How to pick a LinkedIn Ads agency: The 12 questions we'd want asked

Picture the pitch deck moment: a slick agency promises leads at $47 a pop. You nod politely, then remember your product costs $40K and takes nine months to close. That "cheap" lead? It's a vanity metric wearing a business suit.

Here's the uncomfortable truth most LinkedIn Ads pitches dodge: cost per lead tells you almost nothing about whether your spend actually worked. What matters is qualified pipeline, revenue influenced, and how much coordination chaos the agency dumps back on your team. A form fill is not a forecast. A click is not a customer.

The right agency understands the difference. The wrong one will happily optimize you into a spreadsheet full of leads your sales team refuses to call. We run LinkedIn ads for B2B SaaS clients, so we have skin in this game. Here are the 12 questions we'd want any prospect to ask us before signing a contract.

1. What's your primary success metric for a $20K+ ACV client?

Use cost per SQL or cost per opportunity pulled from your CRM as the primary metric, with CPL a lower-level metric. LinkedIn's measurement guidance makes the same distinction: CPL isn't useful in a vacuum unless you know what a lead is ultimately worth to the business.

Ask for a redacted live dashboard. It should connect campaign spend to qualified leads, opportunities, pipeline, and closed-won revenue. If reporting lives in a spreadsheet disconnected from Salesforce or HubSpot, the agency is grading itself on the numbers its platform login happens to show.

A useful dashboard should also make it easy for sales and marketing to work from the same definitions. If one team calls a form fill qualified while the other waits for a verified buying project, a low CPL will conceal the unresolved disagreement.

2. How do you push MQL, SQL, and closed-won back into LinkedIn?

A marketing-qualified lead (MQL) is a lead marketing considers ready for further review. A sales-qualified lead (SQL) is one sales has accepted as worth pursuing. Those definitions should come from your CRM, not from LinkedIn form activity.

The agency should have a clear feedback workflow:

• Capture MQL, SQL, opportunity, and closed-won stages in the CRM.
• Map each stage to the appropriate LinkedIn conversion event.
• Send changes back through the Conversions API, or CAPI, without waiting for a monthly upload.
• Check that event names, timestamps, and campaign attribution remain consistent.
• Document who owns the integration when fields or lifecycle definitions change.

LinkedIn supports granular conversion types, including marketing-qualified and sales-qualified leads. Campaign optimization can then follow your sales team's definition of quality. Each form fill can be weighted accordingly.

Follow up with this question: how quickly after a lead becomes an MQL or SQL does that signal reach the ad platform? "We upload a CSV monthly" is not a feedback loop.

3. What attribution window do you use, and why?

The right window depends on your buying cycle. If your typical deal takes months to move from first touch to closed-won, a short platform window will undercount LinkedIn's contribution. An agency reporting only on CPL can't detect that gap, let alone correct it.

Ask the agency to explain how its platform reporting connects to your Revenue Attribution Report and CRM opportunity history. Compare both with self-reported attribution. Understand what each system measures and where its blind spots are.

Also ask whether attribution windows or stage definitions ever change between reports. If they do, disclose the change before anyone compares one period with another. Otherwise, reporting can make performance appear to improve when only the measurement rules changed.

4. Do you use the Revenue Attribution Report, and which CRM do you connect it to?

The Revenue Attribution Report, or RAR, should connect campaign exposure to CRM outcomes. LinkedIn confirms that RAR supports three CRMs: Salesforce, Dynamics 365, and HubSpot.

Ask the agency to show how RAR and the CRM work together:

• Confirm which CRM is connected and who maintains the connection.
• Report pipeline amount, closed-won opportunities, win rate, and days to close.
• Review performance at the campaign and account levels.
• Check whether the contacts associated with an opportunity can be matched correctly.
• Pair platform attribution with a required "How did you hear about us?" field on the demo form.

Treat any single attribution report as one view of the buying journey. The agency should be able to explain why platform, CRM, and self-reported numbers differ and still make a clear budget decision.

5. Show us your campaign architecture for a 6–12 month sales cycle.

Running only bottom-funnel "Request a Demo" ads with no nurture path is a structural mistake for long-cycle SaaS. Most prospects who fit your customer profile will not be ready for a sales call the first time they encounter the company.

In a credible three-stage funnel, cold audiences learn the problem and positioning before engaged prospects receive deeper proof and education. High-intent accounts receive a relevant sales offer. Stop treating cold audiences like they're ready for sales.

Ask for the split between cold, retargeting, brand, and account-based campaigns. The agency should explain the reasoning behind the split, the content assigned to each stage, and how it suppresses current customers and active opportunities when appropriate.

The architecture should also connect to outbound. If an account engages with educational creative, the next sales touch should reflect that context instead of starting a disconnected conversation.

6. How do you report buying-group coverage, not individual leads?

Complex SaaS purchases involve multiple people, so an individual lead is an incomplete unit of measurement. Ask the agency to report whether the campaign is reaching the right roles across the right accounts.

A practical buying-group view should include:

• Account-level reach and engagement.
• The number of relevant roles reached inside each target account.
• Coverage across business, technical, financial, and executive stakeholders.
• Accounts with repeated engagement across campaigns or content stages.
• Gaps where one role is active but the rest of the buying group is absent.

This view should be usable by sales and available outside the advertising report. An account with several engaged stakeholders deserves different follow-up from an account where one junior contact clicked once.

Ask the agency to show how buying-group activity changes campaign priorities, outbound timing, and creative. Lead reporting without an explanation of account coverage amounts to a lead-generation program. An account-based program requires both.

7. How do your LinkedIn audiences feed our outbound and ABM?

A LinkedIn agency with no access to your CRM, sequence triggers, or intent signals will run a silo by default. Paid media and outbound should work from the same account list and positioning. Engagement data should flow between them, too.

Ask what the agency needs to coordinate with your SDRs. We coordinate through the same Clay table. A recent CRO hire or funding round can build a matched audience and trigger an Instantly or Heyreach sequence in the same week. The same applies to a tech-stack change. LinkedIn engagement can then influence which accounts receive outreach and what the message says.

Cold outreach should account for what the prospect may already have seen. A single campaign should connect the ad and landing page with the sequence, even when three vendors produce the work. Plenty of agencies run great paid media. We're glad to work alongside one when outbound and creative are the gap. A shared operating system keeps the client from coordinating every handoff.

8. What research goes into the first creative brief?

An agency writing from your homepage copy will repeat the same claims every competitor makes. Good answers name the actual inputs: customer interviews, win/loss data, sales-call recordings, product documentation, and a competitor positioning audit.

The first brief should define the buyer's problem, the trigger that makes it urgent, the product's differentiated mechanism, and the proof needed to make the claim credible. It should also explain how the message changes for a cold audience, an engaged account, and a prospect already talking to sales.

If the answer is "we just have a feel for it," that answer relies on taste alone. For a technical product, ask who on the team can read your docs and turn product detail into accurate buyer education without flattening it into generic language.

Could your sales team use the same positioning in a live opportunity without having to correct it?

9. How many variants do you ship, and when do you refresh?

Ask for the exact number and assess whether the agency runs controlled tests and knows why creative changes. Controlled experimentation starts with a testable hypothesis. Random variations are production work.

Ask to see an experimentation register containing:

• The hypothesis behind each concept.
• The audience and campaign stage.
• The single variable being tested.
• The budget and test dates.
• The result and the decision that followed.
• The next variation based on what the team learned.

The fatigue signal to watch is frequency climbing while CTR drops. Small retargeting pools will usually need attention sooner than broad cold audiences, so delivery and performance should determine the refresh schedule.

Write into the contract that you own the experimentation register, source files, copy, and performance history at termination. Otherwise, the next team may have to repeat tests you already paid for.

10. Who is Billing Admin on the Campaign Manager account?

Your company should know who owns billing, data, audiences, and campaign history before work starts. LinkedIn's Business Manager guide says the company that manages billing should own the account and recommends addressing ownership in the statement of work.

Put the operating rules in writing:

• Your company owns the ad account and remains Billing Admin.
• The agency receives only the access it needs to do its work.
• Create campaigns and audiences inside your company's account.
• Creative source files, conversion mappings, and naming conventions remain with you.
• The termination plan covers access removal, handoff timing, and final reporting.

Do not wait for termination to discover that an agency account contains the audiences, history, or billing controls you need. A sensible initial runway can give campaigns time to develop, but the contract should still define a clear notice period and transition process.

Your company should retain ownership of the infrastructure while the agency operates the account.

11. Separate media spend from your fee. How is the fee calculated?

The proposal should separate money paid to LinkedIn from money paid to the agency. A percentage-of-spend model can reward higher spending even when efficiency does not improve. We make that conflict explicit in our agency pricing guide.

Ask the agency to itemize:

• Monthly media spend paid to the platform.
• Base strategy and management fees.
• Creative, landing-page, outbound, or data costs.
• Any percentage-of-spend component.
• Performance bonuses and the CRM metric that controls them.
• Rebates, preferred rates, or other financial relationships.

A hybrid model tied to cost per SQL can sound appealing, but long sales cycles make the variable component hard to price. It also fails if sales and marketing have not agreed on what qualifies as an SQL.

On minimums, we recommend $5,000–$8,000 a month, based on our CPM benchmarks. A program covering cold, retargeting, and ABM often needs $10,000–$15,000. The agency should explain what the budget can support instead of squeezing every tactic into an amount too small to test.

Transparent pricing lets you judge whether the operating model fits your program and prevents comparisons between totals that cover different scopes.

12. Who works on our account day to day, and how many accounts do they carry?

Ask for the operator by name and identify the senior pitch team separately. Then ask how many accounts that person manages, who reviews the work, and what happens when the primary operator is unavailable.

For a high-touch engagement, the statement of work should identify each role, the work that role owns, and the expected client involvement. It should also distinguish strategic review from routine production so you know the experienced team's role in directing the account and its attendance schedule for quarterly calls.

If you're a founder without bandwidth for weekly check-ins, ask what the team needs from you. "An hour every Tuesday" is fine. A constant stream of Slack threads is a coordination job you didn't hire for.

And watch what they ask you. Strong questions about positioning, sales stages, technical buyers, account priorities, and internal constraints reveal how the agency will work after the pitch. If it needs you to translate insights between paid media, outbound, and creative, it hasn't solved the coordination problem.

Coordinate LinkedIn Ads, outbound, and creative with Understory

Understory runs LinkedIn ads and Instantly-powered outbound for B2B SaaS clients. A recent CRO hire can trigger outreach. Recent funding rounds and tech-stack changes can do the same. On-staff creative works under the same team.

A single dashboard pulls cost-per-SQL, cost-per-opportunity, and pipeline velocity from your CRM. If you're tired of coordinating a paid media vendor, an outbound vendor, and a freelancer who don't talk to each other, we should talk.

Book an intro call with Understory and bring these 12 questions. We'll answer every one, walk you through a live client dashboard, and show you exactly how paid media, outbound, and creative operate as one system. No slide decks. No vague promises. Just the same operating model we'd use on your account starting week one.