New
How are you handling semantic layers and metric definitions across BI tools?
dbt semantic layer, a headless BI layer, or definitions living in each dashboard tool? What has reduced the “two numbers for revenue” problem for you?
Evaluation guide
How to evaluate data and analytics platforms
This guide is for data and analytics leaders, data architects and engineering managers replacing a warehouse, consolidating BI tools or building a lakehouse. It assumes your data already lives in several places and the question is where it should be stored, transformed and queried from here.
In short
Use this guide to compare warehouses, lakehouses, streaming and BI tools on what decides the outcome: open formats and data portability, cost control under real query load, performance on your workloads, governance and lineage, and how AI features hold up on your own data. Then shortlist three to five companies on Techarda and test them with your own queries.
What’s changing in Data & Analytics
Before you compare anyone, it helps to know what has shifted in the market, because it changes which questions matter.
Open table formats are separating storage from engines
Formats such as Apache Iceberg and Delta Lake let several engines read the same tables in your own object storage. That makes changing query engines more realistic, and moves the lock-in question to the catalogue that tracks the tables.
Consumption pricing needs active management
Most platforms charge for compute as you use it. That suits uneven workloads, but a badly written query or a dashboard that refreshes all day can run up a bill before anyone notices.
AI analysts are arriving in BI
Vendors now let people ask questions of data in plain language. These features work best on a well-modelled semantic layer and give confident wrong answers on messy data, so test them on your own definitions.
Streaming and batch are converging
More platforms handle real-time ingestion and batch processing in one place. Whether you need data within seconds depends on the decision it feeds, and many teams pay for freshness nobody uses.
What to judge vendors on
Those shifts shape the criteria below. Each one says why it matters and what to ask, so you can take it straight into a vendor call.
| Criterion | Why it matters | What to ask vendors |
|---|---|---|
| 1Open formats and data portability | Your data will outlive any engine you choose. If it sits in a proprietary format, every future change starts with a migration. | “Can our data sit in an open table format in our own storage, can other engines read and write it, and who controls the catalogue?” |
| 2Cost control under real load | Consumption pricing hides risk in concurrency, refresh schedules and inefficient queries. Budgets break on the workload nobody tested. | “Can we set hard spending limits by team or workload, see the cost of each query and dashboard, and get alerts before a limit is reached?” |
| 3Performance on your workloads | Vendor benchmarks use tidy data and chosen queries. Your mix of joins, concurrency and data volumes is different. | “Will you run our own queries on a copy of our data during the trial, at our expected concurrency, and share both the timings and the cost?” |
| 4Governance, lineage and access control | Analytics spreads data to more people. Without lineage and fine-grained access, you cannot say who saw what or where a number came from. | “How do row and column level controls work, does lineage cover transformation and BI tools, and do policies follow the data across engines?” |
| 5Consistent metric definitions | If revenue is defined three ways in three tools, dashboards and AI features will disagree with each other and with finance. | “Where are metric definitions held, can every BI and AI tool use the same ones, and are they versioned and tested?” |
| 6AI features on your own data | Plain-language queries are only as good as the model of your data underneath. A wrong answer delivered with confidence is worse than none. | “How does your assistant use our metric definitions, can users see the query it ran, and can we test its accuracy on our own questions?” |
| 7Ingestion and pipelines | Much of the effort goes into getting data in and keeping pipelines running when source systems change. | “Which of our source systems have maintained connectors, how are schema changes handled, and who fixes a pipeline when it breaks?” |
| 8Residency and data sharing | Sharing features across regions, partners and clouds can move personal information across borders without anyone deciding to. | “Where are our data and metadata stored and processed, and how do sharing features respect residency and consent rules?” |
Trade-offs to settle early
No vendor does well on every criterion, and some pull against each other. Decide where you stand on these before the demos start, or the demos will decide for you.
One platform or a composable stack
A single platform for ingestion, storage, transformation and BI means one contract and one security model. A composable stack on open formats lets you replace parts, and your team runs the joins between them.
Warehouse simplicity or lakehouse openness
A managed warehouse is easier to run and tune. A lakehouse on open formats in your own storage keeps more options open and asks more of your engineers.
Real time or fresh enough
Streaming pipelines cost more to build and run than batch. Match freshness to the decision: fraud checks need seconds, monthly reporting does not.
Self-service or central control
Letting business teams build their own models and dashboards speeds up answers and multiplies conflicting numbers. Decide which metrics are governed centrally before you widen access.
Pay as you go or committed spend
Pay as you go is flexible but hard to forecast. Committed spend brings discounts, and the risk of paying for capacity you no longer need if plans change.
Red flags
As the answers come back, watch for these. One on its own isn’t a deal breaker, but it deserves a follow-up question in writing.
- Performance claims that rest on benchmarks the vendor chose and ran.
- Exporting data in an open format costs extra or needs a separate product.
- No hard spending limits, only alerts after the money is spent.
- AI assistant demos that run on a sample dataset and never on your schema.
- Lineage that stops at the warehouse and never reaches your BI or transformation tools.
- References that have not yet been through a renewal on consumption pricing.
Build your shortlist
Start with the workloads, then use Techarda to narrow the field to three to five companies you can compare fairly.
- List your main workloads (reporting, ad hoc analysis, data science, real time) and the data volumes and users behind each.
- Decide early whether open table formats in your own storage are a requirement. It changes the field more than any single feature.
- Compare like with like: storage and query engines against each other, BI tools against each other, pipeline tools against each other.
- Read recent stories on Techarda for pricing, licence and ownership changes, and plan questions around them.
- Use the practitioner questions below to see where teams like yours hit cost or governance problems, then compare side by side.
Companies to consider in Data & Analytics
Listed by recent activity on Techarda (what people are reading, following and discussing). The order says nothing about quality, market share or fit for your needs.
Latest Data & Analytics stories
Once you have names, recent news is where pricing changes, acquisitions and outages show up first. These are the five newest stories in this category.
What practitioners are asking
News tells you what vendors announced. These open questions show what teams are still trying to work out, with the least-answered first.
New
Postgres everywhere vs purpose-built databases: where do you draw the line?
Vector search, time series, queues, analytics: many teams now start with Postgres extensions. Where has that worked, and where did you need a dedicated engine?
Where evidence is thin
A shortlist is only as good as the evidence behind it, and ours is still growing category by category. Here is what’s missing in Data & Analytics right now.
20 of 20 companies have no published reviews
If you’ve run one of these in production, a short review helps the next team decide. Reviews are moderated before they appear.
2 questions with no answers yet
An answer from someone who has made the same decision is often more useful than any guide. Share what worked, what didn’t and what you’d check next time.
Answer a Data & Analytics questionFrequently asked questions
What should I ask Data & Analytics vendors about open formats and data portability?
Your data will outlive any engine you choose. If it sits in a proprietary format, every future change starts with a migration. Ask: Can our data sit in an open table format in our own storage, can other engines read and write it, and who controls the catalogue?
What should I ask Data & Analytics vendors about cost control under real load?
Consumption pricing hides risk in concurrency, refresh schedules and inefficient queries. Budgets break on the workload nobody tested. Ask: Can we set hard spending limits by team or workload, see the cost of each query and dashboard, and get alerts before a limit is reached?
What should I ask Data & Analytics vendors about performance on your workloads?
Vendor benchmarks use tidy data and chosen queries. Your mix of joins, concurrency and data volumes is different. Ask: Will you run our own queries on a copy of our data during the trial, at our expected concurrency, and share both the timings and the cost?
What should I ask Data & Analytics vendors about governance, lineage and access control?
Analytics spreads data to more people. Without lineage and fine-grained access, you cannot say who saw what or where a number came from. Ask: How do row and column level controls work, does lineage cover transformation and BI tools, and do policies follow the data across engines?
New to some of the terms in this guide? The glossary explains Change data capture (CDC), Data lakehouse, Data warehouse, ETL and ELT, Event streaming and 1 more Data & Analytics terms.
This guide is written by Techarda editors to our editorial standards. No vendor paid to appear in it or saw it before publication. See our methodology, the trust dashboard or go back to the Data & Analytics hub.