Users in some countries (e.g. Yemen) report that all *.supabase.co domains are completely unreachable without a VPN, causing total outages for authentication, database, storage, realtime, and edge functions. This was reported in mid-2026.
“None of the *.supabase.co domains are reachable without using a VPN. Applications that depend on Supabase stop working completely because they cannot connect to the service. This affects applications built on Supabase, making authentication, database access, storage, realtime, and edge functions unavailable for users in Yemen.”
DevOps teams struggle to secure CI/CD pipelines that have broad permissions to sensitive systems. There is a lack of dedicated tooling for monitoring, compliance checks, and automated security enforcement, forcing teams to rely on manual best-practices documents. This pain was raised in 2020 and remains a structural concern given the privileged access CI/CD systems require.
“I want my devops to be focused elsewhere but on the other hand the CI/CD pipeline has permissions to access and do many things in very sensitive places. I know there are 'best practices' documents for how to secure, but are there tools to monitor, offer compliance checks etc?”
Teams with apps that don't run 24/7 (e.g., 9-to-5 business apps) cannot properly configure Datadog to avoid false alerts during planned downtime. Maintenance windows exist but cannot be scheduled on a recurring basis, making availability tracking unreliable for non-continuous workloads. This was reported in 2022 and no resolution is indicated in the discussions.
“The vast majority (500+) of the company app don't run 24*7, they have different business hours (like 9-5) but outside of these business hours, apps are shutdown. And Datadog don't support that. It has maintenance window but they can not be scheduled, the same is valid for monitoring mute.”
As of 2021, developers reported repeated Auth0 outages (four in two months) that knocked out their apps and triggered customer complaints. Auth0's own status page showed 99.87% availability, which users found inconsistent with their lived experience. This pain is from 2021 and may have improved since, but it was a significant driver of alternative-seeking at the time.
“After the fourth Auth0 outage in the last two months has knocked out my app, I'm somewhat at a loss of what to do.”
Multi-stage agentic RAG pipelines in production environments can be extremely slow — 5–10+ seconds before a stream starts — making them unsuitable for real-time user-facing applications. This is a recent (2026) pain point from teams running RAG on cloud AI services.
“Right now, the system is extremely slow (often 5~10+ seconds before the stream starts), and the overall performance and instruction adherence from the LLM are quite poor.”
Teams building AI apps in production are unsure how to implement proper audit trails for LLM interactions — traditional observability tools don't map cleanly to prompt/response logging needs, especially for debugging or compliance purposes.
“with LLM-based apps it feels less clear what the standard approach is, especially if you need proper audit trails for debugging or compliance. curious how teams are handling this in production are people mostly building their own logging pipelines, or are there reliable tools/platforms that help with storing and auditing LLM interactions?”
Developers building or using AI coding assistants find that current tools rely on grep or superficial vector search, failing to understand code semantics, structure, or intent. This is a live, recent pain (July 2025) from practitioners who use these tools daily.
“grep-based search doesn't understand semantics. It finds strings, not meaning. Cursor? A good start with vector search, but still shallow. No deep understanding of code structure or intent.”
Users building or using AI desktop agents are frustrated by the inability to control what the AI can access or do on their system, prompting them to build alternatives with more granular control.
Users report that LinkedIn is being exploited as a vector for malicious activity, such as job offer backdoors, raising concerns about the platform's trustworthiness for professional interactions.
Rotating or expired client secrets in Microsoft/Azure SSO integrations cause production outages with no proactive alerting, forcing developers to manually track expiry dates.
“my company are increasingly rolling this out int”
A recurring problem across multiple tools and time periods: roles assigned in the Identity Provider are not correctly propagated or mapped to the target application, causing users to land with wrong permissions after SSO login.
“After successful SSO login, users with the 'Admin' role in Keycloak are being mapped to 'Viewer' in Airflow. We expect users with the Admin role in Keycloak to retain their Admin role in Airflow as well.”
Java developers on Linux trying to authenticate against a remote Active Directory face poor documentation and complex configuration requirements for Kerberos, with no clear recommended path for remote/cross-platform scenarios. (Older discussion from 2014; may be partially addressed by modern libraries, but the underlying complexity remains.)
“Although there are different solutions for Java based application to authenticate against AD (Kerberos, NTLM, LDAP from Authenticating against Active Directory with Java on Linux), but I don't think they are for remote AD authentication from what I read. Kerberos is said to be the 'new' mechanism for AD authentication recommended by Microsoft. However, it seems that either conf files or system properties are required for it to work.”
Founders observe that even when the core product delivers value, poor onboarding causes users to churn before reaching the 'aha moment', making activation a critical unsolved problem.
“I had a few customers so I know that it is working and providing value. But what's happening next is really stressful.”
Early-stage SaaS products lose the majority of trial users who silently fail to configure the product correctly and never return, rather than reaching out for help. This results in poor activation rates.
“User uploads videos but they did not configure the settings correctly so they end up with bad data. User tries a few more times or gives up or reaches out to me. The majority of people are just giving up and never returning.”
Coordinating user account creation with payment validation (e.g. Stripe) is error-prone — users can be created before payment is confirmed, leading to invalid registrations.
“I want to register the user only if Stripe payment is valid... 3. Validation --> User is created 4. Pay us”
Habit tracker builders report deeply engaged, emotionally attached users who nonetheless do not convert to paying customers, making sustainable revenue a critical unsolved problem.
“My users say they'd cry if my app disappeared. None of them pay”
Many SaaS founders are losing MRR to failed payment recoveries (involuntary churn) without even measuring it as a distinct churn category. This silent revenue leak is underappreciated and under-tooled, with founders lacking automated dunning or recovery workflows integrated into their billing stack.
“Involuntary churn—when customers are lost due to failed payment recoveries rather than active cancellation—is silently eating into the monthly recurring revenue (MRR) of SaaS businesses. A recent pain point analysis from Dev.to reveals that many founders aren't even tracking this metric, let alone optimizing for it.”
Teams generating tens of millions of events per day find that GA and Firebase Analytics do not provide easy access to raw data pipelines, forcing complex workarounds via BigQuery exports.
“My company is using Firebase Analytics to track events from our apps. Everyday, these apps can generate over 30-40 million event tracking lo… I have problem with finding a service which can be alternative to Google Analytics”
As of May 2026, PostHog announced plans to train AI models on developer and end-user data (session replays, interaction data, metadata). Users reacted strongly, feeling the communication was deliberately obscured. This is a very recent and acute trust/privacy concern.
“Posthog wants to train on developer data which means MILLIONS of end-users' session replays, interaction data, metadata, etc. What's worse is they were trying to sugar coat it with their 'funny' emails (even the subject line was hella sugar coated)”
Teams evaluating B2B product analytics tools in 2023 found that no single platform combined good customer support, session recordings, and semi-flexible event/insight creation. Multiple tools were tried and rejected, signalling a persistent gap in the market.
“Looking for something that has - good customer support (I think this is the most difficult one) - Session recordings - Semi flexible in terms of events and insights creation. Tried out Posthog, June, Amplitude, Logsnag and none of them satisfy the conditions.”
Teams evaluating B2B product analytics tools in 2023 report that no single tool adequately combines good customer support, session recordings, and semi-flexible event/insight creation. Multiple tools were tried and found lacking across these dimensions simultaneously.
“Looking for something that has - good customer support (I think this is the most difficult one) - Session recordings - Semi flexible in terms of events and insights creation Tried out Posthog, June, Amplitude, Logsnag and none of them satisfy the conditions.”
SaaS founders report that adding more support tools and headcount paradoxically worsens customer satisfaction, increases response times, and drives churn. Agents end up spending the majority of their time navigating tools rather than helping customers. This is a recent, active pain (2025 discussion).
SaaS founders and support teams report that deploying chatbots inside CRM systems actually lowers customer satisfaction, even when dashboard metrics like response time appear to improve. This creates a false sense of progress while real customer experience deteriorates. This is a recent, active pain (2025 discussions).
“Added a chatbot → customer satisfaction dropped 40%. But hey, the dashboard said 'response time improved,' so it looked like a win.”
As of late 2023, users report that live chat widgets with AI (including Intercom's Fin) hallucinate and fail to provide reliable support, while still commanding high price points. This is a recent, active pain point.
“We are now on our forth live chat widget, they are all embarrassingly bad, hallucinate with their AI, and are almost amusingly expensive.”
Teams using shared inbox tools (Front App, Exchange) report poor UX — missed emails, confusing interfaces, and sync conflicts when multiple agents work the same inbox. Users want a simpler, more reliable collaborative inbox without the full complexity of a helpdesk.
“Problem is I find the UX complicated. It's very confusing and easy to miss emails or responses. My team feel the same way. I also think it's on the expensive side.”
Users depend on automatic bank imports for accurate budgeting, and when providers remove or degrade this feature, it breaks core workflows.
“YNAB will partly remove direct bank imports It was a bummer to me that the famous budget app YNAB announced in a private in-app message that the direct bank data import will be removed.”
Developers relying on API-reported token counts to estimate costs find discrepancies with what cloud cost management tools actually bill, making cost forecasting unreliable.
“Token usage from `response.done` event does not match Azure Cost Management meter data”
Across thousands of SaaS customer reviews, unclear pricing tiers and opaque limits are one of the most universally cited complaints, spanning completely different product categories. This is a persistent, cross-industry pain point as of early 2026.
“Pricing confusion – tiers and limits aren't clear. One negative review is noise. But when dozens of customers say the same thing, it's basically product intelligence.”
A strong, sustained demand (2020–2026) exists for FOSS or self-hosted Notion alternatives, driven by concerns about vendor lock-in, pricing changes, and data sovereignty. The sheer volume of 'open-source Notion alternative' posts across years confirms this is an unresolved market gap.
“we were curious what are some really good alternatives to both which can be collab and used, if it is OSS Solution then would be great, so we can also volunteer for the development too.”
Across many years of discussions (2016–2021), users consistently describe Jira as overly complex, bloated, and difficult to configure — even when they acknowledge its flexibility. This is the single most recurring complaint and spans indie developers, startups, and enterprise teams alike.
“We find Jira very clunky and hard to use, although it's very flexible (though extremely hard to configure).”
Users find feature-heavy tools like Notion slow to open and distracting, preferring lightweight apps that let them focus on content rather than the tool itself.
“Nothing comes close to Windows Notepad. As it opens instantly, you spend all of your cognitive load on the note itself rather than the app”
Engineers feel that tools like Jira prioritize management tracking over developer productivity, creating overhead and enabling micro-management rather than helping engineers ship faster. This is a recent (2024) pain point driving interest in Linear alternatives.
“It is slow, bloated and often acts as a burden to engineers. These tools didn't help engineers in getting the work done faster, they only helped the management in tracking the work which enabled a lot of processes and micro-management which used to kill our productivity.”
Teams receive feedback from many sources (social media, app stores, support tickets, sales calls, Slack, etc.) and struggle to aggregate it into one place for analysis and action.
Developers building multi-app products find themselves repeatedly building custom feedback forms and manually requesting screenshots from users. This friction was highlighted in 2024 as a key motivation for building dedicated tooling.
“I got fed up with chasing users for screenshots and building custom feedback forms for every app, so I built FeedbackThing.Pro to make it easier. Now, all I do is drop in a simple <script> tag, and users can send feedback, screenshots, bug reports, and ratings effortlessly.”
Users note that most feedback tools only capture generic, untargeted input. There is no easy way to tie a piece of feedback to a specific feature, page, or user action — and no built-in screenshot capability for end users to annotate issues in context.
“All I've found so far doesn't emphasize on the need of targeting what feedback relates to what. General feedback is often required. Comments, screenshots (here another problem! there is not such tool -afaik- for providing issues screenshots to a startup, instead, the end user has to do it manually and send it, in case she/he decides to report it), but it's such a mess to define concretely what the feedback is about.”
Developers and PMs repeatedly describe the manual effort of soliciting feedback, screenshots, and bug reports from users as a major time sink that distracts from building.
“I got fed up with chasing users for screenshots and building custom feedback forms for every app”
Founders and marketers find that most AI-powered marketing tools only generate recommendations or content ideas but fail to actually carry out the work, leaving execution still on the user.
“Most AI tools stop at suggestions. Vect AI is built to help founders and marketers actually do the work”