1. Home
  2. Guides
  3. Web development shops
  4. Playbook

Playbookfor web development shops

Web development shop calling playbook: call blocks, cadence and KPIs

Updated September 28, 20264 min read2 primary sources

An SDR standing at his desk mid-call, gesturing as he talks

The short answer

This playbook covers how a web shop runs its calling: prospect lists built one observable finding per row, maintenance and hosting renewal blocks, the honest churned-client call, seasonal windows by vertical, list hygiene, cadence rules, dispositions, three-line sessions with AI summaries, and the KPI ranges to plan the week around.

Step by step

  1. 1

    Build prospect lists one finding at a time

    Every row carries one specific, publicly observable site issue: the sideways menu PDF, the expired SSL, the hours contradicting the map listing. The finding is the opener; a list without findings is a list that gets hung up on.

  2. 2

    Run the local prospect block

    Work the list on three lines, finding first, business cost second, seasonal deadline third. Book consults around a working mockup, and sell the small fix where the budget only fits the small fix.

  3. 3

    Work the maintenance renewal block before renewal dates

    Call care-plan clients with the year's work summarized: updates applied, backups verified, the October emergency. The summary records the renewal decision and the scope conversation for the year ahead.

  4. 4

    Run the churned-client call once, honestly

    Clients who left eighteen months ago get one call: what changed since the redesign, is the site still theirs, anything worth a look. Attempt caps keep one honest call exactly one.

  5. 5

    Review findings and summaries weekly

    Vague findings mean a rushed list; coach them. Verify decision structure and deadlines against recordings before consults get staged.

What this playbook covers

A web shop’s calling has three motions: prospecting local businesses with one observable finding per row, protecting the maintenance and hosting base that pays the quiet bills, and re-engaging churned clients once, honestly. This playbook structures the week around all three, with the hygiene and cadence rules that keep owner-operators answering. It assumes a three-line browser dialer on your own Telnyx numbers, a caller doing the talking, and recordings with after-call AI summaries on every connect.

List hygiene, finding-first

The prospect list is built finding by finding: each row names one specific, publicly observable issue, the PDF menu, the expired certificate, the hours that contradict the map listing, plus the owner’s name where known. Vague findings, “outdated site,” “not mobile friendly,” produce hang-ups, so the weekly list review grades finding quality before any dialing happens. Dedupe across verticals, apply your internal DNC list, and set attempt caps per list.

Two craft rules keep the calling clean. First, describe only what any visitor can see on the public site; no scanning, probing or testing beyond the visitor’s view. Second, honor every stop request the moment it lands. Most of this calling is business-to-business and much of it sits outside the FTC rule under the B2B exemption at 16 CFR 310.6, but the exemption is narrow and owner cells are personal phones, so the TCPA’s wireless restrictions at 47 CFR 64.1200 still govern. Human-dialed lines are the design.

The calling week in blocks

Local prospect block, two to three mornings. The finding-prepared list on three lines, mid-morning before the lunch rush. Finding first, business cost second, seasonal deadline third, who signs fourth. The close books the consult around a working mockup, and the small fix gets sold where the budget only fits the small fix, because a fixed menu page this month is the redesign conversation next year.

Maintenance renewal block, before renewal dates. Care-plan clients get called with the year’s work summarized: updates applied, backups verified, the October emergency nobody invoiced. The call prevents the silent churn of clients who forgot what the plan does, and it doubles as the scope conversation for the year ahead: the client who wants e-commerce in spring is a project, not a renewal. The recording settles the annual dispute about what the plan was supposed to include.

Churned-client block, monthly, small. Clients who left a year or more ago get one honest call: what changed, is the site still theirs, anything worth a look. The outcomes get logged: migrated to a cheaper host and now broken in ways the owner notices, a new marketing person who wants a vendor they can call, or the business closed, which ends the record permanently. Attempt caps keep the one honest call exactly one.

Seasonal block, six to eight weeks ahead of each vertical’s window. Restaurants before tourist season, retailers before holidays, tax offices before April. The deadline is the shared fact, and the summary’s timeline fields drive which rows enter the queue when.

Attempt cadence and the local reputation

Prospects: two attempts two weeks apart at different hours, then a dated re-entry tied to a season or a new finding; a new finding restarts the conversation honestly, and it is how the shop’s second call to a business sounds like service instead of nagging. Renewals: one call per year plus the scope conversation. Churned: one call, ever, per departure. The do-not-call disposition is permanent everywhere, and in a local business community the shop’s manners are a referral asset or a liability, chosen by the cadence.

Dispositions sales and delivery both read

Consult booked, audit delivered, fix sold, maintenance renewed, has a developer, doing it themselves, deferred to next season, churned closed, wrong number, do not call. “Doing it themselves” gets recorded without contempt, because the DIY owner of 2027 is the emergency client of 2028, and the summary’s fields are what make that re-entry call feel like memory. “Fix sold” stays separate from “consult booked” so the shop can see which motion pays for the calling hour.

Three lines and the AI summary workflow

Three lines clear the voicemail layer on a local list, which doubles the conversations per block. After each connect, the AI writes the transcript and pulls the fields: the findings the owner recognized, the business season, who signs, the family dynamics, the next step. The caller verifies deadlines and decision structure against the recording before the next dial, because a consult staged for one spouse when both sign is a wasted Thursday.

The Friday review covers the finding inventory as much as the calls: which findings booked consults, which got fixes, which produced nothing. Over a quarter that inventory becomes the shop’s real lead-gen answer, and it costs one morning block a week to maintain.

KPI ranges and the production benchmark

  • Dials per active calling hour: 70 to 100, median near 85 in DialBreeze production use, 3-line sessions, 90 days.
  • Dials per caller day: 200 to 500 depending on list prep load; finding work happens outside the block, and it shows in the connect quality.
  • Person-connect rate: 12 to 22 percent planned, 17.8 percent production median; mid-morning windows run best for owner-operators.
  • Consults booked per week: your own baseline from the first month, judged by finding type.
  • Maintenance renewals saved: the quiet number that pays the rent; one saved plan covers the calling hour for a month.

The production figures come from 37,411 dials over 90 days to 2026-09-26 in DialBreeze operation, aggregate activity, not customer results and not a promise about your list.

Compliance checklist for the web shop

Human-dialed calls to owner cells, stop requests honored and logged same-day, calling hours respected, recording disclosure for all-party states. Findings discipline: public observations only, no probing. Keep list provenance documented, and when a vertical list is purchased rather than built, the hygiene rules get stricter, not looser. Nothing here is legal advice; run the list policy past counsel where your calling territories are strict.

FAQ

How many dials should a web shop caller make per block?
Plan 70 to 100 dials per active calling hour on three lines, median near 85 measured in DialBreeze production use over 90 days. Because the list is finding-prepared, the quality bar is finding accuracy, not raw volume; a two-hour block with 120 to 180 accurate rows is a strong session.
What connect rate should a local business list produce?
Hold 12 to 22 percent, with 17.8 percent as the production median. Mid-morning blocks reach owners before the lunch rush; measure the windows separately before judging the list.
How many times do we call the same business?
Two attempts two weeks apart at different hours, then a dated re-entry tied to a season or a new finding. Stop requests are honored permanently, and in a local business community, word about the shop that kept calling travels fast.
Do B2B exemptions cover these calls?
Partly. Most B2B calls sit outside the FTC rule's core provisions under 16 CFR 310.6, but the exemption is narrow, owner cells are personal phones, and TCPA wireless restrictions still apply. Human-dialed lines, stop requests honored. This is not legal advice.
What KPIs run a web shop's calling?
Attempts, connects, consults booked by finding type, fixes sold, and maintenance renewals saved. Consult show rate tells you whether the mockup close works. Track which findings convert, because the finding inventory is the shop's actual lead source.

Sources

  1. law.cornell.edu /cfr/text/16/310.6
  2. ecfr.gov /current/title-47/chapter-I/subchapter-B/part-64/subpart-L/section-64.1200

Operational guidance, not legal advice. Rules vary by state and by campaign.

Put the script to work.

Three lines, a recording of every connected call and the notes written after you hang up.

Start a 14-day trialPricing