You've got a list, a window, and a dialer. What you don't have yet is a clear answer to how many phone numbers you actually need loaded in that dialer — and how to rotate them so calls three and four still connect at the same rate as call one. Get this wrong and your first hour looks great, your third hour looks like static.
This is the math and the operational rules. No protocol lectures.
Start with dials per number per day, not total volume
Everyone starts from the wrong end. They take the list size, divide by the window, and buy numbers to match. That gets you a block that's either wildly oversized or, more commonly, so undersized that every number is smoking by lunch.
Start here instead: how many dials can one number make in a day before it starts getting flagged?
There is no universal number, but working ranges most campaign operators land on:
- Cold outbound to mixed carriers: 60–80 dials per number per day is a safe ceiling. Push past 100 and T-Mobile in particular starts labeling you.
- Warm list (past supporters, opted-in): 100–150 is workable because answer and talk rates are higher, which the analytics engines read as legitimate.
- GOTV final 72 hours: you can push harder because you're burning the block on purpose — more on that below.
Once you pick your per-number ceiling, the block size falls out of it:
numbers_needed = (dials_in_window) / (per_number_daily_ceiling * days_in_window)
A 400,000-dial GOTV push over three days at 80 dials/number/day needs roughly 1,650 numbers. Not 200. Not 5,000. That's the ballpark you're negotiating a bulk DID order around.
Match area codes to the list, but don't over-engineer it
Local presence works. A call from a matching area code answers at a meaningfully higher rate than an out-of-state number. But there's a point where the effort of sourcing exact-NPA-NXX matches for every county in a state costs you more than you gain.
Practical rule: match to the area code, not the exchange. If you're calling a list heavy in 313, 616, and 810, buy inventory in those NPAs. Don't chase rate-center-level matching unless you're running a program where the same voter sees your number on caller ID more than once — then the specificity starts to matter for recognition.
More on the tradeoffs in local presence dialing with bulk numbers.
Rotation: the part everyone gets wrong
Most dialers ship with round-robin rotation on by default. That is not a rotation strategy. That is a setting.
What actually protects answer rates:
Rotate by cohort, not by call
Split your block into cohorts of 50–200 numbers. Assign a cohort to a shift, a region, or a script. When a cohort's answer rate drops below your threshold (more on that in a second), pull the whole cohort out and swap in a fresh one. Rotating one number at a time means you're always mixing healthy and burned inventory, which drags your overall rate.
Set a kill threshold and enforce it automatically
Baseline your fresh-number answer rate in the first hour of a shift. If a number's answer rate drops more than 30–40% below that baseline, it's flagged somewhere. Pull it. Don't wait for it to "recover." It won't during this cycle.
Some operators track this manually in a spreadsheet. Don't. Your dialer or your termination provider should be surfacing per-number connect rates in near-real-time so you can pull dead numbers on the fly. If yours doesn't, that's a bigger problem than block size.
Rest, don't retire
A number that got labeled "Spam Likely" today isn't necessarily dead in six weeks. Move burned numbers to a rest pool. If your cycle is long enough, some of them come back clean. In a 90-day sprint they won't — plan to churn.
Reserve inventory: the mistake that ends cycles
Every campaign operator I've talked to who had a bad week had the same story: they ordered exactly the block they thought they needed, ran it hot from day one, and by day four had nothing clean to rotate in.
Order a reserve pool of 20–30% on top of your working block. Keep it cold — no dials, no test calls, no SMS. When your working block starts degrading in week two, you have fresh inventory sitting there. This is the single highest-ROI decision in the whole plan and the one people cut when the budget conversation gets tight.
More on this in rotating flagged numbers and reserve strategy.
Timing the burn against your window
If you're running a 12-week program, you protect the block. If you're running the final GOTV weekend, you burn it — on purpose.
Long programs (persuasion, fundraising, member outreach)
Stay well under the per-number ceilings. Rotate cohorts weekly, not daily. Register CNAM on your primary block so the display name shows up consistently. Treat the numbers as an asset.
Short programs (GOTV, last-week fundraising)
Buy more numbers than you technically need for the volume, push each one harder, and accept that you're not using this block again. What matters is the answer rate on day one and day two — not whether the numbers are usable in December.
Size the block so any single number handles the window without hitting the ceiling before the polls close. Two hundred fifty dials over three days per number is fine. Two hundred fifty dials in six hours on the last Sunday is not.
Attestation and carrier trust, in one paragraph
Since I promised no protocol lectures: carriers score every call for how confident they are the caller is who they say they are. "A-attestation" means the carrier signing your call has verified you own the number — the highest trust level. Ask any SIP termination provider you're evaluating what percentage of your traffic will sign at A-attestation. If they can't answer, or the answer is below 90%, your block is going to burn faster no matter how well you rotate it.
What to do next
Before you place a number order, write down four numbers on paper:
- Total dials in the window
- Length of the window in days
- Per-number daily ceiling you're willing to hit
- Reserve percentage (aim for 25%)
Run the math. Then take that number to a provider and ask for a rate sheet — not a marketing deck. Compare apples-to-apples on origination, termination, attestation percentage, and how fast they can swap burned numbers out mid-cycle. That last one matters more than most people realize until they need it at 9 PM on a Saturday.
A dialer full of the right-sized, well-rotated block will out-perform a bigger block that's mismanaged, every single time.