
value analysis.
How to
measure customer cost and profitability. Understand purchasing patterns
and find out how to apportion promotional spend correctly in order to
maximise return on marketing investment. The approach you take would
depend on how you wish to define Lifetime Value - total sales value per
customer over X years, total sales value by product group after say the
initial relationship building period, profit contribution over time
however defined - i.e. how do you take account of order returns or
transaction cancellations based on the cause, etc. Once defined, you
then need to develop an approach and determine the computation needed.
This would include first bringing together sales values, profit
contribution, marketing spend, sales and customer service costs, etc.,
applying the necessary calculations and then correctly apportioning
values to each customer in a format that suits what you next wish to do
with the information - using for segmentation or profiling, targeting,
looking at change... Techniques to take include file and record merge,
aggregation, file flattening, split file processing, predictive
modelling and ‘marketing effort’ scoring.
Profiling, Database Segmentation and Predictive Targeting.
See also RFM
-
Recency, Frequency, (Monetary) Value.
Who are your customers, what do they look like? How do their
personal, business and trading characteristics differ? When profiling
and segmenting your customer base there are several approaches that can
be taken. For B2B it might be as simple as grouping by company size,
number of employees, etc., or with B2C grouping by age, gender,
demographics, lifestyle, etc. This can be enhanced by merging external
data from mailing lists, surveys and marketing statistics and also by
deriving new fields such as lifetime value. However you can enrich this
process by using statistical analysis. Either look for natural groups or
clusters in your data, letting software such as
SPSS determine the strengths of your variables (fields),
or look for groups (segments and their profiles) based on their
differences to take a particular action - buy or not buy blue widgets,
etc., using predictive modelling. Once segmented and profiled, you can
then see how groups of similar customers compare over time after being
for example, subjected to different kinds of marketing treatment.
Techniques for segmentation and profiling (putting descriptions to the
segments) include correlation, weighting, ranking and scoring, factor
and cluster analysis, predictive and response modelling...
Modelling
for improved campaign response and reduced fraud.
To identify
database segments more likely to respond to specific offers and to
predict future sales and purchase activity together with customers
likely to cause bad debt. This is really an extension of database
segmentation listed above. Techniques typically include sampling,
marketing profit and loss, predictive and response modelling, (inc.
CHAID) modelling, logistic regression, discriminant analysis...
Churn or
customer attrition analysis
-
Find out why
customers defect, develop corrective retention and acquisition
strategies and help define service levels to maximise loyalty.
Techniques usually applied include predictive modelling, competitor
analysis, forecasting...
Forecasting
-
For demand
business forecasting use
Analyse ALL your data
-
FastStats Discoverer
Free
Venn software for graphical overlap analysis
- Venn5
E-mail
broadcast solution -
e-campaigner
REPORTING
As you can see from the above, there is
often more than one piece of software that can be used to achieve a
specific task. In fact, depending on the objective, the quality of your
data and the methodologies used, there may well be a need for several
specialist software tools to maximise productivity and deliver the most
accurate results.
Therefore if you are in any doubt, please
call Tech4T on 01733 890790 and ask to
speak to one of our specialists who will give you some free advice.
|