Red Olive attended the KXEN “Insider Insight” event at the Doyle Bloomsbury Hotel in London recently. I would guess there were about 200 people there. The day was introduced by Matthieu Chouard, KXEN’s European SVP and John Ball, KXEN’s CEO presented a summary of KXEN’s successes to date and their current propositions.
KXEN company plans for 2013 from John Ball, CEO
From a product perspective John outlined the capabilities around operationalizing predictive modelling through KXEN’s “Modelling Factory” product and analysing social media data. He also outlined a customer focus in the sectors of Telecoms, Financial Services and a combination of e-business and retail.
Considering areas for the application of predictive modelling within each organisation, John outlined CRM, Risk, Fraud and Operations. Specific risk control applications were credit scoring and compliance with industry-specific legislation. The applications in operations were much broader, covering KPI forecasting, quality assurance and the identification of anomalies, usage forecasting and pricing optimisation. A customer example included the American Automobile Association who use KXEN software to predict dynamically where breakdowns are most likely given current conditions and then deploy their breakdown vehicles to minimise waiting time for motorists.
John Ball noted KXEN’s main benefits as:
- No organisation should be constrained by their ability to build models. KXEN’s message is that KXEN software makes model-building incredibly quick.
- KXEN is very easy for people to use. You don’t need to be a statistician to use KXEN.
- The improvements in productivity are measurable and significant, both in process output and in people improvements.
- KXEN is particularly suited to high dimensionality “big data” problems where there are many hundreds or thousands of potential input variables, such as those using derived variables in the Telecoms sector.
Continuation by Erik Marcadé, CTO
Erik took over from John and focused on KXEN’s application for e-businesses. The focus was on the need for models handling thousands of variables and the mantra was “develop a very wide model first then automatically slim down.” The thinking is to provide the ability to iterate around new models very quickly, for example predictors of the success of e-mail marketing.
Example clients cited included screenrush.co.uk an on-line films web site where film recommendations are provided by KXEN, and allegro.com where product recommendations are provided by KXEN and have achieved a 500% uplift in sales.
There were many other presentations on more detailed and specific subjects during the day including an interesting one from Dr Margaret Robins, a statistician at Aviva Insurance. Aviva provide general, life and healthcare insurance and use KXEN against their customer database. Their aims are for cross-sell, win-back and acquisition. Margaret explained clearly why multiple models are needed for different customer types, and why she is sceptical both of models containing hundreds of variables and of gains charts which are aren’t smooth.
Overall it was a good informative day for anyone looking to find out more about predictive modelling in general and about KXEN’s offering in particular.
Growing your internal predictive analytical capabilities with Red Olive
We’re finding that many companies with a strong Business Intelligence history are starting to explore predictive analysis and descriptive analysis, i.e. data mining, based on factors such as their overall maturity, the desire to interact in a personalised way with individual customers, the aim to take advantage of new, rich “big data” feeds such as social media and so on. Red Olive has a great deal of experience in this field in many different industries. For more information about Red Olive’s predictive analytical and data mining capabilities click here, or to contact us click here. If you’re aiming to grow your internal data mining capabilities then we can help you with a tailored people development roadmap and development of both soft and hard skills from sharper presentations and business engagement to key statistical understanding and hands-on skills.