How do you score broken lead routing when sales on Outreach and leadership only reviews ARR waterfall monthly on Dynamics 365 ?

The Tuesday meeting where nobody is wrong

Picture the monthly revenue review. Finance pulls up the ARR waterfall in Dynamics 365: beginning ARR, new business, expansion, contraction, churn, ending ARR. New business came in soft — call it fifteen percent under plan. The CRO turns to the demand gen lead, who has receipts: MQL volume was up, cost per lead was flat, form fills grew. Then the sales leader speaks, and he has receipts too: his reps were at target on Outreach activity — sequences enrolled, calls dialed, connect rates holding. Everyone in the room is telling the truth, and the number is still wrong.

This is the signature of broken lead routing, and it is almost impossible to see from the waterfall alone. The waterfall is a financial summary of decisions made eight to twenty-four weeks ago. Routing is a decision made in the ninety seconds after a form submits. Between those two clocks sits an enormous blind spot, and because nobody owns the blind spot, nobody scores it.

The specific pain in this stack is temporal. Sales lives inside Outreach, where the unit of work is the sequence and the unit of time is the day. Leadership lives inside the Dynamics 365 ARR waterfall, where the unit of work is the segment and the unit of time is the month. Neither system natively produces the metric that connects them. Outreach can tell you a rep enrolled forty leads this week; it cannot tell you those were the wrong forty. Dynamics can tell you new business ARR closed at $840K; it cannot tell you $130K of that gap traces to leads that sat unassigned over a holiday weekend in April.

I have watched teams spend two full quarters arguing about lead quality when the actual problem was a routing rule with no fallback branch. A new paid channel started delivering leads with a blank country field. The territory rule required country. Leads with no country matched no rule, so they landed in a default holding queue that no rep had pinned in their view. Four hundred leads accumulated over eleven weeks. Nobody noticed, because every dashboard anyone looked at was denominated in things that still looked healthy: total leads created went up, rep activity stayed flat, pipeline coverage looked fine on a trailing basis. The waterfall finally caught it in September, for leads that arrived in June.

The scoring problem, then, is not "how do we know routing is broken." It's "how do we produce a number, weekly, that a finance-minded leadership team will accept as a predictor of a waterfall bucket they care about." That reframing matters more than any specific field or formula, because it decides what you build. A RevOps team that builds a routing diagnostic builds a dashboard nobody opens. A RevOps team that builds a routing forecast gets invited back to the monthly review.

Adjacent versions of this same problem show up everywhere in the stack, and the same scoring pattern solves them. Call recordings that never attach to an opportunity, so coaching data is invisible in the funnel. Marketing-sourced pipeline that gets re-attributed at close and quietly disappears from the channel it came from. Renewal tasks that fire into a CSM queue where the CSM left the company. All of them are silent-stoppage failures with a long lag before they surface in a financial report. All of them get fixed the same way: define the correct-path condition, measure the percentage that fails it, and translate the failure rate into a dollar projection against the metric leadership already reviews.

How routing failure actually propagates into the waterfall

There are three distinct failure modes, and they produce different signatures. Conflating them is why most routing audits produce a vague "we should clean this up" and no action.

Ghost leads — routed, never touched. The assignment fires correctly. The lead lands in a rep's queue in Dynamics. The rep never opens it. In Outreach, there is no sequence enrollment, no email, no call. This is not a systems failure; it's a capacity or prioritization failure, and it usually clusters. One rep with a big open pipeline stops working new leads entirely for three weeks while chasing a big deal. Or an SDR goes on leave and their queue keeps filling. The waterfall signature is a flat new-business pipeline creation line despite rising lead volume — you added twenty percent more leads and created the same number of opportunities.

Wrong queue — routed to the wrong segment. Enterprise-shaped leads land in the SMB queue and get a five-touch templated sequence instead of an account-based approach. Or SMB leads land with an enterprise AE who ignores them because they don't clear his deal-size bar. The waterfall signature is elevated early-stage closed-lost with a short lag, plus a quieter long-term effect: your enterprise new-business bucket underperforms while your SMB bucket looks fine on count and weak on average deal size.

Stale loops — routed, re-routed, never resolved. Geo mismatch, an owner who declines, a round-robin that reassigns on inactivity, a manual reassignment with no logged reason. The lead bounces. Each hop resets the clock in someone's mind but not in the buyer's. The waterfall signature is aging: deals stuck in qualification for sixty-plus days with no stage change, eventually purged in a pipeline hygiene sweep, which makes them look like a sales execution problem rather than a routing problem.

To score these you need three fields reconciled in one place, and Dynamics 365 is the right place because that's where the waterfall already lives. First, the Outreach first-activity timestamp, synced back onto the lead record — most Outreach-to-Dynamics integrations can write activity back, and if yours only writes a task record you can derive the earliest task datetime with a rollup or a scheduled flow. Second, the assigned queue or owner at time of routing, captured at assignment rather than read live, because live owner reads hide re-routing entirely. Third, an ICP or fit tier on the lead, from whatever scoring model you already have — even a crude three-bucket firmographic tier is enough to detect segment mismatch.

The critical design choice is capturing owner-at-assignment as a separate stamped field. If you only read the current owner, a lead that bounced four times looks identical to a lead that routed cleanly. Add an assignment counter that increments on every owner change within the first seven days, and stamp the first assigned owner and timestamp with a Power Automate flow that fires on create. This is ten minutes of configuration and it's the difference between a scoring model that sees stale loops and one that is blind to them.

Real numbers: thresholds, lags, and what a dollar of routing failure costs

Concrete thresholds beat qualitative judgments, because a leadership team that reviews a waterfall monthly wants a number with a line under it. Here are the ones that hold up in practice, with the caveat that you should recalibrate against your own baseline after four to six weeks of measurement rather than adopting them blind.

The 48-hour touch window. Speed-to-lead research has been remarkably consistent for over a decade: response inside the first hour dramatically outperforms response on day two, and the curve is steep early. For scoring purposes, though, one hour is too tight a threshold — it produces so many failures that the metric becomes noise. Forty-eight hours is a defensible floor. If more than fifteen to twenty percent of assigned leads have zero Outreach activity at the 48-hour mark, you have a capacity or prioritization problem worth naming in the monthly review. Under five percent is healthy. Between five and fifteen is the yellow band where you watch the trend rather than escalate.

Segment mismatch tolerance. If more than ten percent of leads assigned to a given queue carry an ICP tier that belongs in a different queue, the routing logic has drifted from the segmentation the waterfall uses. This one is worth checking in both directions — enterprise leads in SMB is the loud failure, but SMB leads clogging an enterprise queue quietly destroys the enterprise rep's throughput and shows up as an activity problem instead of a routing problem.

Re-route rate. More than one owner change in the first seven days should be rare. Above five percent of volume, you have a rule gap, not a judgment call. Reps reassigning leads is normal; reps reassigning leads *without a logged reason* is what makes the pattern uncountable.

The lag structure. This is the part most teams get wrong, and getting it right is what makes the metric credible to finance. A routing failure today does not appear in this month's waterfall. Map the lag explicitly for your own motion, but a common shape for a mid-market B2B cycle looks like: lead routed in week zero, first meaningful Outreach engagement in weeks one through two, opportunity created in weeks four through eight, closed-won in months two through six. That means the new-business column of any given month's ARR waterfall reflects routing decisions made roughly two to six months prior. Closed-lost at early stages moves much faster — two to four weeks — which makes it the first waterfall bucket where routing damage becomes visible.

The composite score. Combine the three checks into one weekly percentage. Call it whatever your organization will actually say out loud — routing health, clean-route rate, lead velocity. The formula is simply the share of leads routed in a given week that met all three conditions: correct queue for tier, first Outreach activity inside 48 hours, and no more than one owner change in seven days. A single denominator, three pass conditions, one number. Below seventy percent is a fire. Seventy to eighty-five is a managed problem. Above eighty-five, spend your attention elsewhere.

Translating to dollars. This is where the metric earns its place in the monthly review. Take your historical lead-to-pipeline conversion rate and average opportunity value from Dynamics. Multiply the count of failed-route leads in a period by that conversion rate and that value, then apply a recovery haircut — a ghost lead touched late is not fully lost, it's degraded. A conservative haircut assumes you recover somewhere between a third and half of the value on a late touch, and close to none on a lead that was never touched at all. The output is a sentence: "Four hundred and twelve leads failed a routing check this month. At our historical conversion and average deal size, that's roughly $X in new-business pipeline at risk, landing in the waterfall two to four months out."

Be honest about the precision of that number. It is a directional estimate built on your own historical rates, and you should say so in the meeting. A range presented as a range survives scrutiny; a point estimate presented as fact gets torn apart by the first CFO who asks how you calculated it, and then the whole metric loses credibility. State the assumptions on the slide.

The reporting cadence that actually works. Score weekly, review the trend monthly. Sales ops sees the number every Monday with a breakdown by failure pattern and by owner. Leadership sees a single line on one slide in the monthly waterfall review, positioned immediately before or after the new-business column, with a two-to-six-month forward projection attached. That adjacency is deliberate — you want the routing number physically next to the bucket it predicts, so the causal story tells itself.

Trade-offs: what to build, what to buy, and what to leave alone

Not every routing problem justifies the same investment, and the wrong-sized response is its own failure mode. Here is the honest trade space.

Native Dynamics views and a scheduled flow. Cheapest path. Stamp the assignment fields with a Power Automate flow on create and on owner change, build two or three saved views for the failure conditions, and email a weekly rollup. This gets you eighty percent of the value in a week of work. Its weakness is that it's brittle to schema changes and invisible when it breaks — which is exactly the silent-stoppage failure mode you're trying to fix. If you build this, build a liveness check with it: a flow that alerts if the weekly rollup didn't send, so a dead metric announces itself rather than quietly reporting nothing.

Power BI on top of the Dataverse. More work, much better for the monthly review, because you can put the routing score and the ARR waterfall on the same canvas with a shared date slicer. The lag makes this genuinely useful: overlay routing health against new-business pipeline creation offset by your measured lag and the correlation becomes visually obvious, which does more persuasive work in a leadership meeting than any table. Cost is real BI effort and someone who maintains it.

Dedicated routing tooling. Purpose-built lead routing and scheduling products exist and handle round-robin, capacity weighting, holdout logic, and SLA enforcement far better than hand-rolled rules. The trade-off is that adding a tool to fix an observability problem often makes observability worse in the short run, because now the routing decision happens outside your CRM of record and you have another integration to reconcile. Buy it if your routing logic is genuinely complex — multi-product, multi-geo, capacity-weighted. Don't buy it to fix three broken rules.

Fixing the rules versus scoring the failures. There's a real argument that if you know the rules are broken you should just fix them and skip the measurement apparatus. Sometimes that's right — if your audit finds three rule gaps causing eighty percent of failures, fix them and move on. But the reason to build the score anyway is durability. Rules drift. New lead sources appear, new territories get carved, a field gets renamed in a Dynamics upgrade. Without a standing measurement, you rediscover the same class of problem in nine months, from the waterfall, two quarters late.

The adjacent decision worth flagging: whoever owns this score must own it by name. Routing health is the kind of metric that dies when it belongs to "RevOps" generically, because the person who built it moves on and nobody notices the flow stopped running. Name a DRI on the slide. That single line does more for the metric's survival than any amount of dashboard polish.

One more alternative worth considering seriously: fixing the review cadence itself rather than building a predictor for it. If leadership only reviews the ARR waterfall monthly and that's the sole forum where revenue problems surface, the routing score is a patch on a governance gap. Some teams get more leverage from adding a fifteen-minute weekly pipeline-creation check to an existing sales leadership meeting — no new tooling, just a standing agenda item on leading indicators. The score still helps, but it does less heavy lifting when the organization isn't relying on one monthly look to catch everything.

Pitfalls that quietly kill this metric

Scoring activity instead of outcomes. The temptation is to report Outreach sequence enrollments as the health metric, because it's the easiest number to pull. It's also the number reps can satisfy without doing the work — enroll everything in a generic sequence and the number looks perfect while segment mismatch runs wild. Score the condition you actually care about, which is the right lead reaching the right rep with a real touch.

Building the metric on a live-read owner field. Covered above, but it's the single most common implementation error. Reading current owner instead of stamped assignment history makes stale loops completely invisible, and stale loops are often the largest bucket.

Shadow spreadsheets. If the routing score lives in someone's Excel file that gets refreshed manually before each monthly review, it will be wrong within two quarters and nobody will know when it stopped being right. Anything leadership reviews should come from the CRM of record. This is worth being stubborn about even when the spreadsheet is faster.

No rollback plan on rule changes. Routing rules interact. Fixing a geo fallback can silently starve a round-robin. Change one rule at a time, document what it was before, and watch the score for a full week before the next change. Two simultaneous rule changes and a moved number teaches you nothing about which one worked.

Presenting diagnostics instead of dollars. Leadership reviewing a monthly waterfall does not want a routing log. They want three sentences: this is what routing health was, this is what it implies for the new-business column two to four months out, this is what changes if we fix it. Everything else belongs in the appendix.

Silent stoppage. The failure that started this whole problem is the failure most likely to end it. A scoring flow that stops firing produces no alarm — it just produces nothing, and nothing looks like calm. Every automated piece of this needs a staleness check: if the score hasn't updated in eight days, someone gets paged. Build that in the same edit as the score itself, not as a follow-up ticket.

Fixing the instance instead of the class. When you find four hundred leads stuck in a holding queue, the urge is to reassign them and declare victory. Reassign them, yes — but the fix is the missing fallback branch and the alert that fires when the holding queue exceeds a threshold. Verify both directions before calling it done: prove the alert fires on a bad case and stays quiet on a known-good one.

Related questions

How long before a routing fix shows up in the ARR waterfall?

Expect early-stage closed-lost to improve within two to four weeks, new-business pipeline creation within four to eight weeks, and closed-won ARR within two to six months. Set that expectation explicitly when you propose the fix, or the first unchanged waterfall reads as failure.

Can we score routing without an Outreach-to-Dynamics activity sync?

Partially. Without activity write-back you can still score queue-tier mismatch and re-route rate from Dynamics alone. You lose the ghost-lead check entirely, which is usually the biggest bucket. Getting activity timestamps synced back should be the first integration priority.

Should the routing score be reported by rep?

To sales ops, yes — ghost leads cluster by owner and that's actionable coaching. To leadership in the monthly waterfall review, no. Aggregate it. Naming individuals in a finance forum turns a systems conversation into a performance conversation and the fix stalls.

What if leadership refuses to add a weekly metric?

Don't fight the cadence. Keep scoring weekly internally and bring only the monthly aggregate plus the dollar projection to the existing review. The trend line across four weekly points on one slide gets you the same conversation without asking anyone to change their calendar.

Does this approach work for renewals and expansion, not just new business?

Yes, with different fields. Swap first-touch-within-48-hours for renewal-task-acknowledged-before-T-minus-90, and queue-tier match for account-owner-still-employed. The pattern holds: define the correct path, measure the failure share, project it against the waterfall bucket it feeds.

FAQ

What exactly counts as a "broken" route in this scoring model?

A lead fails the check if any of three conditions is true: it was assigned to a queue whose tier doesn't match the lead's ICP tier, it had no Outreach activity within 48 hours of assignment, or its owner changed more than once within seven days. One denominator, three pass conditions. Everything else — lead quality, sequence content, rep skill — is a different problem measured somewhere else, and keeping the definition narrow is what makes the number trustworthy over time.

How do I audit the current setup without buying anything?

Export the last thirty days of leads from Dynamics 365 with owner, created date, and any synced Outreach activity timestamps. Sort by created date, flag every record with no activity inside 48 hours, then cross-tab assigned queue against ICP tier. A manual pass over a few hundred records usually surfaces the two or three rule gaps causing most of the damage. Do this before building any automation — the audit tells you whether you need a standing metric or just a fix.

Why measure weekly if leadership only reviews monthly?

Because the routing signal moves weekly and the waterfall moves monthly, and you need enough data points to show a trend rather than a single reading. Four weekly points on one slide make the direction obvious. One monthly reading is indistinguishable from noise, and a metric that can't show direction won't survive its third review.

What's the minimum viable version of this?

One stamped field for first assignment, one for first Outreach activity, a saved Dynamics view filtering leads with no activity after 48 hours, and a weekly email with the count and the percentage. That's a day of work and it catches the ghost-lead pattern, which is usually the largest bucket. Add queue-tier matching and re-route counting once the first number has been stable for a month.

How do I keep this from becoming another dashboard nobody opens?

Attach it to a decision. The number should end a sentence that changes something: either we fix the top rule gap this sprint, or we lower the pipeline forecast for the quarter after next. A metric that doesn't force a choice gets ignored no matter how well built. Also name a single owner and wire a staleness alert, so the metric announces its own death instead of quietly reporting stale numbers.

Is this specific to Outreach and Dynamics 365?

No. The stack determines which fields you read, not the shape of the model. Any sales engagement platform that writes activity back to any CRM supports the same three checks. The Dynamics-specific parts are the Power Automate stamping flow and the Dataverse-to-Power-BI path for putting the score beside the waterfall. The scoring logic itself is portable.

Sources

• https://learn.microsoft.com/en-us/dynamics365/sales/
• https://learn.microsoft.com/en-us/power-automate/getting-started
• https://learn.microsoft.com/en-us/power-bi/fundamentals/power-bi-overview
• https://support.outreach.io/
• https://hbr.org/2011/03/the-short-life-of-online-sales-leads
• https://www.salesforce.com/resources/articles/lead-management/
• https://www.gartner.com/en/sales/topics/revenue-operations
• https://learn.microsoft.com/en-us/dynamics365/sales/manage-leads

Related on PULSE

• [How do you score broken lead routing when no dedicated RevOps hire yet and leadership only reviews ARR waterfall monthly on Dynamics 365 ?](/knowledge/q10148)
• [How do you score broken lead routing when parent-company rollup reporting and leadership only reviews ARR waterfall monthly on Dynamics 365 ?](/knowledge/q10008)
• [How do you score call recordings not tied to opps when sales on Outreach and leadership only reviews ARR waterfall monthly on Dynamics 365 ?](/knowledge/q10078)
• [How do you score call recordings not tied to opps when no dedicated RevOps hire yet and leadership only reviews ARR waterfall monthly on Dynamics 365 ?](/knowledge/q10358)
• [How do you score call recordings not tied to opps when parent-company rollup reporting and leadership only reviews ARR waterfall monthly on Dynamics 365 ?](/knowledge/q10218)