Sometimes the biggest challenge for larger technology and services companies is that the marketing messages of individual components of suites and solutions often get lost in the broader company positioning. Having watched Ariba over the years, I would say this is certainly true (and this is not a knock on Ariba's marketing, but merely an observation on the challenge of being overly broad and deep, while needing to push a corporate message ahead of individual product and services lines). To wit, "Find it, Get it, Keep it" overly simplifies the concept of spend data management -- let alone advanced analytics -- on so many levels, but such high level positioning is necessary to explain the overall company mantra. In other words, it's a Catch 22 for Ariba from a marketing angle. SAP, Oracle, and Siebel have all faced -- or currently face, in the case of the two giants -- similar challenges as well given their breadth.
But Ariba is quietly beginning to make some noise around enhancements to its spend visibility capabilities. Back on their analyst day earlier in November, Ariba shared a number of spend visibility product updates and enhancements with attendees. I also had the chance to catch up with the solution managers in charge of the product the other day, and learned that Ariba is seeing significant traction (over a 50% bookings increase in spend visibility deals in 2006) of late, and is now serving 150 customers in the area, which puts them from a customer count as one of the largest providers in the market (if not the largest). Like Emptoris, Ariba offers both an installed and On Demand approach -- or a hybrid of both, depending on privacy and data requirements.
Ariba’s recent enhancements are focused on three areas: custom taxonomies, more flexible classification approaches, and a tighter feedback loop with customers. Regarding custom taxonomies, Ariba now enables the ability for users to classify to taxonomies that go beyond UNSPC (down to six levels). These can either be a custom defined taxonomy, or Ariba’s own sourcing taxonomy. For direct materials in specialized industries (e.g., food and beverage), this level of classification can prove critical to understand specific spend and supply elements to drive the most appropriate action and strategy.
In their latest release, Ariba has also integrated Bayesian (statistical) classification technology into their existing rules-based engine (it should be noted Emptoris is doing this as well). While Ariba still views rules-based classification approaches as being the real driver of consistency, Bayesian learning has the ability to improve the level of automation. This is because rules approaches require exact matches which must already exist in the system. In contrast, Bayesian approaches find statistically significant matches that may not already be in the existing rules. Ariba’s solution team told me that their investment in Bayesian learning is primarily to increase the level of automation and hence the speed, especially on refreshes. As the system continuously improves and adapts, the theory goes, everyone benefits with a higher degree of automation. I know the Spend Fool and others will probably debate the limits of this approach in a centralized enterprise deployment environment, but I toss it out here for debate, even if a decentralized network learning type of model might prove even more effective when it’s available in the future.
I realize this post is getting more technical than I would have liked, so I'll pull back the reigns a bit and leave the theoretical debates from the comments section. Ariba's last technical enhancement reminds me of something that Procuri has done a good job with over the years, which is providing specific areas interfaces for user feedback during the standard workflow in the application. In the case of Ariba's new spend visibility release, this enables practitioners to offer feedback on classifications directly via the interface with a customizable approval workflow to ensure consistency. This automates automating the process and ensuring maximum input into the enrichment process.
Since virtually all of the value of spend visibility deployments is based on what a company does with the insight the application generates, integrated expert advisory services can play a key role in helping companies maximize value from their deployments. In the latest solution release, Ariba has added a consultative-based assessment with all its services-enabled deployments in order to help users achieve a more rapid return on their investment by identifying potential opportunities and pointing users to the most useful bits in the reports. In addition, Ariba is now also offering additional fee-based services to help companies kick-start their efforts to identify and quantify initial opportunities that go beyond the free, bundled assessment discussed above. It should be said, however, that Emptoris, Verticalnet, AT Kearney, Procuri and others claim to offer similar types of enabling services capabilities as well with their spend visibility capabilities. But the proof, obviously, is in the individual solution.