Three ways to
work with us.
The case studies show the engine working on public data — commercial research, public-opinion polling, operational data. These are the shapes a paid engagement takes when you apply it to yours.
Who this is for
Insights and research teams with a measurable outcome — purchase intent, brand fit, awareness, churn, conversion, support, vote intention, make-or-miss — and a decision waiting on the answer. Electric Insights is data-agnostic: the engine works on commercial-research data (concept tests, conjoint, MaxDiff, pricing, brand tracking, copy testing), public-opinion polling, and operational or behavioral data (transactional, panel, event-level, sensor). The Approach page covers the methodology in depth and the three buyer worlds in detail.
If you are not sure which tier fits, start with the diagnostic. It is designed to make that decision easier.
A note on terms. We use the term Mirror below for the coupled deliverable: the score, the model behind it, the vulnerability checks, and the interactive tool — shipped as one object rather than as separate documents. The Approach page covers the rationale.
Engagement Options
A short diagnostic, a full applied case study, or an ongoing methodology partnership. Most engagements start with the first and expand from there.
Methodology Diagnostic
A structured look at whether your data supports a coupled release — and what it would reveal.
Fixed fee · 2 to 4 weeks
What you get
- A Design-stage assessment of your data: whether the measured factors support a disclosed explanatory model, and which vulnerability checks are informative to run.
- A pilot explanatory model of the outcome with outcome-level scenario estimates — a slice of what a full Mirror would contain.
- A concise written report on what the approach reveals that standard reporting did not, plus a scoping recommendation for whether Tier 2 makes sense.
- A working demo tool built on your data, hosted by us for 60 days.
What you bring
- An existing dataset you already own, with a measurable outcome — most often a concept-test or brand-tracking column (top-2-box purchase intent, top-box brand fit, awareness), but also churn, conversion, or any other binary the business cares about.
- A short brief on the decision or question the data is meant to inform.
Applied Case Study
A full coupled Mirror for your data — score, model, vulnerability checks, and inspection tools, delivered together.
Scoped · 6 to 10 weeks · Hosted option adds annual fee
What you get
Includes everything in Tier 1, plus:
- A full Design specification — declared outcome, drivers, scenario rules, interval routine — locked before modeling.
- The complete vulnerability check suite: variable ranking, synthetic-variable stress test, plus nonresponse sensitivity for sample-based studies.
- The full interactive tool suite: explorer, model builder, and scenario simulators — so your team can refit on new subgroups, drop or add predictors, and run sensitivity tests after delivery, without a follow-up engagement.
- A domain-specific landing page with your framing and copy, pulling everything into one coupled release.
- Choice of deployment: self-hosted by your team, or hosted by us.
What you bring
- A dataset, codebook, and outcome column. We help with the codebook if it is not yet structured.
- A decision or strategic question the tool is meant to serve.
- A working session or two with your research or domain team.
Methodology Partnership
An ongoing relationship in which the three-stage framework becomes part of how your organization releases research.
Annual · 12-month initial commitment
What you get
Builds on Tier 2 across multiple studies, plus:
- A library of coupled Mirrors across your research programs, sharing a common Design vocabulary.
- Training for your internal team in the Design / Model / Release framework.
- Pre-fieldwork methodology review for new studies, where the framework's leverage is greatest.
- Option to co-author methodological notes positioning your organization as a thought leader.
- Priority scheduling and named methodological support.
What you bring
- Multiple ongoing research programs or a strategic initiative with repeat analysis needs.
- Internal capacity to adopt and extend the methodology over time.
Recent applied work
Three live coupled Mirrors built on public data — one in each of the three buyer worlds. Each one is what a Tier 2 deliverable looks like in practice.
1963 Harris Approval Survey
Historical opinion-research dataset with three binary outcomes (approval, vote intention, tax-cut support). The reference case for the methodology, tied to a paper currently under peer review at Public Opinion Quarterly. n = 1,283.
Open the demo Commercial researchCraft-Beer Concept Test
Concept-test microdata with a top-2-box purchase-intent headline. Five tools live for the brand team to interrogate package perception, claim strength, and segment differences after delivery. n = 803.
Built in partnership with emi Research Solutions, who deliver Mirror engagements to their clients under their own brand.
Open the demo Operational, behavioral & event dataNBA Shot Outcomes
Per-player shot-log data with continuous predictors (defender distance, shot clock) and entity filtering across 33,362 attempts. Event-level census, no sampling.
Open the demoAdd-Ons
Optional components that can sit inside any tier and are priced separately.
Codebook Construction
When your data is not yet codebook-ready, we produce one. Variable definitions, level labels, category structure, and a format compatible with the tools.
Methodology Workshop
A half-day or full-day working session for your research team. Explanatory modeling, scenario framing, and how to read and communicate results in outcome-level terms.
Extended Modeling
When the outcome is not naturally binary or the data has a time structure, we extend the modeling approach to fit. Scoped and priced separately.
Methodological Co-Authorship
For clients who want to publish the applied work. A named co-authorship arrangement that positions your organization alongside the methodology.
How an engagement works
The exact schedule depends on the tier, but every engagement follows the three-stage framework — with a scoping conversation at the front and a handoff at the end.
Scoping
We learn about your data, your outcome, and the decision waiting on it. Standard NDA if needed. No charge, no obligation.
Design
We declare the outcome, candidate drivers, scenario rules, interval routine, and vulnerability checks — in writing — before modeling begins.
Model
We fit the explanatory model, run diagnostics, translate results into outcome-level terms, and execute the declared vulnerability checks. You review drafts along the way.
Release & Handoff
The coupled Mirror is delivered as a single object — score, model, vulnerability checks, and tools together — along with documentation and a working session for your team.
Common questions
A short list of what prospects usually want to know before a first call. The FAQ covers the methodological questions in more depth.
Do you work with confidential or proprietary data?
Yes, under a standard mutual NDA. Data can be analyzed on your infrastructure, on ours, or on a shared sandbox, depending on what your data-governance process requires.
Who actually does the work?
Every engagement is led personally by the founder — see the About page. Methodological work is not subcontracted, and engagements are not run on junior analysts reporting up to a partner.
What happens to the tools after the engagement ends?
You choose at scoping. In the self-hosted option, we hand off the code and you deploy it; your license is perpetual for internal use. In the hosted option, we run it for you on an annual fee. Either way, the model and documentation are yours to keep.
What if my data is not naturally binary?
Many outcomes can be expressed as binary in a way that still answers the decision — top-two-box versus not, renewal versus not, favorable versus not. When that reframing does not fit, we extend the modeling approach; see Extended Modeling under add-ons.
How do you handle scope changes mid-engagement?
Fixed-fee engagements stay fixed against the original scope. If the scope expands — more variables, a different outcome, additional tools — we agree on a written amendment before the new work begins. No hourly creep.
Can we start small and scale up?
Yes, and most engagements do. A Tier 1 diagnostic is the usual first step. If the results warrant more, the scoping recommendation tells you what a Tier 2 or Tier 3 engagement would involve, and the diagnostic fee can be credited toward the larger engagement.
Not sure where to start?
A short conversation is usually the fastest way to find out which tier fits. No deck, no pitch — just a scoping call.