Getting More Out of Dynamics 365 Sales Insights: A Guide for Teams Already Using It
If your organization has Dynamics 365 Sales Enterprise or Premium, you're likely paying for more AI capability than your team is actually using. Sales Insights isn't one feature — it's a bundle, and each piece needs its own setup and its own habits to pay off. Here's what's in it, what it takes to turn on, and how to actually use it day to day.
Predictive Lead Scoring
0–100 score on open leads, trained on your own closed leads.
Predictive Opportunity Scoring
Likelihood to close, based on historical win/loss data.
Relationship Analytics
Engagement signals rolled into a relationship health score.
Conversation Intelligence
Call recording, transcription, and coaching signals.
Sales Accelerator
Smart Worklist and Sequences — where scoring becomes daily behavior.
Goals
Quota/KPI tracking — not gated behind Sales Insights at all.
Predictive Lead Scoring
This assigns a 0–100 score to open leads based on patterns from your own closed leads — not a generic industry model. It's genuinely useful for triaging volume, but it has a hard data requirement.
Data requirement
Your org needs at least 40 qualified and 40 disqualified leads, closed within the training window you select (3 months to 2 years). If you're a smaller team or a new instance, you won't clear that bar yet — and the setup screen will tell you so rather than silently degrading.
Adoption tip
Don't just display the score — act on it. Use it to drive list views and filters, and wire it into Power Automate to auto-route high scorers to your top closers or trigger nurture sequences for low scorers. A score nobody routes on is just a number on the record.
Predictive Opportunity Scoring
Same mechanism, applied to open opportunities — likelihood to close, based on your historical win/loss data. Same 40/40 threshold applies (40 won, 40 lost, within your chosen window). This is the one to lean on for forecast conversations with managers, since it's grounded in your actual pipeline behavior rather than rep gut-feel.
Worth doing
Both scoring models let admins review and adjust the top influencing factors rather than treating the model as a black box — worth doing once you have a few months of results, since the out-of-the-box factor weighting won't always match how your sales cycle actually works.
Relationship Analytics
This tracks engagement and interaction signals (emails, meetings, response patterns) to surface a relationship health indicator per account or contact. It's less "AI insight" and more instrumentation — it needs your email and calendar actually connected and synced before it has anything to say.
Adoption tip
Set your own thresholds for what counts as "healthy" rather than accepting the defaults. A declining health score on a key account is the trigger to actually schedule a check-in — the feature only pays off if someone's watching it, not just displaying it on a dashboard.
Conversation Intelligence
Records, transcribes, and analyzes sales calls (Teams or other integrated telephony), then surfaces call summaries, keyword/talk-time trends, and coaching signals. This requires an explicit enablement step in Sales Insights settings — it's off by default.
Before you turn it on
Loop in whoever owns compliance and recording consent before flipping this on — call recording has legal requirements that vary by jurisdiction.
Sales Accelerator (Actions, Sequences, Smart Worklist)
This is where scoring actually becomes daily behavior. The Smart Worklist replaces a static to-do list with an AI-ranked queue — factoring in lead score, engagement signals, and active sequence steps — so reps open the app to "who to call next," not a manual list.
Sequences act as structured playbooks, with steps that can auto-trigger off behavior like an email open, a status change, or a stretch of inactivity:
Example sequence
Common gap
If your sequences include an email-open or link-click trigger, Email Engagement has to be separately enabled in Sales Insights settings — it's a common gap when sequences "don't seem to be working."
Goals
Goals is Dynamics 365's quota and KPI-tracking module — set targets (revenue, calls, deals) against a period and roll them up by rep, team, or territory.
Worth flagging clearly
This is not part of the Sales Insights settings area or license gate. It's a separate, longer-standing feature, and you can use it fully with a standard Dynamics 365 Sales license — no Sales Insights required. Don't budget it as part of your Premium/Enterprise AI rollout — it's already available.
Rollout Order That Actually Works
For teams adopting these for the first time, sequence matters more than doing everything at once.
Connect email and calendar first
Relationship Analytics and Conversation Intelligence are useless without live interaction data.
Check your data volume
Do this before promising predictive scoring to leadership — the 40/40 thresholds are a real gate, not a soft recommendation.
Turn on scoring, let it run a few weeks
Then review and customize the influencing factors rather than trusting defaults indefinitely.
Only then build Sequences
Sequences built around the scores — built on an untuned or immature model — just automate bad prioritization faster.
The failure mode to watch for
Teams enable everything in one sprint, show it in a demo, and then usage drops off because none of it was wired into an actual daily workflow. The AI here is only as good as the routing and habits built around it.
Related reading
AI TransformationWhere Copilot for Sales and Finance actually earns its keep
Copilot licenses get bought in bulk all the time. Adoption that actually changes how people work is rarer — here's where we've seen it stick.
AI TransformationThe seven-stage AI transformation journey for Dynamics 365
From a clean D365 foundation to multi-agent orchestration — how enterprises actually get to agentic AI, one stage at a time.