At outside the box we are verystrong on data because we know that essentially our remit is to continuallyimprove the ratio of profit to investment from marketing. It is only throughcontinual testing and refinements that this will be achieved – it goes withoutsaying that without data skills and a good solid database for information, wecannot produce effective direct marketing.
We also have data analysts and our own proprietarydemographic software packages; we have a relationship with one of the UK’sleading data specialists.
We have many years of direct specialist experience in dataplanning and management. Our skills in data are wide ranging, including online,mobile and offline...and also include the following areas
- statistical analysis
- psycho-geo demographics
- response modeling – scenario planning
- testing for improvements
Our approach
We would suggest a three phaseapproach – basics-enhancing-innovation:
Phase 1 – basic data management techniques
Our goal is to develop the datafunction for clients so that we can create sophisticated CRM and customeracquisition programmes (by brand) that deliver real ROI. We cannot do this ifwe do not have the basics right.
There is an essential datacleansing and standardisation process that will help provide a solid foundationfor moving forwards.
We could look at some or all ofthe following:
Cleansing the data– cleaning up and standardising address details by comparing to PostcodeAddress File (biggest and most up to date file in UK)– identifying and rectifying poor quality contact name data by comparing to thetotal UKuniverse.
Validating– carrying out name and address validation against the current UKelectoral roll of approximately 40million adults. Telephone validation is alsocarried out at this stage, screening against 17 million records, identifyinglive and disconnected numbers, helping with telemarketing campaigns.
Deduping –We cleanse the data to find duplicates at address level
Suppression– We would identify and suppress gone-aways (using the recognised gone-awaysuppression file 13.4m records, 98.2% accuracy); we would identify and suppressdeceased records (using the Bereavement Register); we would also screen againstthe Mailing Preference Service and Telephone Preference Service.
It is essential that this activity is carried out on a regularbasis to maintain the integrity and legality of the file.
Phase 2 – enhancing the data
At this stage, we can enhance the data by adding some or allof the following profiling information: gender, age, marital status, length ofresidency, personal income, geo-demographics.
We would also look at historical customer information whichwe call RFM or Recency, Frequency, Monetary value.
· Recency– how recently did the customer buy?
· Frequency– how often have they bought?
· Monetaryvalue - how much did they spend?
We should also look to append:
· The type of holiday purchased
· Which incentive a customer has responded to, bybooking
· Which communication channel a customer hasresponded by
Phase 3 – innovation in data usage
Once we have carried out stages 2 and 3 and are adding tothis information on a regular basis, we have the framework to be much moresophisticated with the data usage.
Segmentation– we can clearly define customer segments for effective targeting. We can create ‘pen pictures’ of thesecustomer groups to show visually:
· Who they are
· What they do
· What they buy
· How they buy
· When they buy
For example – ‘Top e-bookers’ (book regularly online, typically youngercouples, spend lots, good, loyal, profitable customers); “one off bookers’(only ever booked once, brand fickle, typically) etc.
Effective CRM recognises the relationship the customer haswith the brand and leverages it effectively. Solid customer information will help us to fully inform the customerrelationship management (CRM) strategy. This is essential to ensure that we arehitting the right individuals with the right message and offer and via the mostappropriate communication channel. RFM analysis will help greatly here as wewill know when to hit them, how often, what it is worth spending on them (basedon their profitability).
We can look at the data and determine where customers may bebetter targeted by other group brands based on their profiling information.Additionally, we can move customers from one database to another as they moveinto different life stages – i.e. Club 18-30 data – where does it go when theindividual reaches 31?
We could also, (should you choose to do so) use the data asan income stream – large volumes of clean, insightful data create additional revenue that can be used to fund newmarketing initiatives.
Customer acquisition-Accurate profiling will help us define the very best cold lists formailing. We simply find our most profitable customers and find more like them,utilizing what we know about offers and timing etc to bolster results.
It will also help us find the best media options to test.
Affinitypartnerships – with accurate profiling information, we can seek outlikeminded, non-competing brands with customers that have similarcharacteristics. We can “piggy back” their mailings and explore the opportunityfor list swaps both of which are very cost effective mechanisms.
We can seek out brands with customers that have similarpurchasing attitudes. For example, late bookers may also have pay as you gophones, underlining their attitude to spend and risk. Or early bookers may beon fixed mortgage deals. The data will help us greatly with this.
We should use all our data information to provide regularreports so we can continue to learn and add innovative thinking over time.
Given the level of activity required to carry out effectivedata management and manipulation, companies generally need to decide whether todevelop the necessary skills in house or outsource.
Response prediction and analysis
It is critical that we apply our experience and knowledgewith regards to likely response and conversion rates. However as there aremultiple factors which affect these numbers we use a predictive modeling toolto consider the scenarios and variables.
We use these scenarios before we design, print or mailanything because we need to make sure that the economics of the project work,the model is attached.
Over time we collect the data from our testing specificallyon your brands and this enables us to refine the model which in turn enables usto increase the ROI on a continual basis.
