The Digital Marketing Flywheel Of Doom

Inspired by some of the recent news on the tradeoffs between brand and performance marketing, I want to explain the mechanics of how digital marketing “best practices” have made it almost impossible to sell a physical product online without some type of discount.

Enter: The Digital Marketing Flywheel Of Doom.

Note — charts below are stylized because the data is for illustrative purposes only, and doesn’t correspond to any individual business.

Budget Setting:

At a high level, the breakdown of costs for an average unit of apparel sold online looks like this (assuming the business needs to be profitable):

The underlying assumption here is that every $1 of digital marketing spend will generate $8–10 in sales. There is no digital marketing channel that drives a return like that consistently, so we have to assume that some sales are coming from elsewhere — either from other marketing activities, overall brand awareness or luck.

This works out to a digital/performance marketing budget that is between 5 -20% of projected annual sales. So if I need to make $10 million dollars next year online, I will have between $500K — $2M to work with.

Budget Allocating:

Every business will have peak times when sales are higher. This could be the holiday season, or a time of year when people use the product more often (ski gear in the winter, BBQ grills in the summer, etc.). In fashion retail, no one strictly needs anything being sold, so many business’ high volume times are going to be driven by sales or promotions.

Let’s say the ecommerce P&L owner is tasked with growing sales 10% higher than last year. A very straightforward strategy would be increasing sales evenly by 10% across each month in the year, and that’s what we’ll look at first for simplicity’s sake.

Once we have a sales target for each month, we can then allocate the year’s digital marketing budget across each month. Again, we’ll keep it straightforward to start and and say that each individual month’s digital marketing budget will be 10% of that month’s projected sales.

Although this is a simplified example, in a lot of companies this is exactly what happens. From a financial standpoint, it is hard to say “we’re going to spend $1 of every $10 earned during a peak sales month, and $1 of every $2 earned in a slower sales month”.

This is where the trouble starts.

Measuring Performance:

Digital marketing investment always produces a higher return when a brand is running some kind of sale or promotion.

The inputs you can control in a digital campaign are:

  • Who you are targeting
  • The campaign creative
  • The on-site experience
  • What part of the product assortment you feature

When a brand is on sale:

  • The product becomes affordable to a larger percentage of the people you’re targeting
  • More people click through the campaign creative, because in addition to finding the product interesting, you’re also attracting deal-hunters.
  • People are more willing to put up with a sub-par site experience to get those sweet deals
  • Broader swaths of the assortment become more relevant to more people, based on each individual’s notion of a “reasonable” price for that item

This results in the ad campaign driving more site traffic for less investment, at a higher conversion rate. Simply: you get more sales dollars in return for your marketing spend. If we’re measuring campaign performance, not long term business impact, a sale or promotion based campaign is going to win every time.

This makes performance marketing appear to be a better investment when the brand is on sale or promotion. On a near term, last click basis, this might be true, as long as you’re not getting so promotional that you lose money on the sale. But on a long term basis, it will lead to big, systemic trouble, as I’ll explain below.

The Outcome

  • Your performance marketing budget will always be higher when the brand is on sale, because sales will be higher
  • Your budget will always be more efficient when the brand is on sale
  • So you will always acquire more new customers when the brand is on sale

At the end of the year, your customer file will look like this:

If you let things play out like this for a few years, your customer file will be full of people that were acquired during markdown. It will become more and more difficult to sell to your own audience at full price.

This is a simplified example that assumes no customers will lapse out of the business — which never happens. Average YoY customer retention rates for apparel retailers are 20–30%. And guess which customers are more likely to stick around in low-cost marketing channels like email? The customers who are waiting for you to break sale.

Some may argue that any customer is a good customer, and the customers you acquire during promotional periods could come back to buy at full price. This is not true in any significant way — I’ll explain why in another post.

Perverse Incentives:

Agencies: most digital agencies are paid a percentage of media spend or client earnings, and the client manages success based on a ROAS goal. This structure encourages the agency to grow their business by pushing for bigger and bigger budgets during markdown events when the KPIs will naturally look better. They don’t have your P&L, so they can’t see that your company is losing money every day you “crush it” with that clearance sale-on-sale promotion.

In-House Digital Subject Matter Experts: Similarly, if you have junior team members running individual digital marketing channels and ROAS will determine their performance review, they are going to push you to spend during markdown. The retail industry doesn’t invest in training anyone, so many digital marketers come to the field with a deep technical knowledge of channel best practices and little knowledge of big-picture business impact.

The Algorithm: Even if you have a dedicated prospecting strategy, you’re still trafficking your ads via algorithms that show people what they like to see. And what they like to see is sale. On an ad-serving platform like Facebook, the ads people interact with most are boosted by the platform, and the ads that people ignore get suppressed. It is very hard to run both a full-price campaign and a markdown campaign simultaneously without the full price creative getting buried.

Your Other Sales Channels: They’re selling a lot of the same things you are, and are faced with similar pressures. Maybe they’re a publicly traded company who has to hit quarterly earnings targets, and you aren’t. Maybe they generate all their profit from selling cloud hosting and also happen to run an ecommerce business that just needs to break even. Even if you can resist the pressure, they won’t, and guess where the customer is likely to shop?


This is the way that things have played out in the retail industry since ecommerce adoption started to accelerate around 2010. We are now in a place where many mature businesses are on sale most of the time, which gives the average consumer the perception that everything will go on sale eventually. This is harmful not only because it erodes profitability, but because it shifts consumer perceptions of value, potentially permanently.

There are no easy answers or best practices that will break a business out of this cycle. But going beyond a surface level understanding of the situation will help everyone make better, more informed decisions. I will also be following up with some posts about how brands can pursue mindful customer acquisition that drives sustainable profitability.

I hope you’ll consider this as you launch that “Black Friday Is Here” campaign today (the Wednesday before Thanksgiving).

I used to design mom jeans (really). Now I help build bridges between quants and creatives and write about the future of retail.