Define the purchase
Start with the work you need done, not a promise of meetings. The same word, “lead”, can describe a contact record, an interested reply or a held conversation. A good evaluation makes that difference visible and records why an option is accepted, held for evidence or rejected.
Swipe or scroll horizontally to read all table columns.
| Purchased output | What still needs checking |
|---|---|
| Contact records | Accuracy, relevance, permitted use and who will contact them. A database match is not interest. |
| Verified records | Verification checks contact data; it does not establish a buying need. |
| Interested conversations | Read the actual meaning and the role. A polite answer is not qualification. |
| Original bookings | Match original creation, attendance and fit separately; remove repeats and duplicate credit. |
| Managed execution | Write down research, messages, approvals, replies and calendar handoff. Discovery and closing may remain with your team. |
Two completed buyer briefs
An agency can test targeting and messages, but it cannot manufacture a credible offer, proof, accessible market or competent sales response. Clarify the service, buyer, commercial threshold, sales owner and evidence before asking a provider to scale.
If the motion is uncertain, buy a bounded learning test rather than a long volume commitment. The test should define the assumptions, sample size, review cadence and stopping rule.
Swipe or scroll horizontally to read all table columns.
| Area | Minimum condition | Resolve first when missing |
|---|---|---|
| Offer | Specific problem and service boundary | Positioning and packaging |
| Market | Identifiable organizations and buying roles | ICP research |
| Proof | Claims the buyer can defend | Case, method or evidence development |
| Economics | Deal value supports deliberate acquisition | Channel and commercial model |
| Sales | Named owner can respond and lead discovery | Capacity and handoff process |
Brief A: suitable for a managed motion
A fictional B2B stock-planning advisory service has sold its defined offer manually. It wants to reach operations leaders at US distributors in English. It excludes existing clients, previous declines, consumers and markets it cannot serve. Its sales owner can take discovery calls on the approved client calendar, but has no capacity to research and handle outreach. Comparable proof can be reviewed privately with permission; the size of demand is still unknown. Decision: evaluate managed execution, without assuming it will create a fixed number of calls.
Brief B: defer the full agency purchase
A second fictional founder has a broad “growth help” offer, no agreed buyer or defensible outcome, and nobody assigned to take calls. A larger sending programme would not resolve those gaps. Decision: clarify the offer and buyer through the founder’s own discovery, or buy a tightly scoped specialist task if needed. Do not award an agency contract merely because it promises volume.
Compare the same scope
Decide whether you need software assistance, a freelancer, staff augmentation, a focused managed agency, dedicated SDR capacity or a broad multichannel provider. Each model creates different management, risk and ownership requirements.
Ask which work is performed by named people, shared specialists, automation and the client. A proposal that calls everything done-for-you without a responsibility map is incomplete.
Swipe or scroll horizontally to read all table columns.
| Fictional option | Work and client responsibility | Fit for Brief A |
|---|---|---|
| R · research only | Supplies reviewed account research. The client writes, sends, handles replies and runs discovery. | Reject for this brief: the client lacks execution capacity. It could suit another buyer. |
| M · managed email/LinkedIn | Research, individual messages, authorized execution and replies; client approves the list, claims and boundaries, then runs discovery/closing. | Hold: scope fits, but capacity, permission and export evidence below remain open. |
| D · dedicated SDR capacity | A dedicated-capacity proposal; exact channels, supervision, availability and data access require written confirmation. | Hold: dedicated capacity is not automatically better. Need and evidence are not established. |
Compare the operating models: Compare agency and software · Compare agencies and freelancers · Compare SDR and staff augmentation
Evidence gates, not a score
Use a reviewable state for each question. The entries below evaluate fictional option M. An acceptable message sample cannot compensate for missing account permission or unclear handover. These are evidence gates, not a score or an automatic recommendation.
Swipe or scroll horizontally to read all table columns.
| Gate and requested artifact | Fictional observation | State, owner and review |
|---|---|---|
| Scope · one common brief | Email/LinkedIn execution and the client’s discovery role are written. | Accept scope only · buyer · day 2. |
| Targeting · account sample with exclusions | A01 fits the agreed role; A02 is outside the serviceable market and excluded. | Accept sample reasoning · research owner · day 2. |
| Message · sourced claim and approval record | A public event question supports the opening, not an assumed buying problem. | Accept rationale · client approver · day 3. |
| Delivery · actual role and coverage plan | The delivery role is known; available coverage and substitution are not evidenced. | Hold · delivery owner · day 3. |
| Qualification · written acceptance/return rule | Fit and attendance are separate; unknown authority returns to clarification. | Accept definition, not future results · sales · day 3. |
| Permission · data/access authorization | Process description supplied; the required account authorization is unresolved. | Hold · authorized account owner · day 3. |
| Reporting · original-ID cohort example | Original booking, held and fit fields are separate; unknowns remain visible. | Accept report design · analytics/sales · day 3. |
| Handover · export and exit inventory | Proposed fields are visible; permitted export scope is not confirmed. | Hold · buyer/contract owner · day 3. |
Ask for the answer and its proof
Swipe or scroll horizontally to read all table columns.
| Ask | Weak answer → specific answer | Evidence and next question |
|---|---|---|
| Why these accounts? | “Our database is accurate” → “Here is the brief, the source date, one inclusion and one exclusion.” | Check the underlying sample. Which disputed account still needs the client’s judgment? |
| Who performs the work? | “Our senior team” → “These roles research, approve, execute and cover absences.” | Request the actual coverage plan. Capacity remains unknown until supported. |
| What qualifies a meeting? | “They replied positively” → “Fit, attendance and acceptance are recorded separately against this definition.” | Read an accepted and a returned record. Who resolves a disputed criterion? |
| What will we see? | “A live dashboard” → “Here is a matched cohort with cancellations, pending calls and sales returns.” | Ask where the original creation date and unknown source appear. A dashboard alone proves little. |
| What if results weaken? | “We optimize” → “We isolate the disputed stage, check contrary evidence and propose one bounded change.” | Ask for the change log and next owner. Do not infer dishonesty from a pause or thoughtful question. |
Review an individual work sample
Give each finalist the same hypothetical or consented account and ask for the research rationale, buying roles, message route, exclusions, reply handling and qualification decision. Do not require free campaign execution; the purpose is to evaluate judgment and process.
Look for dated evidence, restraint, role relevance and clear assumptions. Reject samples that infer confidential pain, invent compliments, misuse personal data or treat a purchased contact as intent.
- Account inclusion and exclusion rationale
- Source links and retrieval date
- Buying-group role map
- Message with claim source
- Positive, objection and opt-out reply routes
- Qualified-meeting decision with reason
A01 is an invented account in Brief A’s market. A fictional operations manager publicly asks how distributors follow up after a regional trade event. The role and serviceability fit the brief; current need, authority and budget are unknown. A02 is a consumer retailer outside that brief and is excluded. A relevant public post earns a question, not a claim about hidden pain.
Fictional opening
Your public question about following up after the regional trade event seems relevant to the planning work we do. Do you handle stock-planning decisions, or does another colleague own that? We could put together a small example of how a distributor might organize the next planning discussion. Would that help?
Fictional brief follow-up
Hi, just checking whether that example would be useful. If it is outside your remit, no problem.
The fact is the public question in this fictional situation. The hypothesis is that a planning example might help. “You are losing revenue through poor planning” would be unsupported and is omitted. A location, event, role or public post can supply relevant context; common ground does not prove intent. Beespoke usually uses an individual contextual opening and one brief follow-up roughly four days later, considering another relevant channel when appropriate. That is a practice, not a universal cadence or an AI-word detector.
Review the message and follow-up: Cold-email copywriting · Cold-email follow-up · LinkedIn messages
Follow the reply to its next owner
Swipe or scroll horizontally to read all table columns.
| Fictional reply | Clarification / next owner | Decision |
|---|---|---|
| “Possibly useful; I do not own that budget.” | Reply owner asks who handles planning and what scope is relevant; client calendar only after agreement. | Hold qualification. Need, authority and timing remain unconfirmed; return with the missing criterion. |
| “Another team handles this.” | Research owner checks whether the original role was wrong. Seek permission for a useful introduction. | Do not count polite redirection as positive demand or auto-book another person. |
| “Please stop contacting us.” | Reply owner records the applicable person/account stop scope and informs the responsible operator. | Stop. Do not restart through another colleague or channel to bypass the refusal. |
Count originals and outcomes
Define contact, reply, positive reply, booked meeting, held meeting, qualified meeting and accepted opportunity. State how no-shows, reschedules, duplicates, existing pipeline and out-of-ICP attendees are handled.
Fictional reporting cohort: 60 unique eligible contacts first reached on days 1–7; original bookings created on days 1–14; outcomes observed through day 28. Six contacts gave relevant positive human replies, four created original bookings. Automatic replies and two non-positive human replies are outside the six. The records below are invented, not a vendor performance claim.
Swipe or scroll horizontally to read all table columns.
| Record | Original / later event | Outcome at day 28 |
|---|---|---|
| B01 | Created day 2; held day 10. | Qualified; sales accepts an opportunity on day 12. |
| B02 | Created day 3; B02-R1 changes the date; held day 18. | Not qualified: wrong responsibility. Sales returns it with a reason. Still one original booking. |
| B03 | Created day 5; cancelled day 9. | Not held; fit not assessed. Billing or replacement depends on the actual agreement. |
| B04 | Created day 6; meeting scheduled for day 40. | Pending, not a no-show; fit and opportunity unknown. |
| F01 | Follow-up with B01 on day 15. | A repeat conversation, not a fifth acquired lead. |
| Duplicate contact row | Second row for the same B02 contact. | Deduplicate before counting; retain the reconciliation separately. |
Review proof without exposing people
Ask references about the same team shape, channel, geography and service scope you are buying. Discuss response quality, visibility, problems, corrective action and handover—not just the number of meetings.
A real proof review needs scope, period, denominator, method and limits. The card below demonstrates the format only. It cannot prove an agency’s results. A truthful anonymized example or a consented private reference can be useful; public disclosure of a customer’s identity is not the price of credibility. Do not expose prospect details, real event IDs, campaign screenshots or a combination of contextual facts that identifies someone.
Swipe or scroll horizontally to read all table columns.
| Proof-card field | Completed fictional example |
|---|---|
| Comparable scope | Managed email/LinkedIn for Brief A; not a claim about calling or a dedicated employee. |
| Period and denominator | 60 eligible contacts, days 1–7; original creation through day 14; cutoff day 28. |
| Outcomes | 4 originals, 2 held, 1 qualified, 1 accepted opportunity; 1 cancellation and 1 pending call. |
| Method | Original-ID reconciliation, deduplication and sales-recorded fit. Missing fields remain unknown. |
| Limits | Tiny illustrative cohort; no revenue result, causal effect or repeatability established. |
| Permission | Invented records only. Real private verification would need the relevant parties’ permission. No reference is contacted here. |
Assign work and approvals
Swipe or scroll horizontally to read all table columns.
| Work | Responsible role and approval | Exception owner |
|---|---|---|
| Target research | Provider researches; client judges and approves list fit/exclusions. | Disputed fit returns to the client, not automatic sending. |
| Messages and claims | Provider writes individual messages; client approves claims and campaign boundaries. | Unsupported claims are removed before authorization. |
| Execution | Provider executes only within authorized accounts, methods and scope. | Missing permission holds launch. |
| Replies / calendar | Provider handles replies, agreed qualification and client-approved calendar handoff. | Refusal, unclear need or calendar blocker goes to the responsible owner. |
| Outcomes | Client/sales records held, qualified, returned and opportunity status in the agreed system. | Unknown outcomes are not quietly zeroed. |
| Discovery and close | Client runs discovery, proposals and closing. | Ask who covers absence; do not invent staffing or response guarantees. |
Inspect data and platform controls
Review account access, data sourcing, privacy responsibilities, approved claims, platform rules, opt-outs, incident escalation, CRM visibility and the provider's use of subcontractors. No responsible agency should guarantee zero account, deliverability or compliance risk.
Confirm the client owns or can export account lists, research, copy, reply history, meeting context, outcome data and learning. Define access removal and handover before the first campaign.
Swipe or scroll horizontally to read all table columns.
| Request | Fictional ticket C01 | Next check |
|---|---|---|
| Provenance and permitted use | Source categories and retrieval dates supplied; permission scope unresolved. | Authorized data owner reviews the applicable basis and scope. No universal legal conclusion. |
| Validation and suppression | A reviewable invalid/duplicate/refusal workflow is shown. | Ask how an objection reaches every relevant operator; do not inspect private contact exports. |
| Platform/access method | Roles and proposed tools described; account authorization still missing. | Account owner decides. A certificate, manual sending or automation label does not prove zero risk. |
| Execution history / incident | Ask for relevant provider errors and configuration-change records, with identifying data removed. | Use maintained primary guidance for precise requirements; no sending-limit or warm-up recipe here. |
| Escalation and ownership | Buyer and delivery roles are named; coverage/substitution evidence pending. | Hold unresolved controls. No live account, DNS or campaign change follows from this article. |
Separate evaluation from commitment
Use a test long enough to produce meaningful evidence but bounded enough to stop. The contract should name responsibilities, billing events, minimum term, cancellation, ownership, confidentiality, data handling and the review process.
Swipe or scroll horizontally to read all table columns.
| Fictional review point | Evidence / decision | Boundary |
|---|---|---|
| Preparation · review day 3 | Approve brief, exclusions, claims, access and outcome definitions; hold unresolved gates. | Not a promised launch date or a free/short pilot. |
| Observation · cutoff day 28 | Keep the creation cohort intact; B04 remains future/pending. Review the available records, even if few. | A commercial minimum term is separate from when evidence is reviewed. |
| Decision · agreed review day 35 | Retain changes/exclusions and concurrent work; accept, narrow, hold or escalate with reasons. | Follow the signed terms. Beespoke’s approved outbound commitment remains three months; no new exit or performance promise. |
Read a complete sample report
Swipe or scroll horizontally to read all table columns.
| Same fictional cohort | Count / interpretation |
|---|---|
| Eligible / positive | 60 reached; 6 relevant positive human replies. Contact accuracy, interest and fit remain different. |
| Original / held / qualified | 4 / 2 / 1; one accepted opportunity. One cancelled and one future meeting remain separate. |
| Ratios, only for this example | 4/60 ≈ 6.7% original bookings; 2/4 = 50% held; 1/2 = 50% qualified. No benchmark or forecast. |
| Unknowns and learning | B04 outcome unknown; authority issue returned on B02; no revenue or complete source attribution claimed. |
Investigate a weak outcome
Swipe or scroll horizontally to read all table columns.
| Explanation to test | Evidence / contrary evidence | Bounded next action |
|---|---|---|
| Measurement gap | A calendar event disappears while original booking records remain. Stable matched IDs would weaken this explanation. | Analytics owner checks definitions and coverage before rewriting messages. |
| Role or message mismatch | B02 returns for wrong responsibility; one case does not condemn the whole audience. | Research owner reviews the disputed role with the client; one approved segment change only. |
| Delivery or scheduling friction | Inspect provider errors, reply ownership and the correct loaded calendar; do not assume a quiet week proves either fault. | Relevant operator records the blocker, next check and approval needed. Outcome stays pending. |
Normalize the quote without guessing
Add setup, retainer, tools, data, variable fees and client time. Compare fully loaded cost over the same period and use held, qualified outcomes as the common denominator.
Swipe or scroll horizontally to read all table columns.
| Compare the same period and unit | Option R | Option M / D |
|---|---|---|
| Included work | Research output only; client execution and reply time are additional. | Confirm exact research, message, channel, reply and handoff coverage. |
| Quote components | Ask about setup, data/tools and client time. Values not supplied. | Ask about recurring fees, extras, minimum term and capacity. Values not supplied. |
| Outcome denominator | Original/held/qualified outcomes are not established by a quote. | Keep unknown outcomes unknown; do not calculate a return from promised meetings. |
This comparison contains no market prices or real provider quotes. A retainer, per-lead or per-meeting label alone does not establish value or eliminate risk. Use the existing approved pricing/cost guides for current facts, and compare a written scope over the same period before doing arithmetic.
Review scope and current cost facts: Outbound cost guide · Appointment-setting pricing · Current Beespoke pricing
Make handover inspectable
Swipe or scroll horizontally to read all table columns.
| Fictional inventory H01 | Observed / unresolved | Verification owner |
|---|---|---|
| Permitted account research and approved copy | Proposed export fields shown; rights and permitted reuse not yet confirmed. | Client/contract owner holds acceptance. |
| Reply context, original IDs and outcomes | Sample mapping shown without personal data; real transfer access unapproved. | Sales verifies completeness only within authorized access. |
| Suppression and open issues | Stop obligations and unresolved B04 outcome remain visible. | Responsible data/operator role follows actual obligations; no invented deletion policy. |
| Account access and receipt | List the approved handover recipients, receipt check and any removal action under the agreement. | Authorized account owner decides; this example removes no real access. |
Record the finalist decision
- Use one brief for every finalist.
- Interview the delivery lead.
- Verify references for the same model.
- Contract a bounded test with a stopping rule.
Swipe or scroll horizontally to read all table columns.
| Fictional option | Evidence-based decision | What happens next |
|---|---|---|
| R for Brief A | Reject: research-only does not solve the client’s missing execution capacity. | Use only if the client can genuinely own execution; no universal judgment about researchers. |
| M for Brief A | Hold: relevant sample/scope, but capacity, permission and export gates remain unresolved. | Seek the named artifacts; accept only after the required evidence and written terms agree. No automatic winner. |
| D for Brief A | Hold: need for dedicated capacity and its actual coverage are not established. | Clarify the missing work before comparing cost. More channels or staff do not decide it. |
| Any full motion for Brief B | Defer: offer, buyer and sales owner are not ready. | Resolve the brief first. Declining a purchase is a valid evaluation result. |
When managed outbound fits
Beespoke’s managed outbound scope focuses on cold email and/or LinkedIn, with founder-led targeting, research, messages, follow-up, replies, qualification and booking against agreed criteria. The client approves claims and exclusions and owns discovery, proposals and closing. Cold calling is a separately quoted activity according to market; it is not part of the standard email/LinkedIn scope. This does not imply a dedicated full-time SDR or guaranteed volume.
When managed outbound is the right next step · Discuss the evidence
Optional private evaluation notes
Use the completed decision above before opening the notes. Record one anonymous option at a time. The notes organize your own judgment; they do not verify evidence, rank agencies or qualify a lead.
Open private evaluation notes
Anonymous notes stay in this page. Nothing is sent or saved. Reloading clears them. Do not enter customer, prospect or account details.
Replace or clear the current notes and decision?
Your evaluation record
Questions
What should I look for in a lead generation agency?
Look for operating-model fit, a credible delivery team, reviewable research and messaging, explicit qualification, prompt reply handling, record-level reporting, responsible platform and data controls, transparent economics and transferable assets.
How many agencies should I compare?
A focused shortlist of three comparable providers is often enough after operating-model screening. More proposals add little when they sell fundamentally different products or use incompatible outcome definitions.
Should an agency guarantee meetings?
A provider can commit to work, service levels or a carefully defined billing model, but a meeting promise does not remove market, attendance or sales risk. Inspect the qualification and replacement rules closely.
How long should an outbound agency test run?
The period should allow research, launch, response cycles and iteration for the selected market without creating an unnecessary lock-in. Define the evidence threshold and decision date in advance rather than relying on a universal month count.
Does a higher checklist score mean a better agency?
No. Evaluate the evidence and unresolved blockers against your brief. Missing authorization, unclear ownership or an incompatible deliverable is not cancelled out by unrelated strengths.
Can I evaluate proof without revealing customers or prospects?
Yes. Use appropriately anonymized evidence and, where suitable, consented private verification. Never treat confidentiality as permission to publish names, contact records or identifying campaign context.
Is a verified contact the same as a qualified meeting?
No. Contact verification, expressed interest, original booking, attendance, fit and sales acceptance are separate states. Write the definitions and record exceptions and unknowns.
Does an evaluation plan change the minimum commitment?
No. Observation and review dates do not override the signed commercial terms. Confirm scope, commitment, notice and handover in the actual agreement; this guide offers no new contract term.
Sources and method
All situations, options, messages and numbers are fictional. Public buyer guides informed the diligence questions, not verified customer results or universal performance rules. Sources reviewed November 3, 2026.
- MyOutreach — Concrete vendor questions; confidentiality and legal details require their own evidence.
- Vierra — Existing source refreshed for definition, ownership and contract questions.
- Flowd — Verification versus qualification and outcome reporting.
- DoneMaker — Weak versus specific answers; no outcome or risk claim is adopted.
- Leadriver — Operating models, rejection rules, onboarding and cohort reporting.
- Woodpecker — A focused specialist may fit a narrow problem better than full execution.
- Google Search Central — People-first usefulness, rather than a ranking promise.