Promotions are everywhere, from the ubiquitous "SALE" signs at furniture stores to the freemium models that are so popular with apps these days. The trouble with promotions is that marketing teams often misunderstand and then overuse them.
Generally speaking, there are three good reasons to do a promotion and a whole host of bad reasons to do one. But before we go there, let's remind ourselves of what the whole point of pricing is. No, it's not to make money. Well, it is, but that's not the right answer here. The point of pricing is to charge customers what they are willing to pay at any given moment. Of course, we have to balance that goal with how difficult it is to actually achieve in practice. In theory, we'd want to negotiate with every single customer. The problem is that this is not a scalable method of selling things. If you are a fruit vendor in a marketplace, you might be able to custom-quote prices for your wares. Now imagine you're selling apples online. You have to set a price and wait as people fill their own carts and check out. So, while we'd love to negotiate, we simply don't have the time. This brings me to the main reason to do a promotion: promotions are used when your existing price model breaks for a certain segment, and you need a special lower price.
The first good use case for a promotion is to add flexibility to your price model to account for customers' varying willingness to pay. The most common version of this is discounts. Salespeople are often given the ability to offer discretionary discounts when negotiating a deal; the theory (and it's often correct) is that the salesperson has the best sense of this particular customer's true willingness to pay. By allowing some leeway in the price, salespeople can customize the price for that customer.
In B2C, this often shows up as coupons. Coupons, especially targeted coupons (over email, for example), are a way to look at a customer and say, "Hmm, you seemed to want this thing. You put it in your cart and everything. You even came back to it a day later. Maybe you just need a discount to get you to buy." Both of these methods account for the fact that the "best" price for a customer segment is really just an average. Allowing salespeople or marketers to give discounts can sometimes compensate for the error that's inherent in that average. There is an entire field called "Sales Operations" that deals with discounting guidelines, among many other things. I'm not going to go into best practices by giving sales and marketing power over setting the price. Let's just say that you should have some discounting guidelines because otherwise, salespeople will go wild.
The second reason to promote has to do with offloading inventory. Imagine you're selling the hottest new iPhone at $1,000. The first people to buy it are the ones who really need it – maybe their iPhone just broke, maybe they love the new camera features, or maybe they're a VC who needs to show off to new founders. Whatever the reason, they have the highest willingness-to-pay. As the phone is out in the market, fewer and fewer people buy it, and as the phone gets to be a year or two old, all the people with a high willingness-to-pay have already purchased it. If Apple still has a bunch of phones left, they'll start lowering the price. After all, it's better to make some money on the extra inventory than throw it away. When the 256GB iPhone 12 came out, it was initially priced at almost $1,000, but by the following year, the price had fallen to $750. By that point, the risk of cannibalization was low, and Apple could safely lower the price and sell to people who don't have as high willingness-to-pay. This strategy is important, but it is unique to companies with an inventory. Why doesn't this work with software or media? Because I can't lower the price of my software without lowering it for everyone. The cannibalization risk is too high. The only time where this works is when you have "versions" of software. For example, Microsoft often dropped the price of Office once the newer version came out. But now that Office is a subscription, they won't do that anymore.
The third reason to promote is my favorite, and it has to do with a fancy economics term called information asymmetry. Information asymmetry, in this case, means that a buyer, before trying a product, doesn't know whether or not the product is actually going to work for them. If I'm buying pizza from a new restaurant for the first time, I might not like it. If I'm buying a new streaming service, I don't know if the shows are any good. If I'm buying some new business software, I don't know if it will have the features I need. Even if you tell me that your pizza is delicious, your shows are entertaining, and your software is useful, I kind of need to experience it myself as a buyer. Put another way, there is risk in my purchase. And when there is a risk, it means I have a lower willingness-to-pay than someone who has had the pizza before or used the software. And this, my friends, is where promotions truly shine.
Promotions help new customers overcome the initial risk of purchasing a product by offering them a chance to try it before making a commitment or by reducing the risk associated with their purchase. Generally, I categorize promotions into four types: freemium, free trial, deep discount, and intro bundle:
Freemium models are extremely popular in Silicon Valley, and honestly, I think they are overprescribed. Freemium models work well when the realized value of your product increases gradually over time. When Dropbox first started, it was the perfect freemium. When you first started using Dropbox, it had zero value. The box was empty after all. Over time, users put more and more into their boxes, and the value grew. All Dropbox needed to do was put a reasonable limit on space, knowing that most users would eventually hit it as long as they got more value out of Dropbox. If, for whatever reason, they stopped putting things into Dropbox, then they weren't getting much value out of it, and the user never converted. How does freemium go wrong? In two ways: either people assume value grows over time when in reality it stays put, or people choose the wrong fence. Both happened at one of Dropbox's competitors, Evernote.
In around 2014, both Dropbox and Evernote were wildly successful unicorn companies. They were founded at around the same time, and both had fancy offices with kombucha on tap. They also had charismatic CEOs who spoke at tech conferences. However, while Dropbox's freemium model was working, Evernote's was floundering. Why? The first reason had to do with the use case. Dropbox was a literal box in the cloud. People used it to store stuff. End of discussion. Evernote, on the other hand, was more like a notebook: sure, you could use it to store things, but you could also use it for cloud-based to-do lists or note-taking. Evernote made two crucial and irreversible mistakes in the early days of the company. The first mistake was assuming that as people used Evernote, its value would grow over time. For some people, it did: some used Evernote as a collection of recipes, class notes, and business cards. However, many people still use it as a cloud-based to-do list. They had a set of notes that they constantly changed or edited, but the total size of the box never grew. In other words, Evernote was an "organizational system" rather than just a box, and it had 100% of its value from the first day the user started using it.
The second mistake they made was picking the wrong fence. While Dropbox fenced their free plan by the amount of storage included for free, Evernote fenced theirs by the number of uploads per month. If both products were a literal box, Dropbox charged by the size of the box while Evernote charged by the size of the opening to the box. For users who were using Evernote as a to-do list, or even a note-taking app, they never bumped up against the fence and never converted.
- Free Trial
The second promotion option is a free trial, which needs no real explanation. Free trials should be used in the opposite situation as freemium, when 100% of the value is realized on day 1. Netflix is the perfect free trial. The value of Netflix is access to its library of shows. It’s not features. It’s certainly not usage. It’s access. So how do you know if the access is worth it? You give someone free access for a limited time. Most media companies offer free trials: Netflix, Disney+, ESPN, and Headspace. Even certain parts of the New York Times are a free trial, although most of the New York Times could be broadly classified as “freemium”. A notable exception is Spotify, which is a freemium. Let’s ask the question: should Spotify be a free trial? Probably. The value to Spotify again is access, not usage. While Spotify may have gotten that wrong, their competitors that followed (like Apple Music, Amazon Music, and Tidal) are all free trials. If I were a betting man, I’d say that Spotify would have been better able to monetize their product if they had been a free trial, but now it is such a core part of their business that they could never take it away from users.
Both the Spotify and Evernote examples illustrate that the decision to be freemium or free trial is often irreversible. In Spotify's case, making the product fully paid for would be a PR nightmare. In Evernote's case, the engineering required to switch the fence to storage would have been immense. There are very few companies that have been able to make that switch; the New York Times is one of them. In the early 2000s, media companies were hammered by low-cost online distribution. The New York Times instituted its 3 articles per month limit in response. While the NYT has been far more successful in monetizing their offering than their competitors, I don't think that the 3-article paywall actually works for them. First off, the paywall is too easily avoidable with incognito browsers. Second, the Times has actually started adding sections to their online presence that are only accessed via free trial: travel, puzzles, and cooking come to mind.
- Deep Discount
The third promotion is a deep discount, and my example here is Verizon. Telecom providers often offer ridiculously low introductory pricing, like 50% off plus a new iPhone if you're switching. There are lots of reasons why they do this, but one is that switching cell phone providers is extremely annoying. You might get a new number, in which case you have to notify all your contacts. It takes time and effort to set it up. And only after all that pain and effort do you see the benefits of presumably increased service levels. Verizon is trying to match this negative or delayed value by significantly lowering the price in those early months. The opposite of this would be an "implementation fee". Imagine if rather than giving you 50% off for the first year, it was double the price and then they charged you a one-time registration fee! That would be preposterous! And yet, so many SaaS companies do exactly that. You may ask, "But I have to cover my costs!" No, you don't. That's cost-plus pricing! Trust me, Verizon does not cover their costs when they give you a free iPhone…
- Intro Bundle
The final promotion style is what I call an intro bundle, and it’s the rarest. Intro bundles work when you quickly realize the value of the product, but then it peters out over time. My example of this is Comcast. When you buy a cable package, the company will often try to upsell you at the point of sale. The theory behind this is that if you buy more items at the beginning of your contract, you’ll get more value out of the product and be more likely to stick with it later on. Gyms will often throw in a free personal training session when you join. Software companies will give you all the enterprise features for the first month. And any company that pays extra special attention to the onboarding process is doing an intro bundle, even if they don’t call it that.
What should you be?
Freemium, free trial, deep discount, or intro bundle? It all depends on how your value accrues to customers over time. Slowly? Freemium. On day 1? Free trial. Delayed or negative? Deep discount. Quickly? Intro bundle.
If I choose free trial, how long should it be?
7 days? A week? A month? I have two answers for you. Answer number 1 is, I have no idea - you should probably test different durations and see what works. Note that promos are way easier to test than prices and infinitely easier to test than packages, metrics, and structures. Answer number 2 is to choose a duration that aligns with how long it takes to derive value. If you're a consumer app, maybe you will understand it in 7 days. For enterprise software, you might need a full quarter.
What are the promotions I don’t like?
Most of them. We've discussed how much I detest promotions that aim to increase sales volume since they seldom yield results. My least favorite promotions are the "always on sale" promotions that appear to be quite popular in furniture stores. I am not fond of such promotions because they rely solely on short-term psychological tactics. If you ever find yourself saying "I'll wait until it goes on sale," that indicates a company that has conditioned its customers to wait for promotions, which significantly impacts profit margins.
When should you use promotions?
Option 1: you need to customize your price to customers' varying willingness-to-pay, e.g. salesperson discounts or marketing promos. I'd put "win-back" campaigns, where you send discounts to churned customers, in that category too.
Option 2: you sell hard goods and need to offload some inventory.
Option 3 (my favorite): you need to de-risk the purchase for a customer in the form of freemium, free trial, deep discount, or an intro bundle. Don't put a sale sign outside your door every day of the year.
Ian Clark has over a decade of experience helping software and internet companies monetize their products. His clients include companies such as Y Combinator, LinkedIn, Eventbrite, and Cloudflare. He is the author of the Monetization Guide.