The Case Against Consumption Based Pricing

Imagine if TV companies charged consumers by the minute to watch a football match.  Or if your office rent increased when you had a particularly busy month.   Sounds preposterous, doesn’t it?  Yet, this is exactly how many media management systems providers are currently pricing their services, and no-one is batting an eye, indeed, it is a model that is being extolled by many. But examination highlights that consumption-based models for management systems are really only beneficial to smaller operators.

The media and entertainment industry was slow to embrace the Software as a Service approach introduced by companies like Salesforce over two decades ago. The last few years has, though, seen a massive uptick in the adoption of cloud services and many media operators are now enjoying the benefits of easy access, instant scalability and the financial flexibility that make SaaS applications so appealing to many other business sectors.  However, with this new way of working comes a whole new approach to cost and, for the media industry in particular, this aspect of cloud adoption has not been without its headaches.

Specifically, anecdotes of ‘bill-shock’ abound as content producers struggle with the apparently contrived complexity of costs associated with downloading or exporting their content from the cloud and many articles have been written (including some of our own) on the infamous egress fees.  But, perhaps the media industry’s preoccupation with egregious egress costs has distracted us from other issues with some cloud-based pricing models which may not be as beneficial as they first appear.

Fees for transcode services and content egress are based on a consumption pricing model which Amazon Web Services (AWS) are credited with pioneering.  It’s not really a new concept of course, we’ve been paying for consumables such as gas and electricity in this way for decades.  In that context, the consumption model for media services is very clear, and inherently fair.

What is more difficult to reconcile, is the application of the consumption model to management systems that simply orchestrate 3^rd party services. For example, some cloud media management providers charge a standing fee for access to their platform and, on top of this, add a variable charge based on the volume of content under management or the 3^rd party services utilised. These management costs are in addition to the cost of the actual consumable service and are categorised as throughput overhead.  Some providers ‘bundle’ a storage or a throughput allowance and charge an ‘overage’ rate if it’s exceeded, a shameful oxymoron in context of cloud.

This model works for smaller operators – in fact it’s seen as the perfect pricing structure for SMEs because it affords access to enterprise class management software, on a pay per use basis, with a very low cost of entry.  Yes, the base costs are a little elevated over the underlying services and each transaction is a little more expensive, but this orchestration tax is worth paying for the operational benefit and convenience.

But, for anything but the smallest operators with only a few TB of content, this orchestration tax becomes noticeable, if not burdensome.  Providers are quick to justify the model by pegging cost to success – the more content a producer has under management, the more work they are doing, the more successful they must be and, ergo, the more revenue there is to spend on services.  This creates a very linear relationship between busyness and cost – which is just about justifiable for video processing services were work is proportional, but that is not the case with management systems.

Like online accounting software tools, media operators expect to pay for management systems based on features, or even user ‘seats’ but will balk if pricing is per transaction.  Moving production operations to the cloud affords limitless scale but that should not limit the economy of scale and media operators should see the cost per transaction flatten out and fall at higher throughput volumes.

In recent years there have been many published comparisons between taking a CAPEX (hardware and perpetual licences) and an OPEX (cloud service) approach to production technology procurement.  In most cases authors tend to present a clear winner one way or the other but the reality is more nuanced and comes down to the business’s operating model.  The significant variations between the different types of SaaS model on offer, and specifically how these suit different business requirements, play an important role here and deserve more scrutiny.

We are great proponents of cloud and SaaS and, in our own procurement of business systems and infrastructure, the model has allowed our business to grow rapidly without taking on burdensome debt or suffering the distraction of equity investment.  But our success has been based on careful assessment of the different SaaS pricing models on offer and how these align with our specific business requirements.  Our advice is that medium and larger scale businesses, in particular, should engage directly with underlying service providers and factor in future growth to identify which model is right for your business.

The acid test here is that operational costs should fall with, not track, your business growth – and if the model on offer looks complicated, look out for the devil in that detail.

This article first appeared in the IABM Journal (119 Page 4) https://theiabm.org/journal-119/