Original practice question. The scoring rubric is our teaching guide. How we research and score answers
TL;DR: Segment customers by how much of your value the provider's feature replaces. Casual users of the generic use case may leave; teams relying on your integrations, workflow, domain quality and controls usually will not. Respond by sharpening differentiation where the provider is weak (specific workflows, proprietary context, enterprise controls, domain evaluation), adjusting pricing for the generic tier, and reducing dependence on that provider so it cannot squeeze you further.
How to approach it
Avoid a reflexive answer. Assess the threat precisely: what the provider's feature does and does not do, which customer segments it affects, and how quickly. Then decide on product, pricing and dependency responses, and how to communicate with customers and the team.
A strong answer
1. Assess the overlap. Use the provider's feature seriously for a week. Compare it with yours on your evaluation set and on your customers' real tasks. Typically the provider's version is good at the generic case and weaker on integrations, workflow depth, domain accuracy, admin controls and compliance features.
2. Segment customers by exposure.
| Segment | Exposure | Likely behaviour |
|---|---|---|
| Individuals using the generic use case | High | Some will switch to the free feature |
| Teams using integrations with their data | Medium | Stay if your context makes answers better |
| Enterprises relying on workflow, admin, compliance | Low | Stay; provider's consumer feature does not meet their needs |
| Domain-specific users (legal, finance, health) | Low to medium | Stay if your domain quality is measurably better |
Look at usage data to size each segment and watch early churn signals in the exposed group.
3. Product response: double down where the provider is weak.
- Deepen workflow ownership: collaboration, approvals, history, the place work actually happens.
- Deepen proprietary context: more connectors, better retrieval over customers' data.
- Publish domain quality: show measured performance on tasks that matter to your segments.
- Enterprise readiness: permissions, audit logs, data residency, admin controls.
4. Pricing response. For the generic use case, consider a cheaper or free tier to keep users in the funnel, and concentrate pricing on the capabilities the provider does not offer.
5. Reduce dependence. If your product works only on this provider's model, they hold your margin and your roadmap. Put a model gateway in place, keep your evaluation suite portable, test alternative providers and open-weight models on your tasks, and be ready to route traffic elsewhere. This also strengthens your negotiating position.
6. Communicate. Tell customers clearly what your product does that the generic feature does not. Internally, give the team a clear plan; platform shocks hurt morale more when the response is vague.
If the overlap is near total. Sometimes the honest conclusion is that the product was a thin layer on the provider's capability. Then the decision is a pivot toward a segment or workflow the provider will not serve, done fast.
The clearest public case is a platform absorbing a product rather than a model provider, and it shows the speed. In its results for the fourth quarter of 2024, Chegg wrote that "we would not need to review strategic alternatives if Google hadn't launched AI Overviews", and reported that non-subscriber traffic was down 49% year on year in January 2025 (8-K, February 2025; the causal attribution is the company's). By October 2025 its board had concluded the review in favour of becoming a skilling-focused business-to-business company (10-K, March 2026). Read it for the timeline: the exposed segment moved within months, and the pivot was to a buyer the platform did not serve.
What interviewers probe next
"Should we stop building on that provider?" Not necessarily; their model may still be the best for your tasks. Reduce dependence so it is a choice, not a trap.
"How do you know whether churn is driven by the launch?" Compare churn in exposed versus less-exposed segments before and after, and ask churning customers directly.
How it's scored
| Signal | Weak | Hire | Strong Hire |
|---|---|---|---|
| Assessment | Panic or dismissal | Compares features | Tests on own evals and real tasks, segments customers by exposure with data |
| Response | Compete on price | Differentiate | Specific product moves where the provider is weak, pricing restructure, dependency reduction |
| Judgment | Not considered | Customer communication | Recognises when overlap means a pivot, measures segment-level churn |
Common mistakes
Assuming all customers are equally at risk.
Competing head-on with a free feature on the generic use case.
Remaining wholly dependent on the provider that just entered your market.
See platform risk and lock-in and the moat ladder.