Amplitude
Each of these is a real decision Amplitude is working through, written so you can make the call yourself — incomplete information, someone who pushes back, no answer key.
Amplitude is a digital analytics and experimentation platform that helps product-led companies understand user behavior in web and mobile products and use that data to drive engagement, conversion, and revenue optimization. Its product portfolio includes its analytics platform, Statsig, and proactive Agents, including the Amplitude Global Agent, Custom agents, and an MCP surface. It serves product, growth, and analytics teams at organizations ranging from early-stage startups to large enterprises through a subscription SaaS model priced on tracked users and event volume. Its in-house OLAP engine, Nova, processes trillions of events in real time, turning raw behavioral data into fast, trustworthy answers that power decisions for thousands of product teams worldwide. Amplitude is a publicly traded company that went public via a direct listing in September 2021 and trades under the ticker AMPL.
Your first one is free. No card.
19 live right now.
The rest
Shape a partner offer for AI measurement so three LATAM buyers sign
Partners are losing AI product conversations because buyers hear separate capabilities instead of one commercial outcome. You must weigh a focused first offer against the breadth enterprise buyers expect.
Turn cohort evidence into a retention case without overpromising
Product managers can find retention patterns but struggle to use them in roadmap funding conversations. You must weigh a sharper investment narrative against claims the behavioral evidence cannot support.
Join AI cost to retention without exposing raw usage
Product managers cannot connect repeat AI-feature use to the token budget behind it. You must weigh a reusable cost integration against privacy limits and uncertain identity mapping.
Prove an activation funnel in 14 days without widening privacy scope
The prospect cannot reconcile its onboarding funnel with the warehouse view, and approval for a production path is at risk. You have to weigh a useful behavioral picture against a data boundary the privacy lead will defend.
Refresh warehouse audiences before Friday’s launch review
The customer’s product managers are rebuilding cohorts by hand because refreshed audiences are not reaching their funnel. You have to weigh a fast customer-specific integration against preserving the event definitions their reports already depend on.
Trace query failures while engineers keep the final call
Funnel and experiment results stall, and on-call engineers rebuild the failure story by hand. You must weigh faster evidence assembly against a diagnosis engineers can actually trust.
Connect agent traces to funnels without rewriting event meaning
Product managers cannot see agent-led steps in the same funnel as customer actions. You have to weigh a useful first surface against durable query semantics and reliable results.
Close a deal without capping product insight
A £720,000 agreement is stalled because procurement needs cost certainty and product needs room for more behavioral data. You must weigh contractual predictability against the usage and partner commitments required for a durable launch.
Settle how usage is estimated so finance stops changing the bill
A finance estimate is blocking a proof-of-concept for the product team’s activation funnel. You have to weigh complete behavioral coverage against a first-year cost range the buying committee can defend.
Lock the retention definition before the data feed retires
The customer has 21 days to retire its existing feed, and product and finance already report different retention results. You must weigh reporting continuity against a definition that remains trustworthy after the cutover.
Reconcile the onboarding funnel before Friday’s product review
One mobile onboarding step now produces conflicting activation rates because its properties split after a release. You must restore a trusted measure without breaking the customer’s historical view.
Choose a value story for enterprise buyers without hiding usage costs
Only 18% of engaged enterprise opportunities reach qualified pipeline after pricing enters the discussion. You must weigh a sharper outcome story against claims that create cost expectations the business cannot support.
Recover trial starts to 13% without overclaiming product value
Trial starts after a retention-view visit fell from 14% to 9%. You have to weigh a faster conversion path against a message that accurately reflects what the product delivers.
Settle how nulls are counted before consumers adopt the contract
Complex funnel queries can return plausible but different counts when optional profile data is missing. You must weigh faster generic-table execution against a compatibility commitment customers cannot easily unwind.
Fix session totals that shrink with profile filters before rollout
Session queries with profile properties return smaller cohorts than equivalent event queries. You need to restore correctness without slowing the analysis path people already rely on.
Prove legacy events survive the schema cutover
A dry run found mappings that preserve names but alter key funnel meaning before an irreversible cutover. You have to weigh broad automation coverage against the evidence needed to trust each critical mapping.
Decide if the migrated history still adds up
Replay tests hit the throughput goals, but the automated summaries disagree on a few identity-merge and late-event cases. You have to judge whether that evidence justifies a cutover you cannot cheaply undo.
Add a review step for reports with sensitive data
Reports tagged with sensitive attributes get exported or left unshared because sharing controls offer no real review step. You have to pick a review model that holds up for enterprise buyers without slowing everyday analysis.