01
Product strategy
Pre-mortem on a feature launch. Stress-test your pricing tier. Surface the assumptions you can't see from inside the build.
Strategic research, productized
Stress-test the
decision before
you ship it.
A parallel team of latest-Opus agents reads your target the way twelve different operators would. You get a six-to-twelve page strategic brief in 48 hours. $97.
Built on the framework used internally for 171 strategic decisions
What you're buying
Twelve research agents in parallel.
One report.
Most strategic research is one analyst reading slowly. A Dialogue is twelve specialized agents reading the same target simultaneously through different lenses, with a synthesis pass that surfaces the disagreements and the convergences. The disagreements are usually where the value is.
01
Latest-Opus models, max effort
Every agent runs on claude-opus-4-7 with extended thinking enabled. Not the cheap model, not the average effort.
02
Multiple angles, in parallel
Skeptic, optimist, customer, competitor, future-self, first-principles. Each agent reads the same target through one specific lens.
03
Synthesis surfaces disagreement
Where the agents converged becomes consensus. Where they disagreed becomes the section you actually need to read.
04
Structured PDF + markdown
Six to twelve pages. Findings, evidence, recommendations, open questions. Sharable with your team.
Best fit
When to commission a Dialogue.
02
Marketing positioning
Have we cornered a category, or written a brochure? Twelve agents read your site like twelve different ICPs and tell you where the message lands and where it leaks.
03
Architecture choices
Postgres or DynamoDB. Monolith or services. Your stack decision walks into a room with adversaries who've shipped both, and walks out with a defensible answer.
04
Competitive analysis
Where the competitor is strong, where they're soft, what they're about to do, and what you should do first. Specific moves, named, with the reasoning attached.
The deliverable
Six to twelve pages.
Your inbox.
Forty-eight hours.
Every report is structured the same way, so you know what you're paying for before you submit:
- 01 Executive read One page. The three things you most need to act on, ranked.
- 02 Convergence findings Where the parallel agents agreed. These are the load-bearing observations.
- 03 Disagreement zones Where the agents split, and why. The sections you actually need to read twice.
- 04 Specific recommendations Named moves, ordered by leverage. Each with the reasoning attached.
- 05 Open questions What we'd want to know to push the analysis further. Often the most useful page.
Format: PDF + raw markdown. Both delivered. The markdown plays well with your own AI tooling.
The math
Ninety-seven dollars to surface what you'd otherwise miss.
Most operators making a strategic call have already spent two weeks reading the landscape. The decision still feels uncertain because reading alone is one perspective at a time. The expensive failure mode is shipping confident in a direction the data didn't actually support.
The cost of getting one strategic decision wrong rarely lands under five thousand dollars in burned engineering, lost positioning, or a marketing arc that ran the wrong direction for a quarter. The cost of getting twelve specialists to read the same target in parallel and surface the disagreements is ninety-seven dollars and one form submission.
The cheapest insurance you'll buy this quarter is the one that tells you what you're missing before you ship.
Common questions
What people ask before submitting.
How fast?
Report in your inbox within 48 hours of submission. Most arrive in 24.
What's actually in the report?
Six to twelve pages of structured analysis. Specific findings, not generic frameworks. Each claim backed by evidence the agents pulled. Delivered as PDF and markdown.
What URLs work best?
Public landing pages, product pages, pricing pages, blog posts, GitHub repos, documentation sites. If a human can read it without login, the agent team can analyze it.
Who runs the research?
Russ Gardner, working alongside a team of latest-Opus agents. The agents do the parallel reading and angle generation. Russ runs the synthesis pass and signs the report.
Confidential?
Yes. Submissions are not published, not used as training data, not shared. The framework is open. Your specific question stays between us.
Refund policy?
If the report doesn't surface at least three insights you didn't already have, full refund. We've never been asked, but the offer stands.
Can I ask follow-ups?
One round of clarifying questions is included. Reply to the delivery email within seven days.
Why $97?
Cheap enough to commission on instinct. Expensive enough that we can deliver the work. The economics are explained two sections up.
Submit your target
Tell us what to read.
Three fields. Stripe checkout. Report in 48 hours.