Explore Problems

After Allstate offered a competitive home insurance rate, its self-inspection app flagged the customer's roof (under 4 years old) as failing, leading Allstate to decline coverage and cancel the policy after already collecting payment through the mortgage escrow. The original insurer, unaware of the switch, kept covering the customer in parallel, leaving them fighting to recover a year of duplicate payments.

Developers writing UI test automation must hunt for selectors and maintain brittle scripts that break as the app's interface changes, and existing AI-assisted testing tools charge per test run rather than per test created. This creates both a maintenance burden and an unpredictable, usage-scaling cost problem for teams adopting AI-driven test automation.

A marketer paid $149 for 1,000 B2B leads and generated zero customers, then discovered the failure wasn't due to messaging quality but the lead source itself. This points to a recurring problem where purchased contact lists are low-fit or low-intent, wasting acquisition budget before any copy or outreach quality comes into play.

Teams managing many complex, overlapping projects and subtasks in mainstream project management software find the experience cluttered and overwhelming, while customization and advanced capabilities remain gated behind higher-priced subscription tiers. This creates friction for teams that need more power but are reluctant or unable to upgrade.

A customer paid for service but the account was never verified, the SIM card was mis-shipped twice, and without a verified account there was no phone number through which to reach support - leaving them paying for a service that never activated. This describes a structural onboarding trap where activation failures block the very support channel needed to fix them.

Teams using AI coding agents like Claude Code and Codex default to expensive frontier models for all implementation work, even routine bounded tasks that lower-cost models could handle under supervision. This creates unnecessary cost without an established way to delegate work by tier while still gating final approval behind stronger judgment.

A buyer paid a $149 nonrefundable fee to ship a CarMax vehicle sight-unseen based on the online listing, only to find undisclosed mildew odor, water damage, cracks, and rough idle upon arrival that were not visible in photos or description. The mismatch between online listing and actual condition, combined with a non-refundable pre-inspection shipping fee, left the buyer out money for a vehicle they would not purchase.

A CarMax customer's extended warranty claim for major engine repairs was denied because routine oil changes had been performed at an independent shop rather than a shop the warranty provider deemed certified. After a poor customer service experience, the customer was offered a refund that withheld roughly 18% of the original warranty cost despite the provider never honoring coverage.

A small business owner paid $750 for a promotional package expecting real leads and job opportunities but received no meaningful results, describing a mismatch between what was promised during sales and what was actually delivered. Customer support offered no acceptable resolution, leaving the business with a real financial loss and no recourse.

Founders and lawyers report spending significant time manually scanning third-party NDAs, MSAs, and DPAs for non-standard indemnity and liability clauses against their company's risk playbook. Experienced feedback suggests the real risk isn't the initial flagging but over-flagging boilerplate, which erodes reviewer trust, and that the most valuable target is ranking which deviations are actually negotiable in high-volume MSAs and DPAs rather than NDAs. The space already competes with in-house counsel doing this scan by hand.

A customer sent a demonstrably broken power tool for repair under its Lifetime Service Agreement, with tachometer measurements proving one speed setting was non-functional, only to have it returned unrepaired with a claim it tested fine. Follow-up support was an unresponsive overseas call center with no callback, leaving the customer with a still-broken tool and no recourse.

A customer explicitly declined to cancel their Allstate auto policy, but the representative canceled it anyway, refused to reverse the action, and hung up. Despite speaking with over 20 different representatives across a month, the customer has been unable to get the erroneous cancellation reversed or the resulting $50 fee refunded.

A Home Depot associate incorrectly confirmed a customer had all the correct bathtub-surround components, an error only discovered mid-renovation when contractors found a broken and mismatched part. The mistake delayed the renovation by five days and forced a hotel stay, yet the store offered only $70 in compensation for the associate's own error.

A customer arranged with a Home Depot store to hold four discounted tools under their name, but the store sold three of the four units before pickup despite the confirmed hold. Subsequent attempts to get the agreed price honored via phone and online support bounced between departments without resolution.

A long-time Notion user says the app has grown noticeably glitchy since early 2026 and has worsened since, and would switch instantly to a more stable planning alternative if one existed. They attribute the decline to recent AI features crowding out core product stability, an active alternative-seeking signal from an otherwise loyal user of a large incumbent tool.

An AT&T internet customer reports frequent outages and support calls that get transferred without resolution. The unreliability directly affects work-from-home productivity, highlighting a gap in proactive outage communication and support routing.

A customer describes routinely waiting 45 minutes to over two hours to reach a live agent for billing or device issues, compounded by difficulty communicating clearly with support once connected. The account points to systemic under-staffing or queue-routing problems in the support channel.

A customer trying to check on a family member's account was left on hold for over an hour with no answer, describing a decline in support responsiveness since switching to a bundled phone-and-internet plan. The experience reflects a broader pattern of degraded customer service accessibility after service bundling.

New users of feature-rich project management platforms struggle to determine the best way to organize work amid many overlapping features and organizational methods, and find that useful capabilities are reserved for paid tiers. This creates an onboarding barrier and limits value for free-tier users evaluating the product.

A user has been unable to locate or cancel their Canva subscription for two months despite the sign-up process being simple. This reflects a broader dark-pattern issue where subscription cancellation is made deliberately harder to find than subscription signup.