We have all been captivated in following data usage--current and future, per subscriber, per device, per technology, per country, and so on. A related metric that does not get quite as much attention--and yet deserves it--is demand.
Usage is a single-number metric that can be defined in many ways but that can be accurately measured for each subscriber, device, or network. Some operators make these data public, and others don't, but they all have the information internally.
Demand cannot not be measured as a single metric as it varies as a function of other variables, such as pricing, ease-of-use, and perceived value of service, or network speed, coverage and availability. As a result, depending on these variables, demand and usage differ, with usage being a point in the demand curve. For instance, pricing is going to affect usage: the more expensive the service it is the lower the usage. Figure 1, below, shows this type of relationship, but the shape of the curve will depend on specific markets--and eventually on individual subscribers, as their elasticities differ. For instance, the curve does not have to decrease with the increase in price for individual subscribers. In a market where prices are high, most users may not be able to afford mobile data at all, and hence only heavy-users, with low price sensitivity or a great need for data connectivity, subscribe to the service, thus raising the average per-subscriber usage as prices go up.
Figure 1: Usage and demand as a function of price. Source: Senza Fili
Generally though, we can assume that if prices were lower, we would see usage growing. Unlimited plans give us a useful glimpse in what usage would be if pricing were not an issue. In the United States, data usage among Sprint Android subscribers is about 50 percent more than for than other operators with most subscribers on metered data plans. Does this mean that if the other operators were to all switch to unlimited data plans, they would see a 50 percent increase in traffic? Most likely not, because heavy users are more likely to choose unlimited plans.
The relationship between demand and pricing is well understood in all sectors, and telecoms is no exception. But there are other variables that affect demand which are specific to wireless data and that may have a powerful--although still rarely leveraged--impact on revenue generation. For instance, data rates exert a strict control over usage. Assuming that the time we have in a day to actively use our smartphones is limited, the amount of traffic we can generate within that time is constrained by data rates. The faster the network, the more you get done, the more video you can watch, and the less time you spend waiting to get the information you need--and the fewer times you give up in frustration. On the Vodafone network, for instance, LTE smartphones generated twice as much traffic as 3G smartphones in March 2013. In part, this is due to the fact that early adopters tend to be heavier users, but a similar effect has been noted when subscribers move from 3G to Wi-Fi as well (although in that case part of the effect has to be ascribed to the fact that Wi-Fi is typically free or cheaper).
The underlying demand may also affect usage depending on how services are structured and priced. In the figure above, we plotted per-GB price. But other ways to charge for service--e.g., by service, such as Pandora or Spotify subscriptions--may increase usage more than the per-GB curve may predict. The ability to use multiple devices in the same plan may encourage migration from fixed to mobile data access (e.g., by increasing the number of tablets with cellular connectivity) and hence increase the mobile traffic per subscriber.
Maximizing traffic however is not necessarily a good thing. If the new incremental traffic becomes too expensive to carry, mobile operators are better off keeping traffic down to more profitable levels. This brings us to the reasons why we should care about demand, and keep it in mind when looking at usage data. Usage levels are often taken to be a reflection of our needs and desire to use to use mobile data, but in fact they are bound by multiple factors--many of which operators can and do manipulate in following their strategy.
Recently, many operators--notably in Europe and in the United States--have started to see a decline in the data growth rate. Per-user traffic keeps increasing, but not as fast as it used to--and the trends are different across markets (see Figure 2, below). Is this a sign of smartphone fatigue from subscribers? This may be a factor for some subscribers, but the change may also be driven by the different demographics of smartphone users (no longer a minority of always-on, highly mobile early adopters), as well as stricter and more frequently enforced data caps from operators.
Figure 2: Usage evolution in the Vodafone network in selected European countries. Source: Vodafone, Senza Fili.
Understanding the demand elasticity that hides behind the usage figures is crucial to continue to refine the forecasts about data usage (for instance, Cisco revised the last VNI forecast downwards presumably on the basis of revised analysis of the demand for mobile data in the current market environment), but also to design new services that can profitably address the unmet demand (and phase out those that don't). More specifically, however, the biggest rewards--and opportunity for differentiation among operators--may be obtained by venturing beyond price elasticity to explore how other factors--e.g., network speed and availability, network congestion, battery life, fear of quotas, device form factor, or content availability--have in different markets and subscriber segments.
Monica Paolini, PhD, is the founder and president of Senza Fili Consultingand can be contacted at [email protected]. Senza Fili Consulting is an analyst and consulting firm that provides advisory services on wireless data technologies and services.