Common questions about BindSignal.

BindSignal is a weekly email for Managing General Agents (MGAs). It reads email metadata across underwriter inboxes on Microsoft 365, groups the traffic by broker firm, and ranks firms by the commercial intent in their conversations and by how far this week departs from what that firm normally does.

It arrives Sunday evening or Monday morning, to a senior team member you choose, naming which broker relationships need attention and which mailbox or person should pick each one up. Nothing to install, nothing to log into and nothing for underwriters to fill in.

When you want to act on a firm, it can also prepare a briefing before the call or draft the next reply. That part is optional, and the two questions below cover how it works.

Only when you ask it to. Which firms appear in the digest, in what order and in which tier comes from our own rules and baselines, with no machine learning.

AI comes in when someone wants help acting on a firm. Press "Brief me for the call" and it reads that firm's recent correspondence and sends you what's open, what changed and what's worth asking. Press "Draft a follow-up" and you get the same with a draft of the next reply. Every figure in a draft has to appear in the correspondence it came from, and a draft that fails that check is left out rather than sent to you half right.

Both are optional. They stay off until an authorised person accepts our AI processing terms, which you can do when you apply for a pilot or at any point afterwards.

No. BindSignal never contacts a broker or sends anything on your behalf. A draft reply is emailed to the person who asked for it, with a button to open it in Outlook, where they edit it and send it themselves, or don't.

A CRM records what closes, and it records it because somebody typed it in. BindSignal answers a question that sits upstream of that: are the right producer conversations getting the right attention, fast enough?

The practical difference is that BindSignal reads the record email already generates, so nothing has to be logged for it to work. It doesn't replace your CRM, integrate with it, or ask anyone to keep it up to date.

Triage tools rank the submissions that arrived. BindSignal works a step earlier, on the broker conversations that produce them, including the ones that never became a submission, and across every connected inbox rather than one at a time.

It's a scalpel rather than a sledgehammer: a narrow read on one problem, and it will not manage everything that lands in your pipeline.

BindSignal places every active producer into one of three tiers each week:

• ACT NOW. Active high-intent signals, such as a bind instruction or a confirmed submission, with traffic going both ways. Capped at three firms a week, because a tier of ten is not a priority list.
• PRIORITY. Commercial activity meaningfully above that firm's own 4-week baseline. Rising submission threads, a first quote request in a while, or a renewal thread reopening after silence. Capped at five.
• MONITOR. Everything else worth seeing, plus any unused slots from the two tiers above. Early signals, and firms with real history that have gone quiet.

Each producer domain is scored weekly using a Composite Opportunity Score (COS) that combines:

• Intent weight. A bind instruction scores highest, a submission next, then quote requests and renewals. General traffic scores nothing.
• Engagement depth. Thread count, traffic going both ways, attachments, and presence in more than one connected mailbox.
• Recency. How recently the commercial signal appeared.
• Abnormality. This week measured against that firm's own 1-week, 4-week and 8-week baselines. The 8-week window is what separates a firm genuinely accelerating from one returning to normal after a quiet spell.

The score is recalculated from scratch each week and capped, so no single firm can swamp the ranking by stacking signals. Comparison is always against the firm's own history, never against other firms. Nobody adjusts a score by hand. During pilots we do hold the digest briefly so an operator can remove anything that looks like an anomaly before it sends, which we would rather tell you about than have you find out.

The digest is a plain email delivered Sunday evening or Monday morning, to a senior team member of your choosing. Each producer appears under its tier with a summary of the commercial signals seen that week, how that compares to the 1-, 4- and 8-week baselines, a suggested next action, and the mailbox or person who should pick it up. For example, "Confirm bind before end of week" or "Follow up on outstanding submission". Where a firm's history is too thin for a trend call, it says so rather than guessing.

There are no dashboards, no logins and no apps to open. Everything arrives by email, including briefings and drafts when you ask for them.

The weekly digest reads email metadata only: sender, recipients, conversation ID, timestamp, whether there were attachments, and the subject line. It does not read message bodies, open attachments, or store email content.

Access is read-only, granted via standard Microsoft 365 app consent. Worth being precise about the subject line, because it does get read at runtime and it is how the intent classifier works. In Microsoft Graph terms a subject is metadata rather than content, so "we never see anything anyone typed" would be too strong. What is true of the weekly digest is read-only, metadata only, no message bodies, no attachments and no persistent storage of email content.

Briefings are the one place message bodies are read, and only if you switch them on by accepting the AI processing terms. With them on, a firm with enough recent correspondence carries a button. If someone presses it, that firm's recent correspondence is read at that moment to prepare what they asked for, and the result is emailed to them. Nothing is read for a firm nobody asked about, attachments are never read, and no message content is stored afterwards.

Each organisation's data is kept separate. There is no cross-tenant data sharing and no persistent storage of email content.

BindSignal is hosted on Microsoft Azure. The connection to your Microsoft 365 environment uses the same security model as any other Microsoft-approved third-party integration your IT team would recognise.

Briefings and drafts are produced by a model on Microsoft Azure OpenAI Service, in the North Europe region, inside BindSignal's own Microsoft tenancy. Nothing is sent to any other model provider.

Your correspondence is never used to train, fine-tune or evaluate a model. Processing is transient: we don't keep the messages read, the briefing produced or a summary of either. What we do keep is a usage record of the firm's domain, the type of request, the time and the tokens used, with no message content and no email addresses. The AI processing terms set this out in full.

BindSignal connects via the Microsoft Graph API using standard organisational app consent. A Microsoft 365 Global Administrator, or an admin with the appropriate delegated permissions, grants the application read access to the mailboxes you nominate, and can revoke it from your own admin centre at any time.

There is no software to install, no agents to deploy, and no changes to how underwriters work. The consent process typically takes under 10 minutes.

Minimal. A Microsoft 365 Global Administrator approves the app consent, which is a standard Microsoft process and typically takes less than 10 minutes. No infrastructure changes, integrations or software deployments are required, before or after.

BindSignal is designed for volume-focused MGAs with 30 or more active producer relationships, where distribution runs primarily through Microsoft 365 email. It works best when:

• Broker contact comes through underwriter inboxes
• Multiple underwriters share distribution responsibility across inboxes
• Several underwriters or distribution staff hold broker relationships across separate inboxes
• Distribution leaders want structured weekly visibility without adding new systems

It is not suited to low-volume, high-touch books where every account is hand-managed and personal judgement is genuinely sufficient, or to teams where broker contact does not come through email. It does not support Google Workspace, and it is not a CRM, pipeline analytics or workflow automation, which is a statement about scope rather than a criticism of wanting those things. We confirm fit before a pilot starts and say so when the answer is no.

Setup typically takes less than one business day from app consent to first data read. Week one establishes the baseline, mapping existing activity without producing a ranked digest. From week two onwards the ranked email arrives weekly, Sunday evening or Monday morning, to whoever you choose to receive it.

Yes. BindSignal does not integrate with, replace, or require changes to any existing CRM or workflow tool. It reads email metadata independently and delivers a digest by email. Your team continues using whatever systems they currently use.

BindSignal sits upstream of those systems, surfacing which conversations deserve attention before anything reaches your CRM pipeline.

BindSignal improves GTM execution quality by making the execution gap visible. Most MGAs have a sound distribution strategy but limited visibility into whether it is being executed consistently.

High-intent producer signals get missed for operational reasons rather than strategic ones. Signals arrive across several inboxes so nobody sees a producer whole, inboxes rank on recency because recency is the only thing they can rank on, and nobody carries a mental baseline for what a normal week from a given firm looks like. BindSignal works on all three without requiring a process change from anyone.

The pilot is a fixed fee of £500 (roughly $750) for four weeks, covering 2 to 10 underwriter inboxes. A typical setup is 4 or 5 individual underwriter inboxes plus a shared submissions or enquiries mailbox. It includes:

• Setup and Microsoft 365 app consent guidance
• Four weekly emails, Sunday evening or Monday morning, with broker firms ranked by commercial intent, a suggested next action, and the mailbox or person who should pick each one up
• Coverage across every connected inbox, so a producer working through three underwriters is read as one relationship
• Call briefings and draft replies on request from the first ranked digest, if you switch them on
• An end-of-pilot debrief covering which threads progressed, which were missed, and where the signal ran ahead of the response

We confirm fit before activation, and if we decide BindSignal isn't a fit, we refund the fee in full. If it turns out not to be right for your operation after week one, we will say so.

At the end of the four weeks we run a debrief, covering which broker relationships moved, which threads progressed, which were missed, and what the four weeks suggest about where your producer book is warming and cooling.

If BindSignal is a fit, the starter package covers up to 20 mailboxes, including 5 shared mailboxes, for under $30k a year. Fixed, with no per-seat creep. Pricing beyond that scales with inbox count. There is no automatic rollover and no obligation to continue.

By week two you'll have a ranked view of your active broker firms. By week three, week-on-week movement becomes visible: which relationships are growing against their own baseline, which are cooling, and which high-intent threads have had no reply from anyone.

Four weeks of your own data is the real test. If it doesn't show you anything you didn't know, it hasn't worked, and we'd rather you said so.

One controlled four-week test against a real MGA's broker traffic. Across 9 inboxes and 47 brokers active in the window, 27 broker threads carrying submission or bind intent had no clear follow-up.

That was one test on one book, run as a test rather than a client pilot, and we don't know what those 27 threads were worth in premium. We'd rather give you the number with its limits than a conversion figure we can't stand behind.