> ## Documentation Index > Fetch the complete documentation index at: https://help.eazybe.com/llms.txt > Use this file to discover all available pages before exploring further. # Deal Analysis > BEA Radar reads your HubSpot deals and the real conversations behind them, then writes its analysis straight onto the deal — so you can see which deals are really moving, which are quietly dying, and which need you today. ## What It Is **BEA Radar — Deal Analysis** reads everything that has actually happened on a deal: your CRM record *and* the real conversations with the people on it. It then writes its analysis straight onto the deal. You don't leave HubSpot. You don't open another dashboard. You open a deal and the analysis is already sitting there beside your own fields — telling you whether the deal is really moving, who owes the next reply, what could stop it closing, and when it is likely to land. Deal Analysis works on **HubSpot deals** and needs a connected HubSpot account. Everything it writes lands as normal deal fields, so you can filter, sort, build list views, and report on it exactly like anything else in your portal. ## Where the Data Comes From Every analysis is built from two sides. The deal itself — stage, amount, owner, close date, pipeline — plus every activity logged against it: calls, meetings, tasks, emails, notes, and the contacts and companies attached to it. The actual chats (for example WhatsApp) and email threads with the people on that deal — the full back-and-forth, including who said what, when, and how long each side took to reply. The two are then read together. **This is where most of the value sits:** the CRM and the conversation often tell different stories, and the gap between them is exactly what usually goes unnoticed. ## When It Runs The analysis runs automatically on a regular schedule. Each run picks up the deals that have had new activity since they were last looked at and refreshes them. You never trigger anything. You open a deal and it's already there — and every deal records when it was last analysed, so you always know how current the picture is. ## What You Get on Every Deal Instead of reading a whole thread to work out where a deal stands, you get the answers waiting for you. Whether the ball is in your court or theirs. Pull up everything where the customer is waiting on you and you have the fastest revenue list of your day. Not just how the conversation looks today, but how it compares to how it was running before — replies getting slower and thinner, or faster and richer. And whether they've gone quiet altogether. The difference between someone describing their own problem, someone weighing up what you do, and someone negotiating price and terms — plus whether the person you're talking to can actually approve it, or has to go get permission. Where the conversation has clearly moved further than your pipeline says — or where a deal was pushed forward more optimistically than the conversation supports. Competitor pressure, pushback on price, silence after a quote, a requirement you can't meet, a timeline that keeps slipping, or an account resting on a single contact. Plus what they're actually objecting to, in their own words. When the customer has said they intend to close, and the amount actually discussed with them — which is often not the figure sitting in your CRM. Alongside these you get a plain-language summary of the whole deal — who's involved, what they need, where it stands, what happened most recently — and concrete suggested next steps for that specific deal. The summary is what you read when you're picking up a colleague's pipeline or walking into a call in two minutes. ## What You Can Rely On Everything written to a deal comes from something that actually happened — in the CRM or in the conversation — and carries the evidence behind it. You can see the exact message or activity a conclusion came from, so nothing is asserted that can't be traced back to something real. Where there's no genuine signal, the deal is left empty rather than filled with an assumption. An empty field is information in its own right: filter for it and you have the deals where that signal doesn't exist yet. A deal isn't marked as gone quiet the moment a reply is late. The customer has to have been sitting on your last message for a couple of days first. If a message goes out in the morning and the analysis runs that evening, they may simply not have got to it. Waiting means that when the flag does fire, it means something — and the list it produces is genuinely worth working. Customers say things like "give us a few weeks" or "after the security review". Those get translated into a consistent set of timeframes, worked out against today's date rather than taken literally. Because everyone's phrasing ends up in the same vocabulary, *"show me everything closing this month"* works across your whole pipeline. Only genuine closing signals count — *"we'll sign next month"* does, *"let's do the demo Friday"* doesn't. Everything is written into its own separate fields. Your deal stage, amount, close date and everything else your team maintains are never edited or overwritten. Where the analysis disagrees with your CRM — a stage that looks wrong, an amount that doesn't match what was quoted — it says so in its own field and leaves yours exactly as your team left it. You decide what to do about the difference. Each deal shows how many conversations went into the analysis. A confident-looking read built on two messages deserves different weight from one built on sixty — so you always know how much material is behind the conclusion. ## Reading It Together The signals are meant to be read as a set. A few patterns worth recognising: ### The deal that's ahead of you The customer owes you a reply, engagement is picking up, they're asking commercial questions, the person you're talking to can approve it, and the conversation has clearly moved past where your pipeline says the deal is. This one is real and moving faster than your CRM knows. Update the stage and push for the close. ### The deal that's quietly dying Nothing looks alarming. They still owe you a reply, they haven't gone silent, and they're still engaged on paper — but replies are getting slower, price landed badly, your contact can't approve it anyway, and the conversation is actually behind where the pipeline claims. Nothing here is on fire, and that's the point. This deal still looks fine in your forecast. It isn't. Intervene now, not at month end. ### The deal that needs you today Whatever else is true — the customer is waiting on you. Start every day with the deals where the customer is waiting on a reply from your side, then look at the ones that are cooling. Cooling fires while a deal still looks healthy in the pipeline — before the stage slips and before the close date moves. ## FAQ No. Everything is written into its own dedicated fields. Your deal stage, amount, close date, and everything else your team maintains are never edited or overwritten. Where the analysis disagrees with your CRM, it records that in its own field and leaves yours alone. On a regular automatic schedule. Each run refreshes the deals that have had new activity since they were last analysed, so active deals stay current without you doing anything. Every deal records when it was last looked at. Because some of it measures time and silence, not just messages. A deal tips into "gone quiet" purely because enough days have passed with no reply — nothing new has to arrive for that to happen. And because warming and cooling are measured against how the conversation *used to* run, the same deal that looked steady last week drifts to cooling and then to silent simply because the pattern changed. The absence is the signal. Yes. It all lands as normal HubSpot deal fields, so you can filter on it, build your own list views, and create dashboards and reports — exactly like any other field in your portal. It won't guess. Everything reflects only what genuinely happened in the data available — if the evidence isn't there, nothing is written. Each deal shows how many conversations the analysis was built on, so you always know how much is behind it. Accuracy is high because nothing stands on its own — every conclusion is backed by the exact message or activity it came from. Nothing is claimed that can't be traced to something real in the conversation or the CRM. You can type over it in HubSpot, but the next scheduled run will replace what you wrote. If you want to record your own judgement permanently, use your own fields or a note on the deal. ## Related The WhatsApp side of BEA Radar — AI analysis written on every chat Filter your Team Inbox by anything the AI has worked out