Grow Users, Lower COGS
Preserve viral growth while moving transactions onto cheaper rails
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 95%
The framework treats adoption and cost as two different optimization problems rather than slowing the whole product to reduce losses. First, determine which transaction path creates the largest variable cost and which users generate activity without durable network value. Then introduce a cheaper rail, migrate legitimate users, and remove behavior driven only by the subsidy. Growth may decline temporarily because headline volume previously included low-quality activity, but retained users become economically healthier. The final test is whether growth resumes after the migration with lower cost per transaction. Venmo applied this by preserving user acquisition while shifting payments from subsidized credit cards to ACH and eliminating users who cycled transactions primarily to harvest credit-card points.
Origin
Andrew Kortina described how Venmo reduced processing costs before its Series B by moving users from credit-card transactions to bank payments and removing points-harvesting activity.
Core principles
- 01Separate user growth from transaction cost growth
- 02Protect valuable users while removing exploitative behavior
- 03Use cheaper infrastructure without abandoning demand
- 04Accept temporary metric declines when the underlying cohort improves
How to run it
- 1
Map the Cost Curve
Calculate the variable cost created by each transaction method, user behavior, or growth channel. Identify the paths that become increasingly expensive as volume rises.
Pro tip Measure costs by user cohort as well as aggregate transaction volume.
Watch out Do not assume that all activity represents equally valuable growth.
- 2
Define Valuable Usage
Separate users who receive the product's core value from users who participate only to exploit a subsidy or incentive.
Pro tip Look for repeated behavior that vanishes as soon as incentives change.
Watch out Avoid labeling heavy users as abusive without examining retention and network contribution.
- 3
Build the Cheaper Rail
Secure the banking, infrastructure, or supplier relationship needed to process legitimate activity at a lower unit cost.
Pro tip Run the new route in parallel before forcing a broad migration.
Watch out Infrastructure negotiations can take longer than the product work.
- 4
Migrate and Prune
Move valuable users to the cheaper path and remove access for users whose activity is economically destructive.
Pro tip Explain the migration around reliability or product continuity rather than internal financing pressure.
Watch out Headline growth can fall sharply when subsidized activity disappears.
- 5
Verify the Recovery
Track retained users, transaction frequency, processing cost, and post-migration growth. Continue only if the healthier cohort resumes organic expansion.
Pro tip Present investors with cohort quality and cost improvements alongside the temporary growth decline.
Watch out Raising capital during the transition can make healthy pruning look like product failure.
In the wild
Venmo was paying roughly three percent to process some credit-card-funded payments while charging users nothing. The company obtained a banking partner, shifted legitimate users toward bank payments, and removed points enthusiasts who were cycling credit-card volume for rewards. Growth declined for several months before beginning to recover.
→ Venmo retained its useful peer-to-peer network while materially improving transaction economics.
Common mistakes
Slowing Every Form of Growth
Reducing acquisition across the board sacrifices valuable network expansion instead of targeting the expensive behavior.
Treating Subsidized Volume as Loyalty
Users attracted solely by a profitable arbitrage may disappear when the subsidy ends and should not be counted as durable demand.
Ignoring Transition Optics
A necessary migration can produce down months that investors mistake for declining product-market fit.
Is it for you?
Best for
It is best for viral marketplaces, fintech products, and networks whose early growth creates significant variable costs.
Not ideal for
It is not ideal when usage costs are already negligible or when migration would destroy the product's core value.
From the transcript
“we would always want like more user growth, but lower cogs.”
“we'll we'll switch people on the bank, uh, which we did.”
“killing all these uh crappy users and retaining the good users and getting everyone move into the bank payments”
From the episode
Venmo (live with Andrew Kortina)
Andrew Kortina