One of the most fascinating, yet relatively unchartered areas of cloud computing’s all-out assault on today’s enterprise is the complex beast that is the consumption-based pricing model.
Technical, operational and of course, security issues aside, it is easy to see how the advent of pay-for-what-I-use, which is commonplace in pretty much all of the leading Public Cloud IaaS services, coupled with a shift from capex to opex spend is extremely attractive to enterprise IT leaders who are looking for new and innovative ways of aligning their IT costs to actual usage. In fact, I would be surprised to hear from any enterprise IT leader who wasn’t sick and tired of shelling out on huge investments in infrastructure and application portfolios, along with their equally steep yearly operating costs, only for some variable percentage of that solution to sit dormant, idle, under-utilized or in some cases, just collecting dust as plain old shelfware.
There has been plenty of excellent discussion recently around whether a growing maturity and accelerated adoption of cloud computing services will have massive shrinking effects on IT budgets. It’s an interesting topic and one to which I am sure you have drawn your own conclusion (as have I) so I will not labor those points, nor wake the learned Mr Jevons again from his well deserved eternal slumber.
But..something doesn’t feel quite right.
For me, the promise of this new financial flexibility in the Cloud, be it IaaS or PaaS, sits on my shoulder like an ethereal vision of an angel, beckoning me ever closer. On my opposite side, there sits a cunning devil, one who promises to unshackle me from the misery of traditional enterprise software, but one who has a very dark and sinister side. His name is SaaS. To the untrained eye, he’s the nicest of devils, quite charming and approachable, but he has two major flaws – he has an identity crisis and he doesn’t understand scale.
Take a look at many of the popular SaaS offerings that are out there today and you will quickly spot a recurring model, sometimes known by the portmanteau word “Freemium”, and often categorized across the “Personal > Business > Enterprise” silos. It may go something like this:
Sign up for FREE ! $12 Per User Per Month ! Call For Pricing !
These omnipresent models are absolutely fantastic if you are a one-man-band, a small organization or have a real niche need for a SaaS-based service within a large organization. It’s the last three words that cause me to wobble. Call For Pricing. I’d rather not. Honestly.
In so many ways, SaaS providers are the antithesis of traditional software vendors. Their shorter product cycles, hassle-free operation, intuitive UIs and UXs, cross-device and cross-platform support and automatic upgrades not only make them attractive propositions, but almost indistinguishable from their ancestors. However, it appears that the DNA runs deep and and leads to an alarming inability for some of these newer players to transcend the “small organization to enterprise” divide whilst understanding that “enterprise doesn’t mean committing to 5,000+ licenses on day one just to get to a lower cost per user in aggregate”.
Here’s a quick example.
Take one large organization, say 15,000 people. Take one “cool” SaaS product, say cloud storage. Take one cost structure of $12 per user per month.
Do a simple linear equation, without calling for pricing. 15,000 * $12 * 12 = $2,160,000 per year.
That’s a lot of money for a simple service that does pretty much one thing. Of course, it assumes that all 15,000 people are active users, daily, but nobody of sound mind would ever pay that much per user for so many users. So, what are the options ? Oh yes, call for pricing.
My guess is that if I did call for pricing, I would be able to negotiate a better aggregate cost per unit, but to do so, I would likely have to commit to a volume of users that I can not guarantee would be using the service on day one (or at all). So now I am not paying for what I use, I am back to a hedging scenario – a lower unit price, but a higher overall commitment and once again, my costs are out of direct alignment with my usage. It is no longer consumption based, as my unit of consumption measurement (my users) is more than are actually using the service. Hmm.
The challenge in all of this is to find a fair model that is palatable for both the customer and the provider. A model that allows the customer to grow into scaled-up actual usage of a service at a price point that is acceptable, representative of the service’s value to the organization and does not dis-incentivize adoption. Sadly, I have no magic formula, but the default position of enterprises having to call for pricing, spend a lot of time and effort to negotiate and then potentially walk away from attractive solutions because the mechanics of the deal “feel too much like the olden days” is something that is causes me to wonder whether this is a phenomenon that is harming enterprise adoption of certain “cool” SaaS products much more seriously than the triumvirate of bogeymen named Security, Compliance and Trust ?
Tweets that mention New Blog Post : "Call For Pricing" - On all things concerning with consumption based pricing. #cloud -- Topsy.com
Feb 24, 2011 @ 23:58:00
Feb 25, 2011 @ 13:59:10
Interesting post. I focus a lot on pricing so that I don’t have to do a lot of work on pricing. We have set prices and publish them on our website. I don’t want a lot of “Call for Pricing” calls nor do i want to track a lot of special deals (a la Oracle – only buy on the last day of the quarter). For us, if we need to lower a price to get a deal it should be indicative that our prices are above market for the value we provide and we need to increase value or lower prices.
We also essentially offer flexible pricing. You can start with the minimum of 10 seats and purchase more over time. Also, it is seat-based so you can rotate users in and out (which you can do with hardware tokens – if you get the token back). Most buyers have to “talk’ to us to understand this. We hope to build it into the product soon.
We also don’t have a lot of marketing around this concept. I’ve been chewing on it for an offering for cloud-based services. You want to add two-factor authentication to your cloud-based app scanner, well, start with 10 WiKID seats and pay as you grow.
great post, love this type of information.
Feb 25, 2011 @ 16:32:53
For pricing on a basic service like cloud storage, shouldn’t the model be simple and straightforward based on actual usage, rather than the number of users? Storage costs XX cents per GB per month or year independent of the number of users. I could see pricing for other services to be seat based, but storage seems simple enough that actual usage would drive cost rather than the volume of users.
Feb 25, 2011 @ 16:37:13
That is part of the point. If you look around at the options, they are based on either an amount of storage or an amount of storage per user count, which the vendors assume is the best they can figure out without “call for pricing”. My point is, this doesn’t work.
Feb 25, 2011 @ 21:50:32
This is somewhat beside your point — which I like — but as someone who used to work for a vendor who had a hugely successful consumer storage-as-a-service, I can sympathise somewhat with your vendor.
The StaaS market has been hopelessly compromised by “fail to scale” price-per-user models. Some of the earliest vendors in this space decided they could differentiate and make their service easier for consumers if they said “hey, just pay us a flat fee per user”. They know full well that this doesn’t scale if users start to pile on the data, but they’re relying on the fact that their piddly DSL and Cable connections stop them from uploading significant volume.
The folks who tried to do “pay for what you use” got squeezed as a more complex, negative-feedback-loop pricing scheme (the more you use the service, the more it hurts).
Then everyone got addicted to the breakage inherent in per-user tiered pricing (e.g. 50Gb for $x) — have folks pay for storage they don’t use. And here we are…