What is account-level intent scoring?

A score per account, not per lead, built from the signals the account and its people give off: hiring, funding, site visits, LinkedIn activity, replies. DemandCompass keeps two scores, awareness and readiness, because a form fill says an account knows you, not that it is ready to buy.

How is the score calculated?

Each of the 81 signal types has a weight, a decay half-life and a cap. Signals roll up from people to the account with a person-intent gate, so an account cannot reach Sales-Ready on company-level signal alone. Weights are calibrated weekly against outcomes and every version can be rolled back.

Where does the data come from?

Your own surfaces first: LinkedIn, inbox, website, events, CRM and email sequencer. Then third-party sources: hiring, funding, news and filings, tech stack, GitHub. If you already pay for an intent platform, it plugs in as one more source.

What does it integrate with?

HubSpot, Salesforce, Slack, Smartlead, EmailBison, Outreach, GetSales and Extrovert for LinkedIn, Clay, Warmly, MailerLite, Luma and Calendly. Nothing is installed on your side.

Can I buy the software on its own?

No. DemandCompass comes only as part of working with us, operated by the team that wrote the model. The free Demand Map is the self-serve part.

Does it replace my intent platform?

No. 6sense, Bombora and similar tools measure third-party intent at the company level; that feeds the readiness axis. DemandCompass adds the awareness axis from your own surfaces, keeps the evidence behind every point, and routes the account with the reason attached. Your vendor becomes one more source.

DemandCompass: RevOps automation platform

The product

See who’s moving.
Know why. Act accordingly.

Every score traces back to the dated signal that created it. Click any number and see the evidence.

Accounts

The layout, the states and the recommended actions are the product’s. The accounts and signals are illustrative.

Who it is for

Same platform. Three jobs.

Marketing, demand gen and sales look at the same market. Each gets the screen its week needs.

If you run marketing

See which activities actually move brand awareness, and prove it.

Every awareness signal is dated and attributed to the activity that produced it: the post, the event, the newsletter, the ad. You watch awareness climb in the accounts that matter, see which of those accounts later turned ready, and show the board the chain from activity to pipeline with the records attached.

What it answers on Monday

Which activities moved awareness in target accounts this month, and which did nothing?

Which awareness-building program is now feeding pipeline?

Can I prove marketing’s contribution without an attribution model nobody trusts?

Good day, DavidLive

MQL count

1,789 +116 this week

Open the report →

1,745 brand-aware (Nurture)44 sales-ready

Same MQL number your funnel reports. The Compass shows what it is made of.

On pace Month target 1,800 · expected ~1,715 by today

Market awareness · how much of your market knows you

7.1% of your 25,274-lead market is aware of you

+1.3 pts over 5 weeks

5.8%Aug 10

6.4%Aug 17

6.6%Aug 24

7.1%Aug 31

7.1%Sep 7

Funnel momentum · are you outpacing decay?

+250 net new aware-or-ready leads over 4 weeks

+161Aug 17▲187 ▽26

+47Aug 24▲47 ▽0

+41Aug 31▲112 ▽71

+1Sep 7▲1 ▽0

Green: leads your programs moved into brand-aware or sales-ready that week. Orange: leads that cooled back down.

Awareness by activity: posts, events, newsletter, adsMovement from Nurture to Sales-ReadyBoard View with the evidence drawerCoverage of the marketChannel effectiveness

If you run demand gen

Control your programs, and see what reps do with the leads you route.

Programs run from one place: audience, offer brief, copy with guardrails, sequencer, routing rules with daily caps. Once a lead is routed, its state stays visible: opened, contacted, replied, meeting, dropped, and why. No more asking sales what happened to the list.

What it answers on Monday

Are my programs running as designed, and which one moves accounts?

What did the reps do with the leads we routed last week?

Which offer converts, and which one should we retire?

01 · Demand createdLive

TAM mapped & sensed13,537 of 33,808 accounts with any signal

40%

Leads aware of you+0.4 pts in the period · of 25,274 leads

7.1%

Accounts activatedcrossed into In-Market / Sales-ready · 0 cooled

+12

Routed leads worked by repscontacted 22 of 24 · 8 replied · 4 meetings

92%

New MQLs by source · latest signal before crossing

Events62

LinkedIn outreach41

CRM activity30

Newsletter12

Cold email7

LinkedIn engagement3

Field events2

Every bar opens the leads behind it. The pitch to the broad ICP is not on this list any more: retired after 5,521 sends and zero positive replies.

Five-step campaign wizardRouting rules and daily capsLead states after routing, per repSignal ledger, row by rowWeekly calibration against outcomes

If you carry a quota

Full context on why you got the lead, and an AI that helps you work it.

Each routed lead arrives with the reason: which signals fired, when, from whom, and the state of every touch so far. The assistant drafts the first message and the call script from that context, and the feedback you leave on each lead goes back into the scoring, so next week’s queue is better than this one.

What it answers on Monday

Why was this lead routed to me, and what happened before it reached me?

What do I write in the first message, and what do I say on the call?

How do I flag a bad lead so the model stops sending them?

Marcus Adeyemi● Sales-ready

1 of 30 in your queue

Director of Engineering · Cipher AI · cipherai.com ↗ · LinkedIn →

Readiness

10

Awareness

10

Connected with

AAlex · since Aug 17

Conversation · every channel

Alex Rivera · email · Aug 17Marcus, quick one. We map which AI-security accounts are moving toward a purchase from public signals and score them the way you would forecast a book of business. I ran Cipher AI through it. Want the read?

Aug 17 · We sent a LinkedIn invite

Marcus Adeyemi · email · Aug 20Interesting. We are rebuilding how we prioritise accounts this quarter, so send me the model when it is ready and I will take a look.

Aug 26 · Accepted the invite

Draft · Alex Rivera · LinkedIn · Proof / case study · Touch 2Marcus, one thing from the last exchange: a security team about Cipher AI’s size cut their weekly review from four hours to forty minutes by scoring accounts before outbound instead of after. I wrote up how in two pages. Want me to send it, or would twenty minutes on Cipher AI’s own accounts be more useful?

Why now — Clay import Aug 19 · GitHub star, eval repo Aug 7 · Registered for AI Engineer Summit 2026 Aug 7 · All 5 signals ›

Send ↵Edit ERewrite with feedback… RSkip S

Why-now brief with every signal and its sourceLead states and timelineAI-drafted messages and call scriptsSuggested replies in the inboxFeedback on each lead feeds the scoringSlack alert, CRM record, call queue