Insight

Occam has built an exceptional team of analysts who support our clients to deliver a host of analytical solutions.  The team works in a multitude of ways to support our clients’ preferred operating models:  a number of clients have our analysts on a retained basis, working as an extension to their in-house analytical teams; others engage with the team on specific projects, or during times of peak demand. For some clients we are the sole supplier of analytical services to their marketing and business planning functions.

A Structured Approach to Analysis

At Occam we understand that customers are the most important part of your business.  As such, we help you understand your customers; who they are, where they have come from, what they have purchased from you, what purchases they are likely to make in the future, and how you can influence them to make those purchases.

By bringing a structured approach to customer understanding, we help you move from knowing very little about your customers to full customer understanding that underpins and influences your business strategy.

We recommend that the following areas are covered in an ordered way, but any single project can also be implemented to help drive increased income for you and your business.

1. Insights Report

The Insights report will give a client an immediate overview of their customer database and will include some of the following:

• Its value and how value is distributed amongst customers/supporters e.g. a Pareto review; RFV analysis (Recency Frequency Value groupings)
• Who their customers are e.g. age and gender profile and where possible a geodemographic profile (e.g. ConCensus profile for charities)
• Where customers are e.g. mapping to show either distribution or penetration of customers across the UK
• Whether churn is increasing (using an Attrition table)
• Which combinations of products are sold by creating a product matrix.

Additionally, we can customise the analysis to address any particular issues. For instance, we can run profiles for different types of customer or do a drivetime analysis to show how closely customers are located to a client’s stores.

Moreover, to aid further analysis work we will conduct a more technical review of a client database:

• How well are fields populated?
• What is the standard of data capture?
• Is any critical data not being captured?

Our experience of working with many client databases enables us to offer advice to clients who may be concerned about the robustness of their data.

2. Profiling

Understanding who your customers are is vital to the successful marketer. Our analysis team can provide multi-dimensional profiles of customers using both known attributes from the database and also by using tagged data from a 3^rd party. 3^rd party data does not necessarily have to be expensive. For example, for charities we can use ConCensus classification (see ConCensus profile below), but we are also able to use free data such as that from the last census.

Profiling outputs can be in the form of a report or a profiling tool. The tool runs in excel and allows the user to select the attributes he is interested in. It also allows for cross-tabulations e.g. do more females buy a type of product or do customers in the South East tend to be younger or spend more.

3. Post Campaign Analysis

Assessing the success of campaigns as they occur will allow clients to ensure that future campaigns are more effective. We can provide post campaign analysis to understand the following

• Response levels, ROI
• Most successful cells
• Trends over time and comparisons to previous campaigns
• Who is responding

We can also provide automated solutions to enable fast and standardised output.

4. Churn Analysis

Whether the economic climate is difficult or not, it is critical to understand how attrition is changing

• How does the lapse rate vary?
• Are changes normal seasonal fluctuations or something more?
• What impact does churn have on revenues, and what impact will it have in the future?
• Who is lapsing?
• Which lapsers should be targeted for reactivation?

We can work with a client to address these questions so that the client can then take steps to ensure attrition is kept to a minimum.

5. Lifetime Value Model

It’s easy to see the immediate effectiveness of a marketing campaign – you simply look at how many new customers or supporters you recruited, how much money it raised and how much it cost to run.  You can then establish whether or not it was a good use of money. However, it’s not so easy to see how these customers perform over the longer term e.g. a five year period, what their attrition levels are and when you will break-even, let alone comparing your different recruitment methods.

Occam has a solution to this problem. The LTV model can be used on any size of customer database – the more records the better! Users have the ability to answer complex questions about many aspects of customer value for themselves. Occam.LTV quickly identifies trends and leads the way to changes in marketing strategy. The ability to run “what if” scenarios across many different variables such as channel cost, attrition rates, interest rates etc, enables marketers to plan effectively for the future.

If the full LTV model is not required, then one-off analysis can be conducted to address specific questions. This is often run shortly before new budgets are set so that the analysis can be used to inform the best allocation of marketing spend to alternative recruitment methods.

6. Segmentation

Every customer on a client’s database is an individual and in an ideal world you would communicate with each person in a way tailored to them.  But clearly this would not be viable. What is viable is segmenting a customer database into groups of similar types.

We will take the customer data and where necessary tag it with geodemographic information so that a full picture of buying history and other characteristics is compiled for every customer on the database. Then we can work with the client to establish high level groupings if required, and finally move onto using cluster analysis to come up with segments.

However, we can tailor the approach used to suit the budget or client requirements e.g. we can base it on RFV or specific characteristics only.

Output is a segment allocation of the relevant customers as well as descriptions of the segments both in the form of pen portraits and more detailed spreadsheet profiles.

7. Propensity Modelling

The advantage of this type of project is that it pays for itself. It increases the ROI of campaigns by enabling the client to target the best prospects on a database and drop those unlikely to respond.

This is done by using data from the client’s existing customer base coupled with external data, such as geodemographic data, to give each individual a score based on their likelihood to respond. The score is developed by statistical modelling, usually regression, although it could be based on profiling or decision trees.

8. Forecasting models

Calculating and agreeing future customer revenue flows can be a time-consuming and tortuous task. Occam has developed a model which streamlines and structures the task by doing most of the hard work behind the scenes. It combines the current number and value of customers with predicted attrition and acquisition to forecast revenues over future years.

The excel based model is based on real data. Attrition rates by recruitment channel are forecast using historic data. The user enters new recruitment streams of customers based on knowledge of the marketing plan, and can also vary assumptions according to their own judgement of future market conditions.

The model then forecasts revenues and customer numbers at the touch of a button.

9. ConCensus Profile  – Charity Only

Occam has developed the ConCensus classification of charitable giving using two sources of data: Reciprocate (containing a wealth of charitable giving information) and 2001 Census characteristics such as age, family status, occupational and educational factors, type of housing etc. This enables us to classify the GB population at postcode level according to level of giving, charity sectors and demographic characteristics.

ConCensus is the only available geodemographic classification focused on charitable giving and is only available to Reciprocate members.

A ConCensus profile will show the type of donors a charity attracts e.g. their relative generosity and which other charity sectors they give to.

For instance do they attract:

• Principled Professionals (young, urban, idealistic supporters of Human Rights, Environment and Homeless charities) or
• Kindly Pensioners (Frequent but low givers to children’s and sometimes animal charities who are elderly and have a low income) or
• Family Fundraisers (generous mortgaged families who often give to Homeless and Mental Health charities)?