Artificia Omnis Integration - Analysis Systems - Artificial Intelligence

In this section ...

Developers of the Internet database for rescued pets in shelters across the UK

Online Solutions

CRM Systems - Art and Science

By Dr. Caroline Wilkins

Customer Relationship Management, CRM, is a concept that means many different things to many different people.  At worst, it is a horrible acronym that strikes fear into those who are too busy to keep up with the latest business jargon.  At best it is the fine art of achieving a good relationship with a customer.

Example - The Restaurant

When business is done on a small scale with a customer base that it is feasible for the business holder to get to know personally, CRM happens naturally.  Take the example of a restaurant.  A restaurant owner may know his regular customers well.  He may have been anticipating their arrival, have saved their favourite table, greeted them by name and asked if they would like their favourite wine.  He may offer greater flexibility in menu and have some suggestions that will suit their tastes.  He may offer complimentary drinks or dishes to make his regulars feel especially valued.  As a result, the customer will visit regularly and tip well.  He may also feel more comfortable passing his compliments to the chef when appropriate and give the restaurant owner a better idea of what his preferences are.  The customer will often go on to recommend his favourite customer to his friends and bring groups back in the future.

The example illustrates a business knowing who their customer base is, what they like, targetted promotions, cross selling, loyalty rewards, customer feedback and member-get-member propagation.  All these are elements that we seek to achieve within a CRM plan for a bigger business.  The challenge is to create a strategy to implement these ideas in a situation where there are too many customers for one person to interact with personally.  We can turn to technology to facilitate.


A common CRM technique is the Loyalty scheme.  Customers will be offered loyalty cards, or internet equivalent personal identifiers, on which they are given points when they spend money.

The points may be accumulated and then redeemed against goods or services.  The system may be as simple as issuing points at a given rate, e.g. 1 point per pound spent, and redeeming against a broad catalog of gifts.  The purpose of the loyalty scheme is to encourage a customer to choose a particular retailer over a competitor because they know they will get points which they can later exchange for something they would like.  A loyalty scheme, in this simplest form, breaks down as a strategy when every retailer in the market place has a similar scheme.  The customer then keeps a wallet full of cards and does not allow the loyalty scheme to affect his choice of retailer on a particular shopping trip.

When this situation occurs and retailers find they have to operate a loyalty scheme simply to keep up with their competitors, the competitive advantage then lies in the more refined areas of CRM - i.e. analysis and targetted promotions.  General promotions are not enough.  A retailer can offer a loyalty promotion such as double points on purchases.  Their competitor soon follows and soon the playing field is levelled again.

The clever retailer then starts to look at the information that his loyalty scheme is bringing in about the behaviour of his customer base.  If the scheme has been designed well and the data has been collected in a sensible way, there is a wealth of information that can help the retailer understand his customers.


Most retail customers can be segmented according to the frequency of their visit and the value of their spend.  In its most coarse resolution, this gives a graph with 4 regions corresponding to the following:

  • Customers who visit infrequently and don't spend much when they do
  • Customers who visit frequently but don't spend much
  • Customers who visit infrequently but spend well when they do visit
  • Customers who visit frequently and spend well each time

Graph showing value against frequency of purchases

CRM Techniques

Having identified four segments, the retailer must now create a strategy to migrate customers towards the High Value-High frequency segment and retain those already there.

To increase the frequency of a customers visit, you might issue them a token which is only valid for a limited period.  If you want to increase the value of their purchase, offer a promotion that is dependant on purchase value e.g. a petrol company might offer a money off voucher to customers who spend more than a certain amount on petrol.  In that particular market, the behaviour is quite easy to identify and manipulate.  Many people put £10-£15 petrol in their car.

A voucher that gets issued when they spend more than £15 will push a large amount of that user group into the next segment of purchase value.  Having issued a token, the customer will want to spend it, so they are more likely to return.  This will increase their loyalty and will also increase their frequency if they were previously less than 100% loyal.

There are many techniques that can be employed to migrate customers to more valuable segments and retain those already there.  A CRM consultancy company like Artificia can analyse the market and make recomendations to help you select techniques and form your strategy.


Where budgets are limited, it may be best to identify a particular segment that has a high potential gross value and seek to migrate them only.  The nature of the market becomes important.  We must ask questions such as:

  • Are the low frequent customers shopping elsewhere?  If so a win-back strategy is in order.
  • Are the low value customers unable to afford greater purchases or is the retailer failing to inspire them?  A greater range or more targetted selection of products may be effective for the latter case.
  • Is it possible to change the frequency with which customers shop?  If the product is petrol and the customer is 100% loyal, then their low value or low frequency may simply be due to a low rate of car usage.  Even the most attractive of CRM promotions are unlikely to make a person want to drive more!  Having said that, all things are possible and it sometimes takes a brave strategy to drive consumers into previously unimagined profitability.  In this case, we might recommend promoting driving as a lifestyle issue.
    We might identify places for our customers to drive their families too at the weekend and arrange concessionary tickets for attractions.

If having evaluated the potential gross value of under-performing customer sectors, you decide it is not cost effective to implement a promotion, it may be that the most sensible strategy would be to concentrate on retaining the High Value-High Frequency segment.  A popular approach to this is the "Gold Star Treatment" i.e. offer your most valuable customers membership/access to a exclusive scheme with special promotions that will make them feel valued and important as customers.  This might be, for example, a "Frequent Flyers" card, a "Gold/Silver/Platinum" credit/charge card or membership of a restricted part of a website.

In the ideal situation, a CRM system will have a range of strategies to achieve positive migration of segments as well as the retention strategy discussed above.  The emphasis may be balanced and tuned to create a feedback system that makes maximum use of the budget invested in the system.  That is where dynamic analysis of data is important.

Artificial Intelligence and the Artificia systems approach

Artificia can perform initial analysis of customer behaviour, based on whatever data is available.  Sometimes this will be transactional data from an existing loyalty scheme.  It could be score tables from computer games or a websites hit logs.  From that, and research of your marketplace, we can identify the initial CRM techniques and form a strategy.

Having rolled out a CRM strategy, it is important to assess its success.  Customers can then be segmented according to their response to a range of promotions.  This gives a better idea of what sort of strategies are going to be most cost effective.  There is no point wasting CRM budget on consumers that wont respond.  It may be that a different strategy is required for a non-responsive sector.

Most systems would involve a human analyst performing this sort of analysis.  A truly sophisticated system would have an artificial intelligence performing real-time dynamic analysis and reacting to changing consumer behaviour.  Artificia intends to lead the world with A.I. driven CRM systems.  If you want to break new ground with CRM, then it is time to speak to us about the possibilities that CRM could hold for your business.

We make CRM a science.

Go to top

Contact ArtificiaA.I. ApplicationsAnalysis SystemsOmnis IntegrationOnline SolutionsHome