Pangea’s growth problem looked like an acquisition problem until the company examined retention closely enough to challenge the premise.
That distinction is the central lesson of this case.
According to a 2023 Reforge case study, Pangea began as a college-focused freelance marketplace. The company had raised $3 million, graduated from Y Combinator in 2021 and expanded its network to 1,800 universities, but it was not seeing the sustainable growth trajectory its team expected.
The initial instinct was to acquire more users.
The retention analysis pointed somewhere else.
This case is useful because it shows how a growth team can move from a channel-level diagnosis to a business-model decision. It is also a reminder that reported case-study results should be interpreted carefully: the figures below come from a Reforge customer story based on interviews with Pangea’s co-founder and were not independently audited by Radar Digital.
The context: scale without a sustainable growth curve
Pangea’s original model connected college students with freelance work.
From the outside, the ingredients looked promising:
- venture funding;
- Y Combinator;
- broad university distribution;
- a clear target population;
- a large potential supply side.
But growth is not simply distribution.
A marketplace can acquire both sides and still fail to create enough repeated value to retain them.
The company had already scaled to 1,800 universities, according to Reforge, yet management did not believe the business was on a sustainable trajectory.
That is an important operational signal.
When a company has already expanded acquisition and distribution, the next question should not automatically be:
How do we acquire even more?
It should be:
Are the users and customers we already acquire creating enough repeat value for the growth model to compound?
The diagnosis changed from acquisition to retention
Pangea co-founder Adam Alpert told Reforge that he initially believed acquisition was the growth problem.
After working through retention analysis, the team reconsidered.
The company plotted retention curves and observed that retention was not strong enough to support the growth trajectory they wanted. Reforge reports that the curve sat somewhere between poor and acceptable and flattened only after roughly 12 to 18 months.
The exact shape matters less than the management implication.
If a cohort decays too far before stabilizing, pouring more acquisition into the top of the funnel can increase the number of users entering a weak retention system.
That can create:
- higher acquisition spend;
- more operational load;
- more inactive users;
- weaker payback;
- pressure to discount or subsidize growth.
Retention analysis therefore becomes more than a product dashboard. It becomes a test of whether the business model deserves more acquisition.
The team moved from quantitative diagnosis to qualitative research
Retention data told Pangea that something was wrong.
It did not tell the company what the new business should be.
That required qualitative work.
According to the Reforge case, Pangea then:
- interviewed hundreds of teams and talent;
- spoke with recruiting and staffing experts;
- tested different positioning and value propositions;
- launched decoy brands and websites;
- attempted to close real revenue on new concepts.
This is a strong example of using quantitative analysis to decide where to investigate, then using qualitative and commercial evidence to decide what to change.
The team did not simply optimize onboarding copy or add another lifecycle email.
It questioned the value proposition.
Why the pivot was a growth decision
Pangea moved toward a contract-to-hire talent marketplace.
The reported changes included:
- opening the platform beyond the original college-only audience;
- eliminating talent-side fees and commissions;
- verifying companies before allowing them to access the marketplace or post jobs;
- focusing the proposition around curated talent open to contract-to-hire opportunities.
Those changes affected more than marketing.
They changed:
- target market;
- marketplace rules;
- pricing;
- trust architecture;
- onboarding;
- supply positioning;
- demand positioning.
This is why the case is useful for marketers.
Sometimes the highest-leverage growth intervention sits outside the marketing department.
If the market structure is wrong, better campaigns will not repair it.
The reported results
Reforge reports three core outcomes following the pivot:
- customer LTV doubled;
- payback period declined by 85%;
- Net Dollar Retention increased 4x.
Pangea also relaunched on Product Hunt and reportedly won Product of the Day and Product of the Week.
These are substantial changes.
But they should be interpreted correctly.
The source does not provide a controlled experiment proving that a specific retention framework caused each result. Multiple elements changed at once:
- audience;
- business model;
- pricing;
- onboarding;
- marketplace design;
- positioning.
The correct conclusion is not:
retention analysis causes 4x NDR.
The more defensible conclusion is:
retention analysis helped the company recognize that the existing model was not strong enough, which triggered deeper research and a broader strategic pivot that the company later associated with materially better economics.
That is still a powerful lesson.
The most important metric was not acquisition
The original diagnosis was “we need more growth.”
Acquisition was the obvious lever because it is visible and controllable.
More ads. More partnerships. More top-of-funnel volume.
But acquisition cannot create durable growth if users or customers do not continue receiving value.
This is especially important for:
- marketplaces;
- subscription businesses;
- SaaS;
- communities;
- creator platforms;
- consumer apps.
A company with poor retention has a growth ceiling.
The faster it acquires, the faster it can discover that ceiling.
Retention analysis should connect to economics
A useful retention program does not stop at cohort charts.
The operator should connect retention to:
- LTV;
- CAC;
- payback;
- gross margin;
- expansion;
- referral;
- channel quality.
Pangea’s reported post-pivot outcomes are notable because they include LTV, NDR and payback.
These are economic measures.
A team can improve weekly active users without materially improving the business.
Economic linkage makes the analysis more consequential.
What operators can copy
1. Start with the growth model
Map acquisition, activation, retention, monetization and referral or expansion. Ask which component is actually limiting sustainable growth.
2. Plot cohort retention
Do not rely only on aggregate active-user numbers. Look at cohort shape, stabilization point, segment differences, time to value and repeat behavior.
3. Translate the curve into a business question
If retention is weak, ask whether the audience is wrong, the job-to-be-done is weak, onboarding is failing, value is delivered too slowly or the business model creates the wrong incentives.
4. Move to qualitative research
Talk to retained users, churned users, prospective buyers, lost deals, experts and adjacent segments.
5. Test the new proposition before a full pivot
Use landing pages, sales conversations, prototypes, concierge delivery and limited pilots. The goal is evidence of real behavior and willingness to pay.
What not to copy
Do not copy Pangea’s pivot simply because it produced strong reported metrics.
The company operated in a specific market at a specific time.
Avoid copying price points, fee structures, marketplace rules, audience expansion or Product Hunt tactics.
Copy the diagnostic sequence.
A useful decision framework
When growth stalls, ask in this order:
- Is acquisition working?
- Are users activating?
- Are they retaining?
- Are the economics sustainable?
- Is the underlying proposition strong enough?
That last question is uncomfortable.
It is also where this case becomes strategically useful.
Limitations of the case
The case has several constraints:
- the principal source is Reforge, which has a commercial relationship with the customer;
- the reported metrics are company-reported;
- detailed baseline values are not disclosed;
- several major strategic variables changed at the same time;
- there is no controlled counterfactual.
Radar Digital therefore treats the numbers as reported outcomes, not independently verified causal estimates.
The operating lesson
The strongest lesson from Pangea is not “pivot when retention is low.”
It is that teams should be willing to escalate the level of the problem.
A weak team asks which campaign it should run next. A stronger team asks which stage of the growth model is broken.
And when the answer points to the product or business model, the organization needs the discipline to stop optimizing the wrong layer.
Pangea’s case shows retention analysis functioning as a strategic alarm.
The important outcome was not a prettier cohort chart.
It was the realization that acquisition was not the problem the company actually needed to solve.
