IT-Analysis.com
IT-Analysis.com Logo
Technology Applications
Business Issues Channels Enterprise Services SME Technology
Module Header
Louella FernandesLouella Fernandes
Louella Fernandes
22nd April - Internet of Things: A New Era for Smart Printing?
Simon HollowayThe Holloway Angle
Simon Holloway
18th April - Virgin Media expose private email addresses
Craig WentworthMWD Advisors
Craig Wentworth
17th April - Box's enterprise customers step forward to be counted
Craig WentworthMWD Advisors
Craig Wentworth
16th April - Egnyte the blue touchpaper...

Analysis

SAS: a platform of choice for managing the customer experience
David Norris By: David Norris, Practice Leader - Analytics, Bloor Research
Published: 28th February 2007
Copyright Bloor Research © 2007
Logo for Bloor Research

It has become fashionable to doubt the ability of SAS to meet the challenge of the new paradigms of productivity and ease of deployment, and to meet the key marketing requirements of personalisation and optimisation. The RDBMS vendors point to their in-database data mining and ask why anyone would need an alternative. Rival analytics vendors have sought to marginalise SAS by emphasising their capabilities in segments such as the high end of precision modelling, or the really rapid delivery of models to staff without statistical and technology skills. Many may indeed have started to question whether it is a safe bet to remain with SAS, or whether they should look to migrate to a new platform.

It has been a few years since I last looked at SAS in any detail, and in the intervening years I have become aware of a big change in the market for analytics. Five or so years ago the requirement was still predominantly for models that accurately represented the core of the market; accuracy meant that very often those models were produced slowly and carefully and, once built, they tended to stay in the analysts hands, outcomes being transferred indirectly. That has been replaced by the rapid production of many more models that address specific business needs and, once modelled, the model has to be deployed embedded in a business process; and of course then has to be regularly monitored to ensure it remains performant. I am aware that all of this represents a change that many analysts feel uncomfortable with and, as so many analysts use SAS, it is often the technology that comes to be associated with that conservatism. So it has been very refreshing to look again at what is probably the most used analytics platform in the market and assess how it stands against the needs of one of the key users groups, Marketing, to manage the customer experience to drive loyalty and profit.

It is my increasing belief that what Marketers want is to be able to obtain industry specific support from one vendor, who is capable of supporting them across all channels used to interface to customers, and that they want global support that recognises local specifics. They want a tool that is easy to use and does not require high levels of technical knowledge to obtain good results, they want high productivity because they need to move and adapt very quickly, they want to be able to deploy solutions into production quickly and with minimal reliance on IT. Marketers are suspicious of big IT vendors because they do not see them as understanding their specific needs and invariably IT vendors see the solution to everything is lots more technology, at ever greater expense.

SAS has a long and established tradition of not being a jack-of-trades, but is a master of one—analytics—and through its use by many of the market leaders in key segments like Telco, Finance, Pharmaceuticals and FMCG has considerable vertical industry knowledge. Specifically in the area of customer experience I think that SAS are able to speak with the conviction and knowledge that most people only associate with specialist marketing support companies. Further SAS is global and is large enough to be able to offer global support; but it is also small enough to avoid the worst of the big company syndrome of anonymous one-size-fits-all mentality.

With SAS® for Customer Experience Analytics, which SAS have developed in conjunction with web data collection specialists speed-trap, SAS have an offering which may well break the grip that the current web analytics specialists have on web based intelligence, because they have taken things to their logical conclusion of not looking at web analytics narrowly but as part of an integrated whole, a view which ties in very closely with what I see being demanded more and more by Marketing executives who are really managing their brand effectively. Customer Experience Analytics enables organisations to collect data across multiple channels and predict future behaviour, and this can be done in real time, so it is possible to really influence the next step taken by the customer.

In terms of productivity, Enterprise Miner is one of the key tools in any analysis of the customer experience, and now it has a new graphical front end. The tool can be used with a standard set of parameters and the GUI will enable good enough models to be built with accuracy and speed. If further tweaking is required you can delve into the depths of the tool and adopt a more expert approach. This is a big change from the rather daunting old style interface and will lead to significant productivity gains. It may still not be as quick to use as a tool like KXEN, but the difference will be marginal and SAS has other attributes that are not present in tools like KXEN. In terms of precision SAS has probably the broadest of ranges of tools, and should have a tool to meet most people's needs. You can use alternative techniques and the system will then look at the results and suggest which is the best model to take. Whilst Fair Isaac may look at optimisation across the portfolio and suggest the use of multiple models on different segments of the data set, that is a probably going too far for the mass of the market.

When it comes to deployment, SAS have a Model Manager as part of their SAS 9 platform that caters for deploying and managing models (management here being the selection, maintenance, and continuous enhancement) across the enterprise and that is probably a more advanced solution than any of the competitors. The Model Manager will also look at how the model is faring in its ability to predict and manage the real world and will identify when it needs to be refreshed to maintain its power.

As for the argument about the in-database capability of Oracle, IBM and Microsoft, they do offer compelling reasons why, if you hold your data in an RDBMS, you should use their tool but that does not answer the fundamental question of whether it is right to hold the data you use for decision-making in the same structure that you use for operational support. I am firmly of the belief that RDBMS are good for the support of the operations of an enterprise, because they are optimised for constant short sharp updates; but they are far from ideal for large complex queries involving table scans and the like. I think that as businesses amass ever more data, holding everything in super large RDBMS data sets and then layering on operational, tactical and strategic operations is increasingly untenable. Once you move the data out of an RDBMS and into a data store more suited to the needs of analysis then the argument about in-database data mining loses its relevance. I believe that as the business benefits of adopting a fact-based decision making culture become more prevalent so there will be a movement away from thinking you can hold all of your data in a system which is not designed to support that way of working, and then the advantages of tools like SAS will become very apparent.

Much of what marketing want is about personalisation getting the right offer to the right person at the right time. But in the modern world that could still mean a plethora of offers all being made to the same candidates in the same time frame, and it is very unlikely that they will buy them all. Optimisation is required to ensure that the right balance is struck. Most people will agree that optimisation is a complex set of algorithms and there are not many who can do it well. Even most of the competition admits that SAS is one of the few that get it right. At present there is a widespread belief that optimisation is only really valid for big vendors like financial services and Telcos, but I see optimisation opportunities everywhere and believe that it has very widespread relevance, with the ability to readily show a positive ROI in most medium and large organisations.

So, overall, it is interesting to see that whilst the others may create niches around the core market, SAS is still probably one of the most competent solutions available, capable of addressing more of the needs of more of the market than the competition. The technology is supported by articulate and knowledgeable solutions support, and as SAS opens up its channels to market to sell through more customer-facing organisations that will continue to strengthen. So SAS users need not fear they are being left behind, and those looking to adopt should include it in their appraisal.

Free Download

SAS have made available freely downloadable PDF versions of this article. Please click the relevant link below:
UK Version | Global Version

PLEASE NOTE: This is an editorially-independent, sponsored article. If you would like your company, products or solutions to be featured on IT-Analysis.com please contact us.

Advertisement



Published by: IT Analysis Communications Ltd.
T: +44 (0)190 888 0760 | F: +44 (0)190 888 0761
Email: