Instacart's S-1 bodes well for the underlying strength of the business
Daniel McCarthy
Associate Professor of Marketing at the Robert H. Smith School of Business, University of Maryland, College Park
Instacart just released their pre-IPO prospectus. I'm neck deep in the middle of the beginning of teaching semester rush, but wanted to share a few thoughts as I was going through the filing.
Bottom line: the underlying fundamentals appear very strong here.
Here's the quick summary for why I feel this way:
Let me share a bit more about each of these points.
Cohort revenue data implies customers are superannuities
Very strong cohort-level performance. Let's unpack these figures:
Each cohort-year, more revenue than the year prior, except for the COVID cohort (2020) (not surprising - many that year were forced in). More impressively, all other cohorts held on to COVID bumps -- eg., 2017 cohort in years 3-5 was 1.74x, 3.0x, *3.2x* of year 1 GTV. The bump stuck!
So while $CART is B2C, it has the revenue durability of a so-so SaaS company. $CART beat us over the head with this superannuity point later on with this figure:
Nothing scary in the fine print of the chart, from what I could tell:
Instacart+ is an impressive driver of growth and durability
It is astounding how big the subscription program
If you pay a fee for free delivery, you're probably regularly using the service for your groceries and are less likely to defect (or use competitors at all - why, if you're now paying a fee that gives you free delivery if you place the order with Instacart?). This implies to me their revenue is very "firm". I have a paper coming out on this very topic.
Impressive cohort profit economics
Gross profit as a percentage of GTV expands meaningfully within-cohort over time. Let's unpack these figures:
For example, consider the H1 '19 cohort: -2.9% GP % of GTV in year 1, vs 3.4%, 5.7%, 6.0%, 7.8% in years 2-5. This too is unsurprising. $CART isn't doling out discounts to high-tenure customers - those customers will come without any pushing. As the business matures, the GP % will go up a lot.
But that's not all. In addition to strong performance within-cohort, we also see attractive performance across cohorts. eg., year 2 GP % was -3.4% for the H1 '19 cohort, vs 4.4, 5.5%, and 7.7% for the H1 '20, H1 '21, and H1 '22 cohorts. Both bode well for GP over time.
COVID is not artificially goosing $CART's numbers anymore
$CART makes a reasonable case for this via a time series model
I reach the exact same conclusion in a recent academic paper I co-wrote with Shin Oblander (https://bit.ly/330SUSJ) - grocery delivery was one of the categories we studied.
The ads business is impressive
30% of revenue is now from the ads business. It is growing like gangbusters. This is very high-margin revenue. It is 2.6% of GTV -- I am not an expert, but it feels like there is still meaningful room to grow from here.
$CART is profitable
Not just 'loss excluding all expenses' profitable. Not 'we had a quarter of profit during peak COVID' profitable. They are 'past 5 quarters in a row, without COVID support, with ramped customer acquisition costs
A small gripe
I'll end with a gripe. I'd love to do a proper CBCV on $CART. But like $DASH, $CART intentionally left out a tremendous amount of data on active customers and gross customer acquisitions over time. Why? They know. They disclosed a smidge of this data. Just not close to enough.
Frustrating!
We could probably back into it, triangulating off of Earnest Analytics data using a methodology like the one from an earlier paper of mind: https://bit.ly/2UlwfxO
But why make us go through those hoops?
In any case, multiple very obvious sources of operating leverage
Your analysis is great but I am also concerned by what was left out. There's clearly been a massive increase in customer acquisition spend, but I couldn't see any analysis of customer acquisition cost or payback. The word "churn" appears twice but no data is given on customer churn and subsequent reacquisition. The growth of their advertising revenue looks good but I wonder if they end up with >100% of their profit from advertising. Overall, it's hard to unravel the true marginal economics and - what I've seen with similar businesses in Europe - is that the parts of the business that are profitable are not growing, and the parts that are growing are not profitable!