293 complaints extracted from 293 reviews of Amazon DynamoDB between 2016-03 and 2026-02, average rating 4.35/5, grouped into the problems users report most. Each problem links to the market it belongs to and to startup ideas built on it.
The problems Amazon DynamoDB users report
DynamoDB is hard and expensive to work with — 57 complaints about Amazon DynamoDB; 64 complaints across 5 products · market pageUsers repeatedly hit the same wall: DynamoDB's query limitations force expensive workarounds (Lambda functions, S3 exports, additional AWS services), the pricing is opaque and punishing for unpredictable workloads, and the developer experience is frustrating compared to alternatives. Specific complaints include inability to do joins, a 400KB item size limit, no straightforward CSV or JSON export from the console, and needing to write 'a ton of lambda functions' for things RDS handles natively with views and stored procedures.
Unpredictable and opaque DBaaS costs — 42 complaints about Amazon DynamoDB; 74 complaints across 6 products · market pageUsers across multiple managed database services consistently struggle to predict what they'll owe before the bill arrives. Complaints range from 'we cannot estimate the cost when data grows and this leads to unexpected cost' to 'pricing model is very complex to understand' to 'data scanning costs are too high for beginner level users if not provisioned correctly.' The core issue isn't that the services are expensive per se — it's that the cost is structurally invisible until it's too late to change anything.
AWS database consoles too hard to navigate — 18 complaints about Amazon DynamoDB; 23 complaints across 5 products · market pageUsers across DynamoDB, ElastiCache, MemoryDB, QLDB, and Timestream consistently describe the AWS console UIs as cluttered, confusing, and hostile to anyone who isn't already an expert. Complaints are specific: 'editing is clunky in the web UI, sometimes it doesn't really work and gets conflicts on what the keys are,' 'accessing the rows and queries is not very user friendly,' and 'a lot of features are dumped on the same screen.' The friction isn't about missing features — it's that the existing interface makes routine tasks like browsing records, running queries, and managing alarms harder than they should be.
No SQL querying on NoSQL databases — 18 complaints about Amazon DynamoDB; 19 complaints across 2 products · market pageUsers of DocumentDB and DynamoDB repeatedly hit a wall when they need to do anything beyond simple key lookups — joins are impossible, triggers don't exist, ad-hoc queries are painful or outright blocked. Complaints are blunt: 'joining tables is impossible,' 'cannot do ad-hoc queries as a traditional database,' 'SQL interface not available which limits ease of viewing the data.' The frustration isn't that NoSQL is bad — it's that analysts, data teams, and developers who inherited these databases have no practical way to ask cross-entity questions without writing application code.
NoSQL data modeling and query complexity — 15 complaints about Amazon DynamoDB; 16 complaints across 2 products · market pageUsers across DynamoDB and QLDB repeatedly struggle with the same core problem: the mental model for structuring data and writing queries is fundamentally different from relational databases, and there's no guardrail to catch bad decisions before they become expensive performance problems. Complaints range from 'query syntax is hard' and 'case sensitive and complex query structure are not up to mark' to 'more complicated schemas can be hard to fit in the range key paradigm' — meaning the pain hits both beginners designing their first table and experienced engineers trying to do something non-trivial with indexes.
Slow autoscaling causing throttling and over-provisioning — 14 complaints about Amazon DynamoDB; 16 complaints across 2 products · market pageUsers are getting throttled because autoscaling takes up to 10 minutes to respond to traffic spikes, but they're also paying for over-provisioned capacity to avoid that throttling in the first place. Complaints range from 'auto scaling should not take 10 minutes' to 'pre-configured max read/write limits do not accommodate spikes' to secondary costs from managing spiky indexes with higher throughput minimums. The scaling-down limit of only 5 times per day compounds the problem — you overshoot on the way up and can't correct quickly on the way down.
AWS-only lock-in blocks local development — 11 complaints about Amazon DynamoDB; 16 complaints across 3 products · market pageDevelopers and teams using DynamoDB, MemoryDB, and Neptune cannot run these databases locally or on any infrastructure outside AWS. Complaints are consistent and specific: 'no local deployment,' 'cannot be installed on individual desktops or servers,' 'only available with AWS and cannot be installed in local systems.' This creates two distinct problems — dev/test cycles require a live AWS connection (and cost money), and teams in regulated or air-gapped environments like GovCloud hit hard blockers around backup and feature parity.
AWS specialty database setup and configuration complexity — 6 complaints about Amazon DynamoDB; 14 complaints across 6 products · market pageUsers across DynamoDB, ElastiCache, MemoryDB, Neptune, QLDB, and Timestream all report the same friction: getting started requires esoteric knowledge most developers don't have — one user specifically called out 'calculating throughput' as a blocker, others describe the experience as 'very tiring' and 'difficult to learn.' The problem isn't that documentation is missing; it's that choosing the right configuration across multiple competing options requires expertise that only comes after you've already made expensive mistakes.
In their words
“UI interaction should be better and lot of features are dumped on the same screen.”0/5 · 2023-06-07 · AWS database consoles too hard to navigate
“Development with DynamoDB was incredibly slow and difficult due to not enough tools to support usage.”2/5 · 2016-03-07 · DynamoDB is hard and expensive to work with
“There is less documentation on structuring the schema, which many developers still don't understand.”2.5/5 · 2023-05-26
“It's kind of hard to configure, and you have to have a lot of esoteric knowledge about calculating throughput in order to set things up.”2.5/5 · 2018-11-26 · AWS specialty database setup and configuration complexity
“The API for Node is pretty horrible and unlike any other database I have used, in a negative way.”2.5/5 · 2016-03-07
“Cost: Most important consideration in my opinion. Depends on the usage throughput”3/5 · 2023-05-30 · Unpredictable and opaque DBaaS costs