What is Ordeen Choice Modelling?

A self-serve conjoint analysis platform. You design a study, we recruit and field real respondents through an integrated panel, and the platform returns hierarchical Bayesian utilities, latent-class segments, willingness-to-pay and a live market simulator. Collapses what used to be a six-week agency project into days.

Who is this for?

Pricing teams, product strategists, insights managers and research consultancies. If you are making decisions where “we think customers will prefer this” would be a defensible answer, Ordeen replaces the vibes with numbers.

Do I need to be a statistician to use it?

No. The hard parts (Hierarchical Bayes estimation, MCMC convergence diagnostics, Krinsky-Robb intervals, multiplicity correction) run automatically. You get a market simulator that answers “what if we charge £5 more?” in plain English. For the quant audience, every result ships with its full method and credible interval — nothing is hidden.

What study types do you support?

CBC (Choice-Based Conjoint), ACBC (Adaptive CBC — with per-respondent build-your-own, screener and tournament), MaxDiff (anchored and unanchored), and MBC (Menu-Based Choice). Latent-class segmentation and a continuous Hierarchical Bayes distribution run in parallel so you can pick whichever story is more defensible on your sample.

How is this different from Sawtooth?

Same statistical rigour (HB with NUTS, non-centred parameterisation, LKJ priors, Krinsky-Robb WTP, BH-corrected interactions). Different pricing (flat monthly instead of per-seat), modern UX, cloud-native, and no desktop install. We publish our methodology; read it.

How is this different from Qualtrics?

Qualtrics is a survey platform that happens to do some conjoint. Ordeen is a specialist choice-modelling platform. The difference shows in the diagnostics (ours are published, theirs are proprietary), the simulator (ours exposes credible intervals, theirs does not), and the price (one flat fee vs enterprise seat negotiation).

How many respondents do I need?

For a main-effects CBC with ~8 attributes you want 150–300 respondents; for ACBC or MaxDiff with many items, 300–500. Latent-class with 5+ segments wants 100–150 respondents per segment as a floor. The platform gives you a design-efficiency score before fieldwork so a weak sample is caught before you spend money, not after.

Can I bring my own panel?

Yes. Upload a respondent list and we field through your source. Or use our integrated Prodege panel and respondent packs — pricing is transparent per complete, and the credit pool is shared across all your studies.

What is the pricing?

Choice Modelling: £2,500/month flat + respondent packs. Price Monitoring: £2,000/month, unlimited watches. Forecasting: £2,000/month, unlimited series. Contract Analysis: £1,000/month, unlimited uploads. No per-seat fees on any product.

Is there a free trial?

Yes — 14-day free trial on Choice Modelling. No credit card. You get the full platform (not a stripped demo); the only limit is you cannot field respondents through the panel until you upgrade.

What is Hierarchical Bayesian estimation?

A statistical method that treats respondents as draws from a population distribution. Each person’s utilities are shrunk toward the population mean in proportion to how noisy their own data is — avoiding both the “everyone is identical” error of aggregate logit and the overfitting of per-respondent estimation. The gold standard for choice modelling.

What is willingness-to-pay (WTP)?

The monetary equivalent of a feature’s utility — the price premium a feature commands in your market. Computed from the ratio of attribute utility to price utility, with Krinsky-Robb credible intervals.

Does Ordeen integrate with Qualtrics / SurveyMonkey?

For CBC, MaxDiff and MBC designs you can export the design and import choice data from any survey platform via CSV. ACBC runs on-platform because the adaptive flow depends on real-time response-driven logic. We recommend running ACBC natively; for CBC, use whichever you prefer.

Where is my data stored?

EU region (Google Cloud europe-west1). We never share respondent data between workspaces, never train models on your data, never sell to third parties. Full DPA available on request for enterprise plans.

Is Ordeen GDPR compliant?

Yes. EU hosting, Data Processing Agreement on request, respondent consent captured and auditable, right-to-erasure supported at study level, 30-day backup retention.

Can I share results with my team?

Yes. Studies, segments and the market simulator are shareable via a private link that keeps live interactive simulation (not a static PDF). Share with internal stakeholders or external clients — no additional seats required.

What export formats do you support?

CSV, Excel, PNG for every chart, PDF for executive reports, and raw JSON for your data team. The simulator can be embedded via a shareable URL that stays live after you cancel.

What support do you offer?

Email at im@ordeen.com, response within one working day. Enterprise plans include setup support, methodology review calls and a direct Slack channel.

Who built this?

A quant-led team with backgrounds in pricing consulting, academic Bayesian statistics and B2B SaaS engineering. The methodology is published, not proprietary — our moat is implementation quality, not secrecy.

How do I get started?

Start a 14-day free trial at app.ordeen.com/auth/sign-up. No credit card. Spin up a study in 5 minutes with sample data, or upload a design brief and have us walk you through it.