Welcome to Marketing BS, where I share a weekly article dismantling a little piece of the Marketing-Industrial Complex — and sometimes I offer simple ideas that actually work.
If you enjoy this article, I invite you to subscribe to Marketing BS — the weekly newsletters feature bonus content, including follow-ups from the previous week, commentary on topical marketing news, and information about unlisted career opportunities.
Thanks for reading and keep it simple,
The World’s Newest Holiday
When you live in Seattle, you encounter lots of people who work for Amazon. Although their employees tend to be tight-lipped, it’s not uncommon for one to mention some secret project launch they can’t talk about, but “you will hear about in the morning.” These announcements are definitely important to the individual, but I have never woken up to anything appearing on the front page of The Wall Street Journal, let alone buried in the business summary news for the day.
Every now and then, though, someone shares an interesting tidbit. Almost by definition, these “over shares” are challenging to verify, but that doesn’t limit their ability to spark questions about Amazon’s motivations for new initiatives. For instance, in the early days of Amazon Fresh, I was told that the REAL purpose of the program was NOT to “master grocery,” but to improve Amazon’s capability to handle returns. How? Well, early users of Amazon Fresh might remember instructions to leave all of the packaging (temperature-controlled bags) on your doorstep, and Amazon would come by and retrieve them. Today, Amazon allows customers to keep or dispose of the bags. As such, if developing strategies for collecting returned items was — as suggested to me by an acquaintance — the original goal of Amazon Fresh, then the program seems to have shifted focus.
In a more topical example, an Amazon employee told me that Prime Day was originally created to test the resilience of Amazon’s systems, bolstering their reliability for the heavy web traffic of Black Friday and Cyber Monday. I heard this story in the context of Amazon crashing last year, just as Prime Day began. A 2018 article from The Verge provided the specific details:
Amazon’s website has been experiencing issues for hours now due to heavy online traffic on Prime Day, the e-commerce giant’s made-up holiday designed to boost sales. According to DownDetector.com, Amazon has been experiencing issues since around 3:05PM ET today, with a large spike in reports immediately after Prime Day officially began at 3PM ET / 12PM PT.
It’s not clear how widespread the outage has been. According to DownDetector, the outage is restricted to the US, and it appears to still be affecting large parts of California and New York as of 5:30PM ET, with a smattering of smaller outages in the Pacific Northwest and other parts of the country.
A corporate post-mortem revealed that the site crashed due to server issues, causing frustrations for customers during a window of approximately two hours. I asked my friend if heads rolled because of the disruption and he replied, “Absolutely not. That’s not how it works at Amazon. We take risks and problems happen. It’s all about how people solve the problems when they happen.” This sounds a bit too Pollyanna, right? Turns out my friend might have been correct — the outage did not seem to negatively affect Amazon’s sales that day. As noted in MarketingDive:
Amazon said in 2018 that Prime Day that year surpassed Cyber Monday, Black Friday and the previous Prime Day. Cumulative sales for the first 12 hours of the 36-hour period were up 89% over last year, according to data from Feedvisor. [Emphasis mine]
If “testing the system” was, in fact, the original reason for running Prime Day, then these results indicate that — like Amazon Fresh and returns — experiments can yield unexpected results, informing new strategies and goals. Prime Day is now bigger than Black Friday and Cyber Monday; if anything, those Thanksgiving events are now Amazon’s preparation for Prime Day.
So, then, what is the purpose of Prime Day?
If you analyze the behavior of Amazon’s competitors, the rationale for Prime Day is clear: “sell a bunch of stuff at a discount.” From the MarketingDive article:
Prime Day tagalongs are becoming as much of a tradition as Amazon's self-started holiday. RetailMeNot estimates 250 retailers will launch sales on Prime Day this year, up from 194 last year and just seven in 2015, the first year of the event.
In particular, Target is riding on Amazon’s wake. Consider the language in the company’s own press release:
“Last year’s Target.com One-Day Sale was one of our biggest days of the year for online sales,” said Mark Tritton, executive vice president and chief merchandising officer, Target. “This year, we’re giving guests more discounts across even more of our assortment with two days to save on hundreds of thousands of items and offering the best options in retail for delivery and pick up on their terms, including same-day.” [Emphasis mine]
Notice that without mentioning “Amazon” or “Prime,” Target describes an event that: (1) is online-only, (2) was one of their biggest sales days last year, and (3) will expand to two days this year. If it walks like a duck…
Why are competitors increasingly offering such blatant clones of Prime Days?
The answer is simple: other companies look at Amazon’s massive sales numbers, and they want a piece of the action for themselves. The recent advertisements for Prime-like events demonstrate that rival companies are emulating what they BELIEVE to be the winning formula: dramatically discount their own merchandise in order to move a high volume of products. Unfortunately, like many external attempts to duplicate successful companies, Amazon’s competitors are copying the superficial elements of the day, while missing the REAL drivers of Prime Day’s success.
Professor Peter Fader is a renowned marketing researcher at the Wharton School of Business, with expertise in analyzing behavioral data to forecast consumers’ activities and companies’ future sales. When I interviewed Fader for my book, he described why substantial discounts on days like Black Friday are usually a terrible business decision:
Black Friday customers are the worst customers. Customers acquired on that day have low LTV [Life Time Value] universally. Most won’t buy again until that time next year. Many other customers who buy that day would have bought anyway at full price later.
This idea is so important that it’s worth repeating: the majority of Black Friday customers fall into one of two (undesirable) categories: (1) people who would have made the purchase at full price, and (2) people who are unlikely to purchase again until the next big sale. If your company offers deep discounts, you leave money on the table when it comes to the first type of customers (those willing to purchase at full price). But it’s the second type of customers — the ones who won’t make a purchase anytime soon — that are the real issue. You might hope those new customers will “trial” your product, returning later to buy at your normal prices; alas, Professor Fader’s data suggests that idea is just wishful thinking. The customers who wait and buy when prices are low? They will continue to wait and buy when prices are low. They are exactly the type of customer that you DON’T want.
Companies receive a big lift from discounting, but that lift usually comes at a dangerous price. In last week’s “Presidential Emails” newsletter, I wrote about the risks of spamming your customers with low-quality emails:
The true, unmeasurable (or at least difficult to measure) consequence [of sending too many emails] is not just the short-term subscription cancellations, but the long-term impact: the reduced trust your subscribers have in you. Every time a subscriber has a “bad experience” from reading your email — or even seeing it in their inbox — it reduces their chances of opening the next email. Which, in turn, reduces subscribers’ respect for your brand, as well as their trust that you (and your emails) will provide them with value. The erosion of trust is obviously difficult to evaluate, but it reveals itself through metrics that you CAN readily observe: lower future open rates, lower future conversions, lower future click-through rates, and lower future brand traffic.
In the same way that every low-quality email reduces customers’ trust in your brand, every deep discount impacts customers’ perception of your pricing. The first time your company runs a discount, you will get a lift, BUT you will also train your customers to believe that your product does not deserve the price premium you charge every other day. Offering a discount won’t destroy your brand, but it CAN cause a small amount of long-term damage. Sales will increase during the discount period, but then drop ever so slightly after the discount ends.
Suppose your company wants to offer another discount; you could expect another lift, but it will probably yield a smaller return than the first time. Results from your subsequent discounts will be lower still. And then what happens? Eventually you face a situation where you can only move high volumes by offering a discount. And when a 10% discount doesn’t move the needle, you increase to 20% — until that level doesn’t achieve the lift you want to see, so then a 30% discount becomes your new normal.
I can’t think of a better example of this cycle than Art.com. An online décor company, Art.com stands out from competitors because they offer far more inventory than anyone else, including options to have items prepared and framed in a limitless combination of sizes and materials. Given the vast scale of the home décor industry, a company that provides high-quality products for every budget and taste should be able to charge a price premium.
Back in 2015, Art.com was recruiting me for their CMO role. After analyzing their situation, the problem was clear: they had forced themselves down the discount trap. What started at 10% discounts moved to 20%, then 30%, and they eventually started resorting to 40% discounts on days when they “wanted a lift.” In our conversations, everyone agreed they needed to escape from their predicament. Unfortunately, when it comes to discounts — just like dieting — saying you want to change and then taking action are two very different things. I finally unsubscribed to their emails in April of last year, but not before screen capturing the last dozen messages I received from the brand:
Think about those subject lines — “Say Hello to 60% Off.” What does that tell the consumer about the value of your brand? Where can you go from there? I would not be shocked if the company is playing with discounts of 70% for “special occasions.”
Is Amazon digging the same hole as Art.com?
In Jeff Bezos’s famous shareholder letter, he explains why Amazon always behaves like it’s Day 1 (filled with vitality and new ideas), instead of acting like Day 2 (mired in stasis and headed for decline). For some time, I thought that discounts might eventually drag Amazon down, at least a little. Similar to the way that Art.com increased the SIZE of their discounts, was Amazon simply increasing the LENGTH of their discounts? I was beginning to wonder if Prime Day Two signalled that Amazon itself was beginning its own “Day 2”? That is, until I started reflecting on what’s REALLY for sale during Prime Day.
Why would Amazon double the length of Prime Day?
As I wrote in the opening paragraphs, the “original” motivation for Prime Day was (at least according to an employee I know) to test the resilience of Amazon’s systems ahead of Black Friday and Cyber Monday. Now that Prime Day has eclipsed those sales events, let’s consider a couple of possible reasons for Amazon to expand their mid-July sale.
Possible reason 1 — In 2018, the servers crashed right after the 3:00PM ET launch. With a two-day version of the event, Amazon could spread the anticipated web traffic over more hours, reducing the strain on their server load. I can’t imagine Amazon decided to extend the length of Prime Day just to avoid overtaxing their servers. Yes, running the sale for two days instead of one might reduce peak loads, but the impact would likely be marginal. Moreover, it would be far easier — and much less expensive — to simply increase the number of servers required to ensure a disruption-free event.
Possible reason 2 — Every year, Amazon sells a lot of “stuff” on Prime Day. By doubling the length of the sale, they could (potentially) sell double the amount of stuff. Perhaps Amazon was worried about potential declines in year-over-year sales, so they planned a two-day event that would be guaranteed to “beat the Prime Day record” for total customer spending.
All things considered, I don’t find either of those ideas — managing server loads and increasing sales — very compelling explanations for why Amazon added an extra day. I think, actually, that we need to consider an even bigger question…
What’s the point of Prime Day?
In other words, what does Amazon have to gain from a summer sale that some cynics describe as a “giant yard sale”?
Probable reason 1 — Awareness might seem like a trivial concern for Amazon; after all, who ISN’T aware of the (sometimes) most valuable company in the world? That said, awareness is a matter of degrees. Although Amazon might be the default shopping option for you and your circle of friends, the company’s share of retail is still only 5% in the United States — and even lower internationally. Many people never even consider Amazon as an option, especially when shopping in categories where the retail experience might matter. By launching a giant “Prime Day” event, Amazon attracts a significant amount of attention, from both media and word-of-mouth. In turn, that attention drives (on the margin) increased awareness and consideration for everyone who hears the message. Prime Day is, essentially, a giant commercial for Amazon, supported by product discounts instead of paid media. Suppose Amazon continues with the annual Prime Day as a way to boost awareness and consideration — then why expand the event for an extra day? Running a two-day version will cost twice as much in discounts, while only drawing marginal increases in free media coverage.
Probable reason 2 — The more I think about it, the more I think the rationale for Amazon Prime Day is right there, hiding in plain sight — the lead sentence of Target’s press release:
Target Corporation (NYSE:TGT) today announced Target Deal Days — its biggest sale of the summer — with no membership required to shop thousands of deals. [Emphasis mine]
Target is not alone in using this tactic. Competitors, perhaps because they don't see any other ways to counter Amazon, have jumped onto the idea that "you don't need membership to shop with us." For instance, check out the clever one-minute video that eBay created to mock Prime Day:
As described by John McCarthy at The Drum:
A human girl named (you guessed it) Alexa mocks the 'parade' of deals on offer during Prime Day (15/16 July) while reminding viewers that it [Prime Day] will just be... “a holiday Amazon totally made up to get people excited about its parade of deals, real parades don't charge a membership. They will mostly have deals on random stuff nobody wants, they are mostly trying to clear their warehouse.” [Emphasis mine]
All of these companies are learning the wrong lessons from Prime Day. Rivals believe that membership fees reduce the value of the discounts, but they are missing the point — for Amazon, the discounts encourage people to sign up for memberships. You need a Prime membership to receive the discounts on Prime Day. So how many new memberships is Amazon selling on Prime Day? We can piece together some details from an article in Forbes:
Last year’s Prime Day was reported by the company to be its biggest day ever for new Prime members, though it did not reveal how many signed up. But to place that in context, Hitwise estimates Amazon signs up 200,000 new Prime members per month.
So, this is my takeaway conclusion: on Prime Day, Amazon is not just selling discounted products, they are offering customers an easy way to make future purchases. Once you commit to a $120/year membership, it greatly simplifies the process of buying the things you need (or just plain want). In the past, consumers faced an annoying decision: “Should I buy this small item on Amazon and pay the delivery fee, or should I try to get my basket to $35 for the free shipping? Maybe I should just make a quick trip to the store?” With a Prime membership, ordering via Amazon becomes second nature; click a few buttons and the item will appear on your doorstep — delivered for free — within two days or less (and soon to be even shorter).
That level of convenience explains why Amazon extended the event to two days. Not so they can sell more “stuff” this week, but so they can sell more — substantially more — stuff this year. Remember Professor Peter Fader’s conclusion that Black Friday shoppers are typically the infrequent, discount-driven customers that companies should avoid? Prime members are the complete opposite. So when Target brags about how their discounts do not require memberships, they are playing directly into Amazon’s hands. Amazon Prime Day buyers evolve into long-term shoppers at Amazon. Target “Discount Day” shoppers, on the other hand, take their deals and run; maybe — depending on the scale of future discounts — they MIGHT return.
Once you understand that Prime Day exists to sell memberships and not Instapots, you can recognize Amazon’s rationale for including sideshow events. From The Wall Street Journal newsletter (with no corresponding web page):
...it's interesting that Amazon.com Inc. is tapping [Taylor] Swift to promote its annual summer shopping extravaganza Prime Day with a concert to be live-streamed July 10 at 9 p.m. EST on Amazon Video… Sure, the concert helps Amazon promote its video streaming service that's an also-ran to Netflix Inc., as well as its music-streaming service that competes with Apple Inc. and Spotify Technology SA.
No one should be surprised by Amazon’s promotion of other services that accompany the Prime membership. Does it matter that some of those features are second class compared to Netflix and Spotify? Not at all. You know what DOES matter? The fact that those services are all rolled into your Prime membership. Join Prime for the discounts, stick around for Taylor Swift and The Man in High Castle. If Amazon can hold your attention on their various services, it’s that much easier for you to make an incremental purchase.
Speaking of purchases, Amazon’s video game streaming service, Twitch, joined the Prime Day fray this year. As explained by Sarah Sluis from AdExchanger:
Twitch streamers will sling products during two 12-hour streams on July 15 and July 16 to drive Amazon Prime Day sales.
The Twitch Presents channel will be taken over by different streaming celebrities selling products ranging from electronics and gaming to kitchenware and kitsch. It’s QVC, but with gamers making the sale.
If you’re not familiar with the video gaming community, many of the popular players are VERY famous — more so than Hollywood stars and pro athletes. By visiting Twitch Presents, you can watch these online celebrities pitch discounted Prime Day products in a modern-day version of a shopping channel marathon. Quite simply, the Twitch Presents channel might provide a way to convince younger consumers to sign up for Prime memberships. If Amazon’s experiment works, you could expect more executions like this next year. Why not have multiple channels available through Amazon Fire TVs, with celebrities from their stable of exclusive TV shows hawking products?
Bottom line: I was wrong to believe that Amazon might have started falling down the discount trap. Amazon may be slipping to Day 2 in some ways, but Prime Day Two is a great Day 1 initiative. Just don’t tell Target.
Keep it simple,
If you enjoyed this post, I encourage you to click the little heart icon below my bio. Thanks!
Edward Nevraumont is a Senior Advisor with Warburg Pincus. The former CMO of General Assembly and A Place for Mom, Edward previously worked at Expedia and McKinsey & Company. For more information, including details about his latest book, check out Marketing BS.