We offered our first discount in 8 years. It didn’t go well.

In December, we did something extremely unusual for our business: We ran a discount.

And the result was … a total flop.

In fact, here’s an actual chat from the moment we sent the email out:Screen Shot 2018 01 19 at 9.47.48 AM

I’ve been publicly against discounting for a long time.

In general, discounts are a lazy way to convert customers to buy. Most marketers lead with price, discounting by 25%, 40%, even 50%, watching a surge of orders come in, and assuming they’ll sell those customers more on the back-end.

What they don’t realize is that those discount customers will hardly ever convert to full-price buyers. These discount customers generate a disproportionate number of refund requests and dramatically cut into profit.

These businesses are now in the discount spiral of creating more and more discounts for the same result. They stop innovating and creating truly must-have products, instead leaning on the convenient crutch of discounts. Over time, frequent discounters like Gap train their customers to wait for the next sale.

It’s the rare company that holds the line on pricing, creates incredible products, and charges a fair price to customers who are delighted to pay. This is what we did in our business for years and years.

So why did we try a discount?

And why did it fail?

Why we ran a discount for the first time in 8 years

We’re always looking at our business and testing our assumptions. Here are a few core assumptions we’ve tested:

  • People won’t pay for content online (they do — our top 1,000 customers spent more than $8 million with us)
  • People won’t read long pages (our top sales page is more than 75 pages long)
  • You need to leverage social media to sell (well, this might be true … but my social media doesn’t sell much of anything)

We want to balance our core values — which don’t change — with testing different tactics to reach our market. Those tactics change over time. As Marshall Goldsmith says, “What got you here won’t get you there.”

So when we took a holistic look at our business, paying special attention to the hundreds of thousands of readers who have never purchased anything from us … we noticed that “price” was one of the major objections.

If you had a group of hundreds of thousands of loyal readers who’d never bought from you, and their primary reason was “price,” what would you do?

We thought about this carefully.

  1. We knew that we had a large segment of readers who had never bought from us and that “price” was a major barrier for them.
  2. We knew that when a customer buys from us once, they tend to love our material and buy from us again. The key was getting them to buy for the first time.
  3. Therefore, we concluded that if price was a concern — and we could remove that barrier — we might build a group of long-term customers who otherwise would not purchase from us.

Hence the idea of a one-time discount.

We also studied a wide variety of other companies that discount, including discounting a can of soda cheaper when it’s cold outside, the SaaS product that uses discounts to encourage yearly payments, and premium brands that use “whisper sales.”

We decided to make a bet and try a one-time promotion offering special pricing for select courses.

How we rolled out our special pricing campaign to 120,000 people

Here was our plan:

Step one: Send an email to a small cohort with a subtle promotion. Watch reactions.
Step two: Depending on reactions from the test cohort, we could continue by sending to a larger cohort.
Step three: Send a second, shorter email that more directly mentioned the promotion to everyone. Sales would close that night.

We had two rules:

  1. No leading with price. We would not create a gaudy $9.99 $3.99 pricing page.
  2. We would limit the promotion to a few select courses.

Our goal here was “tip” prospects who were on the fence about trying one of our courses, knowing that our product quality and customer service teams would convert them into long-term customers.

There were a few possible outcomes we were prepared for:

  • Our readers revolt and feel slighted that we offered discounts after years of discouraging them. And what would the people who have already paid full price think?
  • It blows up and we convert a massive crop of students who wouldn’t have normally signed up.
  • It’s middle of the road, with a few folks signing up and little impact to our bottom line or brand.

After agreeing on the approach, we set aside a week on our calendar and hoped for the best.

Here’s what email one looked like:

Notice that the email led with value. 2/3 of the way down, we introduced the special promotion. There was no mention of it in the subject line. This was all intentional.

We took great pains in writing this email, drafting several versions with multiple edits before settling here.

When you clicked, the page linked here:

This page featured short snippets of the selected products. Each button highlighted the savings (“Save $500”). In retrospect, we should have spent more time describing the products. It may have helped to include prices of the products.

Remember: we wanted to send to a tiny subset of people first to avoid totally misreading the situation and angering our readers. So we sent to a VERY small group of 2.6k. And then, this happened:

First cohort (2,600 people):IWT Email 2 1

What you’re seeing is a 32% open rate and a 1.6% clickthrough rate. These are relatively normal numbers for us. And most importantly, we didn’t get any negative reactions about offering a discount.

But…

There were no responses either. There were also no sales. This may have been a sign no one cared. Or that we sent it to too few people. We weren’t sure.

So, per our plan, we sent to more people (but not EVERYONE ELSE) later in the day.

Second cohort (30,000 people):IWT Email 4

All of these numbers are fairly standard. In other words, people were getting the email, reading it, and some were clicking. We got similar responses — in other words, no emails back (positive or negative).

And also…

No one was buying.

No response. No sales, no email replies … nothing. It was like we were shouting into a dark void.Screen Shot 2018 01 19 at 10.00.15 AM

You would expect something. Maybe angry emails. Maybe people saying “FINALLY! I’ve been waiting!” But to get nothing at all?

Ugh.

In fact some of the only reactions we received were readers (rightly) wondering why we were going back on one of our own rules.Screen Shot 2018 01 19 at 9.52.27 AM

It was time to make a call. We’d seen barely any email responses and fewer than 10 sales. We decided to send to the rest of our segment and watch the reactions overnight.

Third cohort (86,000 people):IWT Email 3

And the result was the same as the previous two cohorts: nothing.

Everything was technically working. The sales forms worked. Clicks were being tracked.

Eventually, we started to see a few sales. But no questions about the promotion or the courses. It was like we threw a party and nobody came.

Here’s my question for you: If you were in this situation, what would you do? Here’s the scenario:

  • You just emailed your entire segment with a sale offer, but you’ve gotten poor response: A few sales, no reply emails.
  • You’ve doubled-checked your technology and everything is working correctly.
  • You have a second email scheduled for tomorrow.

What would you do?

We huddled and decided we should do something uncharacteristic for us. We quit. We cut our losses and didn’t even send the second email.

MA = marketing automation team. EOY = end of year. SOD = start of day.
Ryan is ordering that the second promotional email be cut.

We decided to reconvene in the morning. Maybe people would read it and buy later. Maybe people were at work and wanted to get home before buying. At this point, WHO THE HELL KNEW? Certainly not me.

My fiancée noticed I was walking around aimlessly with a blank look on my face. “What’s going on?” she asked. I told her. (She used to be a retail buyer for 8+ years.)

Her response:

“Oh, boohoo … you run a business where your customers don’t love discounts and only want to pay full price. What a tough life!”

DAMN, roasted. LOVE YOU BABE!

I woke up the next morning and asked my team for the numbers. What did weeks of planning, lots of rewrites, and endless wrangling over the positioning get us?

$42,384.

Now $42,000 is real money to a solopreneur or small team. But when you factor in our opportunity cost — all the hours of brainstorming, copywriting, design, and testing, plus what we could have done with other marketing strategies — this was a loss.

What we learned

There are lots of possibilities for why this didn’t work. In descending order:

  • The promotion wasn’t big enough (10% off doesn’t motivate as much as 50% off)
  • The promotion wasn’t clear enough (it was buried at the bottom of the text of one email, obscured by a subject line of “I’m glad my teachers taught me this”)
  • The timing was off (after Christmas, typical retail logic = people have spent their money)
  • Our prospects/customers aren’t motivated by price or 10% promotions … or non-buyers claim price is a barrier but it’s really not (this one has a lot of possibilities and you should carefully think about this)
  • A combination of the above
  • Something totally different we didn’t even consider

Here’s the part of a GrowthLab post where we usually tie it up with a bow with lots of takeaways. But the ugly truth here is that business is messy. It’s not always entirely clear why something works or doesn’t work. And sometimes “test your assumptions” means those tests will fail. I wish I could point to one of those bullets as THE reason. But I can’t.

What’s even more confusing is that this is a strategy that can’t really be A/B tested. As I previously wrote on a Quora post about A/B tests:

You generally can’t A/B test strategy.

With really major changes like offering a new product or which product to offer — changes that can potentially provide HUGE ROI or cost you a lot — you can educate yourself with customer research, etc, but ultimately you have take a risk and make a decision.

Will we run a promotion again? I don’t know. If we do, we’ll take a fresh approach and build from the ground up.

The bigger lesson is that after 14 years of running this business, I still have a lot to learn. As your business gets more complex, there are more and more choices to make — and fewer and fewer exact formulas to follow.

I love my team’s willingness to experiment with new approaches, even radically new ones. And we’ll continue sharing our lessons learned — from shutting down a $2 million subscription product, how 1,000 customers make us $8.6 million in revenue, to why it’s important to rewrite your story — here at GrowthLab.

While I can’t pick out one of those bullet points as a lesson for our business, I do know my personal lesson: Even after 14 years of doing this, I still have a ton to learn. In fact, we all have a ton to learn, all the time. Despite what dozens of blog posts will tell you, sometimes there’s no way to know the exact reason something worked or didn’t work. There’s no fool-proof system.

Those messy truths are the kinds of things we love to write about here on GrowthLab — and I’m glad you’re along for the ride.