Finding the Voice in Subscription Services

This essay ended up being a bit longer than I expected. This is a partial draft of what I hope will become an academic paper on subscription services operations and metrics. This essay is mostly about metrics, but I tried not to get too technical and boring. I got the idea to write it after reading Hirschman’s Exit, Voice and Loyalty, a treatise on consumer behavior published by the Harvard professor in 1970. I believe widely used subscription services metrics are biased towards Exit. The focus on financial targets and cancelation rates hinders sound decision making. In this essay I challenge this Exit bias and present the case for Voice and Loyalty metrics. I assume subscription services metrics are not common knowledge so this introduction is manly a presentation of widely adopted metrics, based on Gupta et al (2006). If you are familiar with subscription services metrics I invite you to jump directly to the “Shortcomings of Subscription Metrics”.

Subscription services’ managers are always searching for ways to predict and optimize customer flow. Each month, the service needs to create news ways to lure more signups than cancelations. This is not an easy task and it comes as no surprise that subscription services main processes and metrics reflect those flows.

Customer Acquisition: Growth Rate brings in new revenue, but managers do not have a lot of control over it. They can hire market research firms to decide on pricing, they can launch marketing campaigns and they try to create the best service they can think of but at the end of the day they can only hope people will come. If managers are to succeed, they need lots of metrics around Growth Rate. How is pricing affecting it? How can a campaign be more effective? Will a longer trial period do any good?

Customer Retention: Churn Rate is the one managers need to minimize. Even though there are external economic factors that may affect churn, it is much more manageable than growth. There are a lot of second tier metrics that can help a manager predict and avoid churn. How long does a customer remain as a subscriber? Why did a customer leave? What are the signs that a customer may leave? What is the price/quality sensitivity? How do customers respond to retention offers? Most metrics that answer those questions are within reach of managers. They can tweak the product to enhance customers’ experiences or they can monitor customer engagement to predict churn probability.

Acquiring customers has a lot of hit or miss and most tier 2 metrics will be the responsibility of marketing/sales departments. Once customers are in, all efforts should be to retain them. Most other departments (Tech/Products) metrics are about avoiding Exit. Of course, the best-case scenario would be to get churn to zero, but it is probably not financially feasible. Furthermore, as both churn and growth rates are affected by price and quality, these metrics are not independent.

Clearly, not every company wants to balance acquisition and retention at the same point. In some industries, such as low-end, used-car retailing, to take an extreme example, retention strategies have no leverage, because the intrinsic “retainability” of customers is simply too low. In others, the relative importance of retention changes as the industry evolves. - Blattberg & Deighton, 1996

CLV: In order to create boundaries for decisions on maximizing growth and minimizing churn, subscriptions services need financial metrics. One way to do this is to balance customers’ acquisition costs with the expected profits generated by them. Luckily, managers have been using this kind of metric for a while in all sorts of business. It’s called Customer Lifetime Value (CLV).

CLV is the present value of all future profits generated from a customer. One common approach is to assume we know how long a customer will be with a firm and then generate a discounted cash flow for that time period” — Gupta & Lehman, 2003, p. 10.

Blattberg and Deighton (1996) offers another piece of evidence from an interview with a senior executive at McDonald’s: it is “easier to get a current customer to use you more often than it is to get a new customer.”

Customer Expansion: Not every customer generates the same value. Managers should try to convince customers to upgrade to higher margin plans. To do this, both acquisition and retention metrics should come into play. Expansion offers should be desirable enough for the customer to want them but at the same time the current offers should not be degraded or else the service risks losing customers.

Shortcomings of Subscription Metrics

By themselves, those metrics are not enough to manage a business. Growth rate only measures customer influx, a number. Even tier 2 growth metrics will usually classify customer influx by its estimated CLV but this gives little insight on who is the person signing up. For what reason did they choose one service over the other? When did this person decide to subscribe? Is this decision process replicable?

If managers do not have answers for these questions, customer acquisition will forever be delegated to marketing departments because Operations people will not have enough insight in order to adapt the service to be more attractive. Answering those questions will also lead to a better estimated CLV and there are ways to get this kind of insight. Targeted advertising platforms have bulk information on customers interests and they are getting better at pinpointing when exactly a person decided to become a subscriber.

Companies should also be wary of industry reacquisition dynamics. This happens when churned customers from company A are acquired by company B and churned customers from company B are acquired by company A. This type of dynamic may indicate dissatisfaction with all companies in an industry metrics won’t show it. Growth rate and churn rate will be in equilibrium but customers will be trying to find a better service. This type of dynamic may have happened with PayTV until VOD presented itself as an alternative.

Churn rate has a similar issue, as in it only shows the outflux number. It’s tier 2 metrics can get very complex, such as Tech Service Level and Product Quality and all of them respond to the same action: Exit. But should the loss of customers that had been loyal for years be treated the same way as customers that had been around for some months? Maybe those groups have different needs and reasons.

A larger issue is that Exit pressure happens after it is already too late. Instead of focusing on avoiding Exit, services should change their mindset to focus more on rewarding Loyalty. Some Chinese services do it very well. Loyal customers are rewarded with VIP status, early access and prizes. Those loyal customers will be brand ambassadors and are key both to acquire and to retain other customers.

Subscription services metrics are biased towards Exit. CLV dress up as a lifetime metric but in the end of the day it is a way to measure profits before Exit. It does not reflect customers satisfaction during their lifetime. Quite the opposite, the longer unhappy customers remain as subscribers, the better CLV will be. What it does not consider is the negative effect this may have, with brand detractors and high costs for reacquiring customers.

Engagement metrics could be a measure of Loyalty but the Exit bias turns them into an Exit monitor. Engagement reductions are a warning signs for churn. Streaming media can give us a good example of this mindset coming into play. A consultant from a large company that is famous in the VOD sector stated at a conference that if a client remains inactive for 3 months, services should try to engage them because they have high risk of churn. But “there’s causal relationship between inactive customers and churn after six months. The worst thing you can do is talk to those customers.” Does this sound customer centric to you? The Exit addiction forces managers to take actions that are worse for the customers and for themselves on the long run. The only communication heard from customers are their decision to take their money elsewhere.

Changing this Exit mindset is not an easy task and managers will need other tools in order to truly build loyalty. Enter the Voice. Hirschman (1970) gives insight from an economist’s perspectives on the dynamics of Exit, Voice and Loyalty. In perfectly competitive markets there is little differentiation between companies, so any small difference in price or quality will lead customers to Exit a business and choose another. But when the competitive landscape is not perfect, such as an oligopoly, then a Voice effect may come into action. According to Hirschman, customer will use their Voice to express discontents with a company when they believe quality has dropped. The more loyal a customer is, the more he will try to give the company a chance to do better before he decides to Exit. Companies that are Loyalty focused will have more leeway than the ones that aren’t. Loyalty focused companies may bleed less customers when things are not good and will have time to adjust. At the same time, companies that are Exit focused will lose customers to Loyalty focused companies when things go south.

In an age of large data processing capacity, individual targeting and social networks, customers expect to have tailored services, but they also expect high levels of interactivity. Focusing only on Entrances and Exits, where nameless customers flow in and out is a company centered approach. Metrics that treat customers like nameless points in an average are not enough and are not fair to clients. They deserve to have their Voice heard and customer centric competition will do just that.

Case Study: Subscription Video on Demand Services (SVOD)

Launching a new SVOD platform to compete with the big ones is no easy task. Online distribution models have lifted a lot of geographic and technological barriers for video delivery, but new barriers have quickly come into play. Rights for popular appealing shows and movies are finite and usually exclusive, meaning that no two services may hold the same rights at the same time. With big investors pumping money into SVOD services, competition forces licensing prices to go up. Even large TV companies have a hard trying to outbid A-list services. Additionally, if a service is to seriously compete with A-list SVOD services, it will require heavy investments in data intelligence, marketing and original productions that can easily reach $ 15 million per episode. Geographic reach is limitless, but it ends up working in favor of global services. Production cost are sky-high and even wealthy local media giants are increasingly finding themselves having to share rights in order to be able to afford original shows and movies. Catalog size is also limitless but so are cloud costs. And if customers expect an almost infinite catalog, this means a lot of sunk cost on shows that will rarely be watched. Today, high-level SVOD market could only be an oligopoly.

I believe Exit, Voice and Loyalty dynamics are at play when it comes to SVOD. Of course, other competitive forces are at play as well. Netflix has lots of long term debts and they also have early mover advantages. Being a global company, they have scale advantages but they also have network effect advantage. This type of competitive advantage happens when having one client brings more clients that brings more clients. If new clients enjoy the service, they may also convince friends to stay, reducing churn.

Netflix was able to build a following. Their shows are always trending on social media and Netflix is proficient in fueling and steering social media engagement. There are other large SVOD services that do a really good job when it comes to production value and rights acquisition, but no one does public relations — or have such strong network effects — quite like Netflix. Observing Netflix numbers, it’s like they have no tradeoff: huge subscriber base, one of the lowest churn rates, huge engagement and they can also make money. Netflix’s edge is not only on its shows and usability. There are other services — although not a lot — with great shows and usability as well. Netflix’s edge is very reliant on Marketing.

Keep in mind that when I say Marketing I’m not referring to marketing departments. Any large service should be able to create funny campaigns and flood social media with their brand shows. Successful subscription services need to understand marketing as more than that. Marketing activity should be managed as extensions of the product/service sold. This means that Marketing and Operations should become more intertwined, working together both inside the service and outside it. The customer journey is never-ending. It starts outside the service. After the customer subscribes, the journey keeps on going when using the service, trying to reach the right amount of engagement for each customer. It also keeps on going outside the service, listening and interacting with the client’s Voice. And it is never-ending because even if customers unsubscribe, they become targets for reacquisition.

Sounds simple enough, right? It’s nothing new. And yet most companies are not able to do that. In fact, their own metrics won’t allow it to happen, because they are Exit biased and missing the point. Subscription services should foster Loyalty.

Don’t take me wrong, I don’t think Netflix is unreachable. I don’t even think Netflix has fully mastered it yet but it is the industry best at this moment in time. Netflix hears the Voice of its customers and is on its way to get high levels of Loyalty from them — they have a low churn rate for SVOD which is still pretty high when compared to traditional media but they are lowering it each year.

As they should. There are some countries where almost all households have broadband access. And where almost all households with broadband access already subscribe to Netflix. There is no possible further growing, so retaining and expanding customers are the main goals. Soon Netflix will hit this ceiling in many relevant countries and their managers know it. As a one product company, it remains to be seen what they can do create more value through customer expansion. For the time being the only thing they can do is raise prices and hope people stay. Only through Loyalty they can achieve that.

Obviously, this doesn’t mean that Netflix is not using Exit metrics as well. In fact, they probably use all the main Exit metrics and their ramifications. But the company tries to make a point that those metrics are not their main goal whenever possible. For instance, instead of cashing in payments from a customer that has been inactive hoping that they never remember to unsubscribe, Netflix automatically cancels an inactive account. This is simply a sign that the company does not want to play clients for fools. It may cause a little revenue loss, but Netflix knows it has more to gain in reputation.

Netflix’s Edge

Netflix has one of the best social media strategies of any company. What to others is just a marketing activity delegated to interns allows it to be constantly monitoring the Voice of customers. Netflix seems to be capable of monitoring networks and sorting what is just a complaint from what is a movement and what is just a suggestion from what is a demand or threat to Exit. And Netflix excels in responding to clients. They seen to have some well thought processes that their social media workers have to follow, because they can respond in all social media platforms and in many different languages while appearing to be the same person.

Every company has fluctuations in quality, this is inevitable even for Netflix. This is not a simple subject, because in the media industry higher quality for one person can be seen as lower quality for another person. In other words, if Netflix develop a companywide project to improve quality it could, paradoxically, be harmful to the service, losing subscribers and receiving complaints. Hirschman argues that, all else constant, a company will gravitate towards an offer that has an average quality, and it looks like Netflix is ​​at this point. Every customer should feel that Netflix has some wonderful productions and also some crap. The magic lies in some set of customers loving what other sets of customers considers low quality media.

Of course, some customers will always be price sensitive while others are quality sensitive. Hirschman calls those quality sensitive share of customers connoisseurs. These loyal customers accept price increases, as long as quality is maintained in their opinion. And they also defend the company’s brand. They attract new subscribers and help retain existing ones. Netflix encourages this behavior by creating exclusive content for social media that fuels conversations. The network effect of generating conversations brings loyalty. In this context, Loyalty means that there is moral suffering if a customer choose to leave the service. One example is the fear of missing out.

On the other hand, connoisseurs punish drops in quality. Netflix has managed to show that it listens to people’s demands. If a customer expects to be heard, it has a higher probability of choosing to use Voice pressure instead of Exit pressure. A loyal dissatisfied customer will complain before canceling. Netflix has already pushed several price increases only to watch its subscriber base grow even more.

Every set of customers in a service will have more activist members and quieter members. This is a good thing, as these silent customers form a slack capacity of Voice. The company knows that activist clients will help marketing campaigns and will give day to day feedback. But the company should also know that slack customers will only use their Voice when their situation gets truly uncomfortable. I believe that Netflix uses social media data and service use data to plan their responses to user’s Voices. Netflix appears to pinpoint with high level of accuracy the moment when the common demands of its most active users start to get echoed by their quietest users. Netflix can identify demands and can act quickly. Even when the company is not able to satisfactorily solve the demands in the short term, it shows customers that the service is trying its best to meet their demands. This is another opportunity that Netflix uses to demonstrate that it is listening to the Voice. If customers have satisfying past experiences with the company, it may be enough to maintain their loyalty.

In order to achieve Loyalty, Netflix has to show their word is worthy. Netflix strives to demonstrate its concern for customers. Many companies have a genuine concern for their customers, but concern must be shown, verbalized and perceived. Canceling inactive subscriptions, for example, is a good demonstration of thoughtfulness. It shows intention and it shows that the company will go out of its way and even harm itself for the customer. It was also a smart competitive move. From a client’s perspective every other competitor that does the same is just copying Netflix and would be cashing on incautious clients if it were not for Netflix. Interaction is also important to generate value for a company’s word. In a two-way relationship, the customer will be more willing to continue paying and wait for the company to solve issues.

Netflix knows that retaining a customer is cheaper than reacquiring one. A customer who enjoys the service and left for financial reasons is cheaper to be reacquired than a new customer. But loyal customers who left because they felt betrayed are the most difficult to win back. They will need a demonstration of quality above the level when they left. If they ever return at all.

A Matter of Loyalty

Notice that to succeed, Netflix probably needs metrics that are different from standard subscription services metrics.

Exit and Entrance flows are not enough. Lost customers loyalty levels correlate with the chance of reacquiring them. Each customer Loyalty should be considered. Companies should work to determine what is “quality” for each customer.

Income is not the only value created by customers. Customer Lifetime Value should take into account Loyalty and network effects such as the ability to influence other customers. Losing a connoisseur is a far worse sign for a business than losing a customer who usually signs up just to use a little bit of the service and bounce.

Engagement, which informs Exit metrics, should also be analyzed from the perspective of Loyalty. Different groups of people have different patterns of engagement, which is not to say that just by being less engaged there is higher chance of leaving. Data intelligence must allow services to be able to sort these groups and understand true risk of Exit. Companies that actively listen to customer’s Voices learn how to differentiate real discontent in each group from what is just a hiatus of engagement.

Subscription services should develop Voice metrics as well. Any metric used must consider different needs from the most energetic people and the Voice slacks. Dynamically identifying variations in this slack and responding quickly to changes in each group of consumers is an ability that is difficult to replicate and creates competitive advatage.

A Matter of Focus

Netflix has only one offer for clients. This helps the service avoid the trap of having customers that are much more financially valuable than others. Services that combine many types of plans with access to different services are likely to face more difficulties when they need to make decisions and prioritize customers’ Voices. Most likely, a free customer will be less heard than a paying customer. And the more one pays, the more prestigious is the Voice. It is a trap that leads managers to justify mediocre service levels. I believe this was the reasoning behind Roku’s spinoff from Netflix. It is hard to tell if the deal would have gone any differently today. I believe that expanding out of their focus is the main challenge and a litmus test for Netflix. In countries where Netflix has reached the limit of subscribers, other services are also gaining popularity. Netflix will have to expand its operations if it is to continue growing.

Another important matter of focus is that Netflix subscriptions are not part of a bundle, unlike Amazon Prime Video. Neither does Netflix have any other major source of income, unlike HBO Max. Netflix results are heavily reliant on management’s ability to provide a good service. There is no way to mask bad results and big mistakes could mean the end of the company.

On the other hand, a pitfall that can hide inefficiency is that streaming platforms today are sensitive to new customers or revenue but not profit. Even Netflix, despite making a profit, does not rely heavily in this metric. Clearly Netflix cares about profit and has been managing to increase its financial success but the tradeoff between profit vs service level is a balance that must always be reassessed. Spencer Neumann, Netflix CFO, made it crystal clear during Q4 2020 earnings call:

So we’re going to continue to invest aggressively into the growth opportunities that we see. And that’s always going to come first. But beyond that, if we have excess cash, we’ll return it to shareholders through a share buyback program.

Netflix is the best in class, but they still have a long way to go in their journey to excellence. They are yet to find success with interactive content in their service. And, unlike Chinese services, Netflix does not reward Loyalty inside their platform. There are not many ways to be social either. And of course, as Netflix reaches subscribers quantity ceilings, they will need to offer more value for the service. Rewarding loyalty appears to be a low hanging fruit. Building successful interaction in-platform is a lot trickier. I am curious to know how Netflix’s ability to listen to customers’ Voices will allow the company to disrupt other branches of the media industry. But one thing is certain, Netflix will keep trying to develop new functionalities and they can count on their customers to help them in their journey. I hope that more companies will be able to overcome their Exit-bias and do the same.

Sources:

BLATTBERG, R. & DEIGHTON, J. Manage Marketing by the Customer Equity Test. Harvard Business Review. July-August 1996.

GUPTA, S. &LEHMANN, D. Customers as Assets. Journal of Interactive Marketing. Vol. 17, n. 1, pp. 9–24. Winter 2003.

GUPTA, S. et al. Modeling Customer Lifetime Value. Journal of Service Research. Vol. 9, n. 2, pp. 139–155, November 2006.

HIRSCHMAN, A. Exit, Voice and Loyalty, 1970.