Coupa and Ariba -- Analyzing Comparative Claims (Part 1)

Last week, Coupa launched an attack campaign against Ariba. Today, I'll start a series in which I'll analyze Coupa's claims and compare them to reality. My goal is to clear the air for those who read the campaign's claims and want an unbiased, outside opinion. In this first post today, I'll tackle a few of Coupa's initial claims and see whether they hold weight -- or not. Later this week, I'll analyze the rest of Coupa's claims against Ariba, and in a final post, I'll put out a few claims I'd make if I were Ariba going up against Coupa (I'll also provide Coupa the opportunity to respond to these in a later column, but to be fair, only after the post goes up; after all, Ariba did not have any advance notice of Coupa's campaign against it). Let's begin.

First, let's tackle ease of use. Coupa claims that its system is "designed with the end user's experience in mind, and characterized by broad adoption at customers" compared with Ariba, which Coupa claims is "not known for ease of use." In its description, Coupa appears to confuse some of its features (e.g., item reviews, supplier ratings) with actual usability, but that's not my beef with this claim. What I dislike about it is that Ariba is not as bad as Coupa makes it out to be. In fact, I recently interviewed a shared-service buying resource at an outsourcing firm and asked him which system he preferred -- Oracle or SAP. He responded, "Ariba."

Now that's from a user -- a power user, in fact. Moreover, Gartner published a review in June 2009 calling Ariba's P2P application "excellent" for ease of use. So what's the verdict on Coupa's claim? It's neither right nor wrong, but it is misleading. Coupa may be easy to use for simple requisitioning in the SMB market, but when it comes to a complex environment, Ariba comes out on top over SAP and Oracle (and Coupa, for that matter, which has only a limited track record of scaling in Global 2000 environments, at least to date). Seriously, riding a Vespa might be easier than riding a Ducati, but we all know it's not a fair comparison. Still, both can get you killed if you don't wear a helmet, or if you do something stupid, so be forewarned, whichever P2P or motorized path you plan to take.

Next, let's talk configuration. Coupa claims that its "Quickstart deployment is so easy, you'll be able to generate your first PO in less than an hour. Application configuration is so simple, business users can make changes on the fly." In comparison, Coupa claims, Ariba has "potentially time-consuming and expensive deployments and customization requiring extensive IT and consulting resources. Maintaining the system and changing workflows may require dedicated technical resources." Again, Coupa is neither right nor wrong, but it is quite misleading. Sure, with Coupa, you can start using the system in an hour, but chances are, you won't have it configured to the full extent possible. A few weeks to a month or two is more like it for a broader roll-out (Coupa's first larger-scale enterprise deployment took much longer, mind you, but you won't read that in the comparison).

Likewise, Ariba takes time to roll out, but it's getting better at it in a SaaS context, provided that back-end environments are not overly complex (you're probably fine if you're less than $7.5-$10 billion in revenue). Still, count on 3-6 months to be fully up and running, although it's possible to get started in targeted areas more quickly. A more disconcerting part of Coupa's analysis is its claim of the level of customization required with Ariba. With Ariba's SaaS platform, there's been a move towards platform configuration over customization whenever possible. This marks a big shift from its installed past, and Coupa is simply wrong and misleading in its claim about Ariba SaaS.

So what's the verdict? No doubt, Coupa can be up and running faster, but in most cases it's working in much smaller environments than Ariba. If a $2.5 billion company were comparing Coupa and Ariba, I'd bet the actual initial company-wide deployment time for each would be within two months of the other for full rollout -- not enough to move the savings needle if Ariba is the better solution for the organization. In my view, the better question to ask is the time it would take to get a specific dollar amount (e.g., $300 million) under management and full control.

The next topic we'll skim over quickly: "Value." I'm sorry, value is a not a fair comparison metric in my book, at least the way Coupa lays it out. It's so nebulous that, unless you want to get into total cost comparisons, the mere mention of "value" on such a high level should cause you to look circumspectly at the comparison. Sure, Coupa may not have Ariba's network fees or configuration costs, but Ariba may allow you to to drive better overall procurement controls and the ability to work more closely with preferred suppliers. This in turn allows you to capture rebates based on negotiated savings (which are tracked in spend and contract management tools), and profit from dynamic discounting programs more easily. See, it's not a fair comparison metric unless you want to create a customized grid based on your organization and hundreds of variables. The verdict: Coupa loses on this one by getting its claim wrong and by using "value" as a basis of comparison without all the details, including the areas in which Ariba is stronger.

Fourth, and finally for today, let's turn our attention to benchmarking. Here Coupa claims that its "multi-tenant cloud architecture allows for performance benchmarking against the community across dozens of KPIs" versus Ariba, which it claims does not offer similar capabilities. This is only partially true. Ariba does offer some benchmarking capabilities in other solution areas, yet not P2P, at least that I'm aware. Still, its SaaS architecture could allow it to move in this direction, no doubt. In Coupa's case, the move to offer cloud-based benchmarking is quite cool indeed, although I'm not sure how useful the actual benchmarks will be until it begins to cut them by industry and company size. You can't compare a $20 million non-profit/university with a $250 million financial services firm and hope to glean anything from a benchmark comparison. Still, I commend Coupa for moving in this direction. (Incidentally, Coupa shares a similar vision here with Ketera, which is also moving in this direction, and offers related benchmarking capabilities across its network today.)

Stay tuned for more analysis tomorrow, and plan to put on the flak-jacket, because I won't hold back from calling it as I see it.

- Jason Busch