Stop running experiments that tell you nothing.
For researchers who plan their own experiments and do not have a statistician to ask.
| Run | Temp °C | Cure min | Catalyst | Result |
|---|---|---|---|---|
| 01 | 60 | 20 | A | 48 MPa |
| 02 | 60 | 20 | B | 60 MPa |
| 03 | 60 | 40 | A | 50 MPa |
| 04 | 60 | 40 | B | 62 MPa |
| 05 | 80 | 20 | A | 54 MPa |
| 06 | 80 | 20 | B | 66 MPa |
| 07 | 80 | 40 | A | 56 MPa |
| 08 | 80 | 40 | B | 68 MPa |
Five stations, start to finish.
Describe
Tell it what you want to find out.
You write your goal in plain words and list what you can change.
- goal in plain english
- no statistics to choose
Define
Say what you can change.
A range you can set anywhere in, a few fixed settings, or a set of options.
- temperature 20 to 80 °C
- stirrer 200 / 400 / 600 rpm
- catalyst A or B
Design
It picks the design and explains why.
The design that fits your variables and your budget of runs, with the reasoning written out.
- full factorial
- d-optimal
- explained in plain english
Record
Run the sheet, type the results in.
Results stay attached to the runs that produced them.
- export to excel
- upload results back
- saved to your account
Analyse
Find out what the data says.
The model tells you which variables matter, how much, and what settings give the best result.
- effects ranked by size
- optimum settings found
- plain english summary
Built for people who have to defend the plan.
You get a run sheet, not a wall of text.
A numbered list of experiments you can print and work from in the lab that afternoon - and a record of exactly what you ran, so you can show your working later.
Every term is explained.
Statistical language appears with a plain English definition on hover, so you can defend the design to a supervisor.
Your work is saved.
Run sheets and results live in your account, not in your browser, so they survive a different machine.
Analysis that shows its working.
Effects, model fit, residuals, and an optimum — computed from your actual results, not generated by a language model. Every number is the same every time.
Designing experiments is free.
Analysis of your results is the only thing that will ever be paid for.
Free
Available nowEverything you need to plan a campaign and record what happened.
- Unlimited experiments
- Plain English design recommendations
- Run sheets you can take into the lab
- Export to Excel and email
- Record and upload results
- Everything saved to your account
Analysis
Available nowMake sense of the results once the runs are done. See which variables drove your outcome and what to do next.
- Which variables actually mattered and by how much
- Whether each effect is real or could be noise
- Model fit, prediction ability, and reproducibility
- The settings that give your best result
- Plain English summary of everything the data says
- Multi-response optimisation and sweet spot analysis
- Mixture experiment support
Frequently asked questions
Do I need to understand statistics to use this?
No. You describe your experiment in normal words. There are no statistical terms to choose from, and where one appears in the explanation, hover over it for a plain English definition.
How is this different from asking ChatGPT?
A chatbot can design an experiment, and often does it well. What it cannot do is hand you a numbered run sheet to work from, or remember any of it next week. ExperimentIQ keeps the design, the runs and the results together, so months later you can still show what you ran and why - same experiment, different machine.
What does it cost?
Designing experiments, generating run sheets and recording results are free. Analysis is £19 a month (regular price £29) while the launch offer lasts. See pricing above.
Who can see my experiments?
Only you. Experiments are tied to your account and access is enforced at the database, not just in the interface.
Design your first experiment.
Free. No card, no setup. Sign in with Google and start.