An SDR with 200 accounts is not doing 200 tasks. They are running 200 small projects, each with a state, a next action, and an end. The tooling presents it as a task queue, which is why the work feels endless and the reporting measures the wrong thing.
Borrow three ideas from project management. They are not a methodology, and adopting them takes a week.
Accounts have states, not positions in a list
A task list answers "what is next". It cannot answer "how is the portfolio doing", which is the question a manager actually has.
Give every account one of a small number of states:
| State | Means | Exit |
|---|---|---|
| Unqualified fit | On the list, not yet checked against the ICP | Checked, in or out |
| Waiting on signal | Fits, no reason to call today | A trigger fires |
| In cadence | Actively worked | Reply, no, or exhausted |
| Held | Real reason to pause, with a date | The date arrives |
| Closed out | Done for now, with a reason | Re-entry window |
Five states, mutually exclusive. The value is not the taxonomy, it is that every account has exactly one and can only be in it for a defined reason.
Most territories, examined honestly, turn out to be 70% "waiting on signal" mislabelled as "working". That single discovery usually changes how a rep spends a week.
Limit work in progress
The classic project management finding transfers directly: too many things in flight slows all of them.
An SDR running 150 accounts in cadence simultaneously is not working 150 accounts. They are touching whichever ones the tool surfaces, which is not the same as a decision.
Set a WIP limit on the "in cadence" state. Thirty to fifty is a common landing place, and the exact number matters far less than having one. When the limit is full, nothing new enters until something exits, which forces the exhaustion exit to actually exist. That is the same discipline as the defined cadence end in cadence design.
The uncomfortable part: a WIP limit makes the backlog visible. Teams often discover the territory is three times larger than the capacity to work it, which is a staffing or targeting conversation that was previously hidden inside "the rep is busy".
Every account has one next action with a date
Not a sequence step. The next human decision, and when it happens.
"In cadence, day 5 of 7" is a tool state. "Call the VP after their conference ends on the 14th" is a next action. An account with no next action and no date is not being worked, whatever the activity log says.
This is the check that finds the quiet dead weight: pull twenty accounts and ask for the next action and its date. The ones without an answer have been consuming list capacity for weeks.
The weekly review that replaces activity reporting
Activity reporting asks how many calls and emails. It measures effort, it is easy to game, and it tells a manager nothing about whether the portfolio moved.
A state review asks four questions:
- What changed state this week, and why? Movement is the signal.
- What did not change state, and how long has it been static? Static
accounts are the real finding, and they are invisible in activity reports.
- What is at the WIP limit and should be exited?
- What entered, and on what signal?
Fifteen minutes per rep. It surfaces stalled portfolios that a dashboard of call counts hides completely, because a rep can hit every activity number while nothing moves.
Related: SDR manager training for the coaching cadence this sits inside.
What this is not
Not a new tool. Every state above can live in a CRM field and a saved view. Buying software to fix a definition problem is how teams end up with three systems that disagree.
Not project management ceremony either. No stand-ups, no burndown charts, no sprint. Three ideas: states, a limit, and a dated next action.
Where this sits
This is the Motion pillar of the Tenbound Pipeline Architecture Standard: how the work gets sequenced and run. Salesforce frames the SDR role as opening conversations and handing them to closers. Portfolio thinking is what makes that repeatable across two hundred accounts instead of whichever ten are top of mind.
Place your team on the maturity model. Teams at Manual run task lists. Teams at Orchestrated run states, because a system cannot supervise a to-do list.