Now that we’ve dispensed with much of the cursory coverage of the SAP and Ariba acquisition announcement (you can read our main coverage thus far here, here, here, here and here), we thought it time to take out the elephant gun and take aim right at the pachyderm (that’s “very large animal with tough skin” for the non-vocabulary nerds) staring us down. If you’re curious about what the giant metaphorical creature is that’s not sure whether it is in “truth” or “dare” with us, you shouldn’t be: it’s Ariba’s supplier-funded, network-based business model. Over the years, we have attempted to analyze who pays for how much transactional connectivity through Ariba. In short, our analysis suggests that larger suppliers with high dollar volumes transacting through the network (regardless of the number of actual documents changing hands) with multiple trading partners are far more likely to pay substantial annual fees than smaller suppliers, those transacting infrequently and those with low-dollar volume transactions.
The fact Ariba has many suppliers paying thousands – sometimes tens of thousands or more, per year – to transact with multiple trading partners has helped offset the cost of Ariba’s P2P application for buyers, allowing it to compete on price as well as product. Yet many procurement professionals believe that suppliers will factor any additional transaction costs they may incur into their pricing. Still, Ariba has had the upper hand in this area because their SaaS-based P2P deals require companies to only work with companies through the network (Ariba CD customers have options to take suppliers off the network, something providers such as Vinimaya, Hubwoo, WALLMEDIEN, OB10, Transcepta and others can theoretically enable – and have, in certain cases).
Ariba believes it has pricing power in the network. In analyzing Ariba’s recent quarterly results, we explored the topic in detail:
A few years back, Ariba made the important strategic decision to shift away from buyers carrying the bulk of the burden of P2P fees (at least future costs) and putting these costs onto suppliers in the form of transaction-based (dollar volume) network fees. The result of this shifting business model gave Ariba much greater flexibility to pursue aggressive pricing for its SaaS modules (e.g., P2P, sourcing, spend analysis) to the point of, in a number of cases we are aware of, offering prices to win select deals that competitors would just walk away from. Curiously, these prices were not just limited to the eProcurement and e-Invoicing areas. Specifically, we can reference a number of situations where Ariba's quoted pricing for SaaS-based options over a 3-year term were between 15-30% that of the competition. However, the beauty of this pricing for Ariba was that if they could eventually drive significant P2P network volume in these deals -- even those starting outside of transactionally-focused areas – they would more than make up the difference by having supplier pay greater and greater fees. tweet
There are many differing perspectives to the merits of this strategy. By obfuscating the actual total costs to both buyer and supplier at the signing of a deal, Ariba is at an advantage in most cases because buyers need less of a capital budget to invest. However, some organizations believe that suppliers will ultimately just raise their prices to compensate for higher transaction costs over time. Spend Matters' own analysis at the time of Ariba's last major price increase was based on discussions with a range of buying organizations. A number of these companies planned to take select suppliers with high-dollar POs/invoices off the network because they are the ones who are hit hardest under the new pricing rules (these companies had the choice because they were CD customers – SaaS customers do not have this flexibility). tweet
After running the numbers for a larger buying organization transacting $500MM in annual volume using either Ariba Buyer, P2P or just the Ariba Supplier Network, we concluded that for the great majority of suppliers, the fee increases [and the fees themselves] are negligible. In fact, 70% of suppliers (as of the fall of 2010) would continue to transact for free, according to Ariba. However, high dollar volume transaction suppliers faced a significant increase under the new pricing model. Specifically, our analysis suggests that larger, financially healthy suppliers (with a low cost of capital) – and with multiple trading relationships and revenue going through the Ariba network – submitting fewer documents per annum are those that stand to disproportionately assume the costs of the network fee increases. Looking a this, we opined that "it could be argued that the new Ariba network fees place a larger burden on large suppliers to subsidize smaller ones, those most likely to leverage Ariba as a primary sales channel to win new business." tweet
Under the current fee structure, as of the 2010 price increase, our analysis suggests that roughly 12% of transacting suppliers were hit by network fees (this was using one transacting relationship example). But large suppliers like a Dell, Grainger, or Staples transacting $1.5 million per year would pay $2,325 in transaction fees annually (versus $1,500 before the price increase). This number maxes out per trading relationship at $20,000 annually, which would represent $12,903,225 in volume (up from a cap of $10,000,000 or $1,500 in annual transaction-based fees).
On the recent earnings call, Ariba hinted there might be room to raise network pricing further. In this regard, Bob Calderoni suggested, “I've always felt there's an opportunity to get a better yield than we currently do out of the network. I think today we have an effective yield of only about five basis points. Our pricing is 15 basis points, but it's free for some suppliers up to a certain level. So we got a net yield today of about five, maybe a little bit more than five basis points. The opportunity is certainly there to drive that yield up through business model changes, whether we change the basis points or we change some of the caps or some of the floors. I like to put some time in between those changes. And so I don't anticipate any changes in the near-term here, but that opportunity remains. And like we have in the past, you'll probably see, every 2-3 years, there'll be some tweak...whenever we do a model change, we tend to be above 30% [network growth]. So without a model change, we'll be similar to what you see today, high-20s, somewhere between the 20% and 30% range.”
The fundamental crux of the deal and whether it is a huge financial success (or not) based on the current network operating structure and arrangements for SAP and Ariba will be whether or not the continued fee structure remains viable. Regarding this specific topic, we observed a couple of weeks back that:
As Ariba debates fee increases – an increase in 2013 would be within the 2-3 year window that Bob suggests in the above quote – its competitors for network connectivity are increasingly providing identical transactional services at fixed (and sometimes free), volume-based prices (e.g., Basware, OB10, SAP/Crossgate/Hubwoo, Tradeshift) that buyers have the option of paying for suppliers in which they often receive a material discount for doing so. Moreover, some competing connectivity offerings are beginning to move further down the path embedding discounting programs as a core component of their offerings where for Ariba it is secondary … For Ariba to justify any type of fee increase to suppliers, we believe they'll need to embed low APR early payment financing options at the core (which could easily pay for itself by allowing Ariba to maintain its margin while in fact getting rid of basic transaction-based fees, provided they played the model correctly). tweet
Overall, it's clear that Ariba network fee increases are once again on the table for Ariba customers -- whether they occur in 2012, 2013 or 2014, they are definitely coming. If Ariba can succeed in wrapping new value around the supplier network for its larger transacting suppliers, perhaps these new fees will be accepted without protest, and without price increases on a SKU basis for buyers that would otherwise find themselves paying for them, one way or another, in their larger transacting relationships. tweet
That last point is the one that SAP and Ariba will need to completely flesh out if they opt to exercise network-pricing power over suppliers. Done right, with added benefits for vendors, SAP could come out looking much better than Ariba did after the last price hike, when it faced initial pushback. But pursued incorrectly, with any price increase SAP will open itself up to dozens of lower or fixed-price per document pricing models (not to mention free approaches) performing the same commodity service for cases where SAP/Ariba customers are not locked into their proprietary network.
Where this is all going is simple. SAP and Ariba will need to de-commoditize supplier network connectivity to preserve pricing power to drive, maintain and grow the high margin network/cloud connectivity business. The combination of an expanded network value proposition focused in the areas of supply chain finance/discounting, supplier management (and risk management), advertising/ discovery, supplier benchmarking and related areas could prove critical defining elements to drive differentiation for SAP and Ariba to justify – and even raise – their premium connectivity pricing.
- Jason Busch