Blue Apron spent $154.5M on marketing and revenue fell anyway
Blue Apron’s biggest marketing year was followed by a 24% fall in revenue, then a 32% fall after that. The filings show what happens when acquisition spend is covering for a retention problem.
One vote per reader. You can change it. No account needed.
Four years in one table
| Year | Net revenue | Marketing | Marketing as % of revenue |
|---|---|---|---|
| 2016 | not compared here | $144.1M | — |
| 2017 | $881.2M | $154.5M | 17.5% |
| 2018 | $667.6M | $117.5M | 17.6% |
| 2019 | $454.9M | $48.1M | 10.6% |
Two things in that table are worth sitting with. The first is that the biggest marketing year, 2017, was followed by the first big revenue fall. The second is that when marketing was cut by 59% in 2019, revenue fell by 32% — not by 59%, but not by nothing either.
What that second number means
If marketing spend had been buying customers who stayed, cutting it would slow new growth while the existing base held. That is not what happened. Revenue fell nearly a third in the year the spending stopped, which tells you how much of the revenue depended on a constant supply of new customers rather than on returning ones.
The treadmill, described plainly
A meal-kit subscription has a natural enemy: the customer eventually decides that cooking the same box every week is a chore. If the average customer stays a few months, the business has to replace a meaningful share of its base every quarter just to stand still.
While the acquisition channel is cheap and the money is available, this looks like growth. Revenue rises, the marketing dashboard reports a positive return on each cohort, and the retention problem is invisible in the top line. It only becomes visible when the spending stops — which is exactly when the company can least afford what it reveals.
How to tell whether you are on it
Look at revenue from customers acquired more than twelve months ago. If that number is flat or falling while total revenue grows, the growth is being bought, not earned. This is one query against your own orders table and it is the single most useful thing in this article.
Compare payback period against average customer life. If it takes five months of margin to repay the cost of acquiring a customer who stays six, the business is working extremely hard to break even.
Do not let the blended number hide the cohorts. A healthy old cohort and a terrible new one average out to something that looks acceptable and is not.
The fair caveat
Blue Apron was operating in a category with genuinely difficult economics — perishable goods, expensive delivery, low switching costs — against competitors with more capital. The spending was a symptom of the category as much as a cause of the decline. But the pattern generalises well beyond meal kits, and the diagnostic above works in any subscription business.
Questions readers asked
Is high marketing spend always a warning sign?
No. Spending 17% of revenue on marketing is normal for a consumer subscription business in a growth phase. The warning sign is what happens to revenue when the spending changes.
What is a reasonable payback period?
It depends entirely on how long customers stay. The useful comparison is not against an industry benchmark but against your own average customer life: payback should be a fraction of it, not most of it.
Did cutting marketing cause the 2019 revenue fall?
The two happened together and the filings do not separate cause from effect. The point of the article is narrower: revenue that falls 32% when acquisition spending stops was substantially dependent on that spending.
Where do these figures come from?
Blue Apron Holdings filed 10-K reports with the SEC for each year shown. Net revenue and marketing expense are audited lines in those filings.
This analysis is based on publicly available filings and reporting, all linked above. It represents our editorial interpretation of those sources. We have no affiliation with, and receive no payment from, any company named in this piece. If you believe a figure here is wrong, tell us — corrections are made publicly and the article is marked as updated.
Put your budget and channels into the simulator and see the range published cases landed in.