Feature · Diagnostics

Outreach Diagnostics — find out why replies went quiet.

A strong score that isn't converting isn't bad luck — it's a pattern. Outreach Diagnostics reads your own CRM results and names the pattern out loud, so you fix the actual leak instead of sending more of the same pitch.

It runs on plain local code reading your pipeline — zero AI calls, zero cloud, works offline.

The four diagnoses

  • High Score / Low Reply — your targeting is right but the pitch isn't landing: usually generic openings or buried asks. Fix the first two sentences.
  • Ghosting — conversations that stalled mid-thread get flagged with their last stage, so a single polite bump can revive them.
  • Pitch Fatigue — the same opener sent repeatedly with declining replies tells you the template expired. Rotate angles per niche.
  • Niche Mismatch — bookings clustering in one niche while effort scatters elsewhere shows you where to concentrate — and which portfolio to lead with.

Plain English, like everything else that breaks

Diagnostics shares DNA with Self-Healing, Orty's error philosophy: when something underperforms, you get a readable explanation and a suggested next step — not a shrug and not a dashboard of vanity metrics. If a panel ever errors itself, you get the same treatment: a readable message plus a local log file you can paste into a support email.

Private by construction

Because diagnostics compute locally from your own CRM, there was never a privacy trade-off to argue about. No uploads, no provider keys needed, no internet required. It's analysis your spreadsheet could do — if it had seven stages and followed up for you.

Frequently asked questions

How much history does it need before saying something useful?
Roughly a dozen logged outcomes is enough for the first flags. Before that, Diagnostics stays quiet rather than pretending — false confidence is worse than silence.
Does this need an internet connection or API keys?
Neither. It reads your local pipeline with deterministic code — same machine, works offline, sends nothing anywhere.
Can I disagree with a diagnosis?
Please do. Each flag cites the underlying counts from your CRM so you can judge whether the pattern is real. It's a second opinion computed from your data, not a verdict handed down from a server.